» wen“! ~. llllllllllllll'llllllll lillll l'llllfw M99 LIBRARY 193 10575 3226 ~ Michigan State University This is to certify that the dissertation entitled . Development of Microprocessor-Based Steering Control System For An Apple Harvester Utilizing Non-Contact Sensing presented by C. Bert McMahon has been accepted towards fulfillment of the requirements for ph.D. degree in Ag. Egr. fiwflm Major professor Date 7/0 82’ MS U i: an Affirmative Action/Equal Opportunity Institution 0912771 MSU RETURNING MATERIALS: Place in book drop to remove this checkout from LIBRARIES £gg=:,!=!;_ your record. FINES will be charged if book is returned after the date stamped below. F 3 fl ’" flow-- '3" (2/... ”J: '73“: N> I.’ ~’ 7 J ll DEVELOPMENT OF A MICROPROCESSOR-BASED STEERING CONTROL SYSTEM FOR AN APPLE HARVESTER UTILIZING AN ULTRASONIC SENSING SYSTEM By C. Bert McMahon II A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Engineering 1982 ABSTRACT DEVELOPMENT OF MICROPROCESSED-BASED STEERING CONTROL SYSTEM FOR AN APPLE HARVESTER UTILIZING AN ULTRASONIC SENSING SYSTEM By C. Bert McMahon II An automatic steering control system is needed for the USDA over- the-row-apple harvester because the operator cannot accurately steer the apple harvester for long periods of time. The objectives of this research were to develop a non-contact sensing system and an automatic steering control system for the USDA over-the-row apple harvester. The automatic steering control system was required to accurately steer the apple harvester's front wheels such that each tree stayed within the harvester's allowable zone. This allowable zone was 45 cm wide, 409 cm long and centered on the har- vester's centerline. The tree row was required to be either straight or a smooth continuous curve. A micrOprocessor-based steering control system and a non-contact sensing system were designed to control the steering of the harvester's front wheels. The non-contact sensing system consisted of five sonar units which utilized ultrasonic transducers and circuit boards which were made by the Polaroid Corporation. The sonar units were used to measure the distance from each sonar unit to the tree trunk as the har- vester moved over the tree row. The steering control system used a C. Bert McMahon II microprocessor to perform a proportional control algorithm where the wheels were turned to a steering angle which was proportional to the harvester's position error. This control system used the sonar measure- ment as the feedback signal. A simulation model was developed to simulate the harvester's closed-loop steering control system. The model reasonably predicted the motion of the harvester for a curved tree row and for a row with a step change. The steering control system with a non-contact sensing system was tested using a simulated tree row. The results of the tests showed that for a straight and curved row the steering control system was effective at keeping each tree within the harvester's allowable zone at a harvester ground speed of 0.8 km/h (0.5 mph). During these tests the maximum deviation of the harvester centerline from the tree row centerline was 8.9 cm for the curved row and 5 cm for the straight row. Approved by: .34.... //W Major Professor . flaw/ghee Department Chairman ACKNOWLEDGEMENTS The author wishes to express his sincere gratitude and appreciation to the following: Dr. Thomas H. Burkhardt, the author's co-major professor, for his help in developing the technical content of this dissertation and for his many hours of reviewing and commenting on this manuscript. Dr. Bernard R. Tennes, the author's co-major professor, for his guidance, support, and encouragement given during the development of the steering control system for the USDA apple harvester. IJr. Ajit K. Srivastava, who served on the author's guidance committee, for his assistance in the development of the simulation model and guidance with this research. Dr. P. David Fisher, who served on the author's guidance committee, for his guidance and suggestions for this research and for the knowledge gained in digital electronics due to the courses taught by Dr. Fisher. Dr. Clark J. Radcliffe, who served as the outside examiner of this dissertation, for his suggestions and comments. Richard K. Byler, fellow Ph.D. candidate, fiw~rfis assistance and encouragement during this research. The author wishes to acknowledge that the following undergraduate students and technical staff who helped accomplish this research effort; Joe R. Clemens, Shaun F. Kelly, Mary E. Maley, Mary L. Heyn, William A. Heyn, Paul E. Speicher, Theresa L. VanSlyke and Richard J. Nolthuis. The author wishes to express his gratitude to the United State Department of Agriculture for providing financial support for this research. The author is also very thankful for this opportunity to earn this dotoral degree and for the support from his wife, Sally, and from his parents and grandparents. TABLE OF CONTENTS Page LIST OF TABLES O O O O O O O O O O 0 O O O O O O O O O O O O O 0 v LIST OF FIGURES . ..... . . . . . . . . . . . . . . . . . . vii CHAPTER 100 INTRODUCTION 0 O O O O O O O O O O O I O O O O O O O O 1 1.1 The Need for Automatic Steering Systems . . . . . 3 1.2 Comparison of Sensing Methods . . . . . . . . . . 6 1.2.1 Contact Sensing Systems . . . . ..... 7 1.2.2 Non-Contact Sensing Systems . . . . . . . 8 1.3 ObjeCtive O O O O O 0 O O O O O 0 O O O O O O O O 11 2.0 LITERATURE REVIEW . . . . . . . . .......... 13 2.1 Automatic Steering Control Systems Using Mechanic Contact Type Sensors . . . . . . . . . . 13 2.2 Automatic Steering Control Using Optical Type Sensor . . . . . . . . . . . . . . . . . . . 18 2.3 Automatic Steering Control System Using Buried cable 0 O O O O 0 O O O O O O O O O O O O 22 2.4 Automatic Steering Control Using Spacial POSition senSing O I O O O O O O O O O O O O O O 23 2.5 Sonar Sensor - Ultrasonic Transducer . . . . . . 25 3.0 DESIGN REQUIREMENTS 0 0 O O O O O O O O O O O O O O O 29 3.1 Alignment . . . . . . . . . . . . . . . . . . . . 29 3.2 Tree Spacing and Row Curvature . . . . . . . . . 30 CHAPTER Page 3.3 Operating Environment . . . . . . . . . . . . . . . 30 3.4 Interface with Harvester Steering System and Design Constraints . . . . . . . . . . . . . . 33 4.0 CONCEPTS TO DETERMINE TREE POSITION FOR AN AUTOMATIC STEERING CONTROL SYSTEM . . . . . . . . . . 35 4.1 Method 1 - Two Sonars Used to Determine Tree Position by Triangulation . . . . . . . . . . 38 4.2 Method 2 - One Sonar and One Angle Measurement Used to Determine Tree Position by Triangulation . . . . . . . . . . . . . 40 4.3 Method 3 - Two Sonars Used to Detect the Tree Presence Zones . . . . . . . . . . . . . . . . 43 4.4 Method 4 - Two Optical Sensors and Two Angle Measurements Used to Determine Tree Position by Triangulation . . . . . . . . . . 45 4.5 Method 5 - Two Optical Sensors Used to Detect Tree Presence in Zones . . . . . . . . . . . 48 4.6 Method 6 - Multiple Sonars Used to Measure Tree Position . . . . . . . . . . . . . . . . . . . 50 4.7 Final Selection of Tree Sensing Concept . . . . . . 54 5.0 APPLE HARVESTER STEERING CONTROL SYSTEM . . . . . . . . 55 5.1 Dynamic Considerations of the Automatic Steering Control System . . . . . . . . . . . . . . 63 6.0 SONAR SENSOR DEVELOPMENT . . . . . . . . . . . . . . . . 66 6.1 Sonar Sensing System . . . . . . . . . . . . . . . 70 6.1.1 Description of Sonar Circuits . . . . . . . 72 6.1.2 Description of Interface Circuit . . . . . . 78 6.2 Sonar System Testing and Results . . . . . . . . . 83 6.2.1 Sonar System Accuracy Tests and Results . . 84 6.2.2 Sonar Beam Angle Tests and Results . . . . . 94 ii CHAPTER 7.0 8.0 9.0 10.0 11.0 12.0 APPENDICES A. B. SIMULATION MODEL WITH INTERACTIVE COMPUTER GRAPHICS 7.1 Model Requirements . . . . . . . . . . . . 7.2 Simulation of Steering Control System 7.2.1 MicrOprocessor Simulation Model . . 7.2.2 Harvester Motion Simulation Model 7.2.3 Model Verification and Validation . 7.3 Simulation Results . . . . . CONTROL SYSTEM SOFTWARE . . . PERFORMANCE TESTS AND RESULTS . . . . 9.1 Configuration of the Steering Control System . 9.2 Test Procedure . . . . . . 9.3 Results of Straight Row Tests 9.4 Results of the Curved Row Tests 9.5 Test Results - Row with Step-Change 9.5.1 Reliability Problems During Test SU MMAR Y O O O O O O O O O I O O 0 CONCLUSIONS . . . . . . . . . . . SUGGESTIONS FOR FURTHER STUDY . . DATA FROM SONAR ACCURACY TESTS . COMPUTER PROGRAM FOR HARVESTER SIMULATION iii MODEL Page 101 102 105 109 111 115 124 131 149 149 152 159 163 173 181 185 196 198 201 212 CHAPTER C. Page ANALYSIS OF THE HARVESTER'S INERTIAL EFFECTS DUE TO GROUND SPEED . . . . . . . . . . . . . . . . . . . 223 COMPUTER PROGRAM FOR THE HARVESTER'S MICROPROCESSOR- BASED STEERING CONTROLLER . . . . . . . . . . . . . . . . 242 DATA FROM PERFORMANCE TESTS OF THE HARVESTER'S STEERING CONTROL SYSTEM . . . . . . . . . . . . . . . . . 256 LIST OF REFERENCES 0 O O O O O O O O O 0 O O O O I O O 0 O O O O O 265 iv Table 6.1 6.2 8.1 9.1 A.1 A.2 A.3 A.4 A.5 C.1 E.1 E.2 E03 E.4 E05 LIST OF TABLES Data from accuracy test for sonar unit done at 25.500. . . . Data from accuracy test for sonar unit one at -1.0°C . . . List of the Microprocessor I/O Ports . . .......... The X-Y Coordinate Position of Tree Stands Curved Row Tests . . . . . . . . . . . . . Sonar Distance Data from AccuracyoTest for Unit-2 at Air Temperature of 25.5 C . . . Sonar Distance Data from AccuracyoTest for Unit-3 at Air Temperature of 25.5 C . . . Sonar Distance Data from AccuracyoTest for Unit-4 at Air Temperature of 25.5 C . . . Sonar Distance Data from AccuracyoTest for UHTt-S at ATP Temperature Of 2505 C o o o Sonar Distance Data from AccuracyoTest for Unit-2 at Air Temperature of -1.0 C . . . for the Sonar Sonar Sonar Sonar Sonar Position Coordinates of CG.from Dynamic Model for Vehicle of Weight 89,000 N and 22,200 N . Harvester's Position Data Collected for Three Performance Tests with a straight Row . . . . . . . . . . . Harvester's Position Data Collected for Three Performance Tests with a Row Containing an 8 cm Step Change Harvester's Position Data Collected for Three Performance Tests with a Curved Row on a Campus Lawn . . . . . . . . . Harvester's Position Data Collected for Three Performance Tests with a Curved Row on a Concrete Driveway . . . . . . Computed Position Coordinates of Front-Point A and Rear Point-B for Three Performance Tests with Straight Row . . . . . . . . . . . . . . . Page 86 90 143 158 202 203 204 259 260 Table Page E.6 Computed Position Coordinates of Front-Point A and Rear-Point B for Three Performance Tests with Curved Row on a Campus Lawn . . . ..... . . . . . . . . . . . . 262 E.7 Computed Position Coordinates of Front-Point A and Rear-Point B for Three Performance Tests with Curved Row on a Concrete Driveway . . . . . . . . . . . . . . . . 263 E.8 Computed Position Coordinates of Front-Point A and Rear-Point B for Three Performance Tests with Row Containing 8 cm Step Change . . . . . . . . . . . . . . . . 264 vi LIST OF FIGURES Figure Page 1.1 Perspective View of the USDA Over-The-Row Apple HarveSter I I O O O O O O O O O O I O O O O O O O O O 0 O O 2 1.2 Top View of the USDA Over-The-Row Apple Harvester . . . . . 4 1.3 Front View of the USDA Over-The-Row Apple Harvester . . . . 5 3.1 Top View of Harvester Showing the Allowable Zone for Tree Trunks During the Harvesting Operation . . . . . . . . 31 3.2 Diagram of Tree Row Showing the Minimum Radius of curvature O O O O O O O O O O O O O O I O O O O O O O O O O 32 3.3 Schematic Diagram of the Apple Harvester's Hydraulic Circuitry for Front Wheel Steering . . . . . . . ..... 34 4.1 Method 1 - Tree Sensing Using Two Sonar Units to Determine Tree Position . . . . . . . . . . . . . . . . . . 39 4.2 Method 2 - Tree Sensing Using One Sonar and One Angle Measurement to Determine Tree Position . . . . ..... . 42 4.3 Method 3 - Tree Sensing Using Two Sonar Units to Detect the Tree Presence in Zones . . . . . . . . . . . . . 44 4.4 Method 4 - Tree Sensing Using Two Optical Sensors and Two Angle Measurements Used to Determine Tree pOSition O O O O O O O O O O O O O O O O O O O O O O O O O 47 4.5 Method 5 - Tree Sensing Usng Two Optical Sensors to Detect Tree Presence in Zones . . . . . . . . . . . . . . . 49 4.6 Method 6 - Tree Sensing Using Multiple Sonar Units to Determine Tree Position (Top View) . . . . . . . . . . . 51 4.7 Method 6 - Tree Sensing Using Multiple Sonar Units to Determine Tree Position (Front View) . . . . . . . . . . 52 5.1 Diagram of Major Components of the Automatic Steering cont "01 syStem O O O O O O O O O O O O O O O O O I O O O O 56 5.2 Diagram of Tree Position Error . . . . . . . . . . . . . . 59 vii Figure 5.3 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 6.10 6.11 6.12 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 Block Diagram of the Apple Harvester Automatic Steering cont r0] system 0 O O O O O O I O I O O O O O O O O O O Polaroid Ultrasonic Circuit Board . . . . . . . . . . . Polaroid Ultrasonic Transducer . . . . . . . . . . . . Timing Diagram of Ultrasonic Circuit Board . . . . . . Diagram of Sonar Sensing System . . . . . . . . . . . . Sonar Circuit Diagram for One Sonar Unit . . . . . . . Timing Diagram of Sonar Circuit for Objects for Range Less Than 240 cm 0 O O O O O O O O O O O O O O O O O 0 Timing Diagram of Sonar Circuit for Objects With Range Greater Than 240 cm . . . . . . . . . . . . . . . . . . Interface Circuit Program . . . . . . . . . . . . . . . Timing Diagram for Interface Circuit . . . . . . . . . Distance Data for Sonar Number One at 25.50C . . . . . Distance Data for Sonar Number One at -1.0°C . . . . . Plot of Sonar Data to Determine Transducer Beam Angle . Mode 1 Graphical Output From The Simulation With K = 1.0 O O O O O O O O O O O O O O O I O O 0 O 0 Mode 2 Graphical Output From the Simulation With K = 2.0 I O O O O O O O O O O O O O O O O O O O 0 Flow Chart of the Harvester Steering Control System Simulation . . . . . . . . . . . . . . . . . . . Flow Chart of the Simulation Model for the Harvester's Microprocessor Functions . . . . . . . . . . . . . . . Diagram of the Harvester Steering Geometry . . . . . . Diagram of the Harvester for Data Collection Test . . . Sonar Data and Wheel Position from System Test at 0.8 km/h Using a Continuous Wooden Fence . . . . . . . Flow Chart of Harvester Motion Simulation . . . . . . . viii Page 64 67 68 69 71 73 75 77 79 82 87 89 96 104 106 107 110 114 116 118 123 Figure 7.9 Graphical Output of Simulation - Mode 2 with K = 0.5 O I O O O O O O O O O O O O O O I O O O O O O 7.10 7.11 7.12 8.1 8.2 8.3 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 9.9 9.10 9.11 9.12 Graphical Output of Simulation - Mode 1 with K = 0.5 O O I O O O O O O O O O O O O O O O O 0 O O O Graphical Output of Simulation for Curved Row MOde 2 With K = 0.5 O O O O O O O O O O O O O O O O O Graphical Output of Simulation for Curved Row MOde 1 With K = 0.5 O O O O O O O O O O O O O O O O 0 Block Diagram of Steering Control System . . . . . . Flow Chart of the Computer Subroutine Used to Turn the Harvester Front Wheels . . . . . . . . . . . Flow Chart of the Steering Control Program for the 1302 Microprocessor . . . . . . . . . . . . . Diagram of Simulated Tree Stand Used in Steering Control System Tests . . Diagram of Linkage that Supports the Front Pen . . . Plan View of Harvester Showing Pen Locations for Performance Tests . . . . . . . . . . . . . . . . Plan View of Test Configuration Used for Straight Row Test . . . . . . . . . . . . . . . . . . Path of the Harvester for Straight-Row Test Number 1 Path Path Path on a Path on a Path on a Path on a Path on a of the Harvester for of the Harvester for of the Harvester for Concrete Driveway . of the Harvester for Concrete Driveway . of the Harvester for Concrete Driveway . of the Harvester for Campus Lawn . . . . of the Harvester for Campus Lawn . . . . Straight-Row Test Number 2 Straight-Row Test Number 3 Curved Row Test Number 1 Curved Row Test Number 2 Curved Row Test Number 3 Curved Row Test Number 1 Curved Row Test Number 2 ix Page 126 127 129 130 132 134 137 150 153 154 156 160 161 162 164 165 166 168 169 Figure 9.13 9.14 9.15 9.16 9.17 9.18 9.19 9.20 A.1 A.2 A.3 A.4 A.5 8.1 C.1 C.2 Path of the Harvester for Curved Row Test Number 3 on a Campus Lawn Path of the Harvester Predicted by the Simulation Model for Curved Row . . . . . . . . . . . . . . Path of the Harvester for Row with Step-Change-- Test Number 3 O O O O O O O O O O O O O O O O O O O O 0 Path of the Harvester Predicted by the Simulation Model for Row with Step-Change with Time Constant Equal to 2.0 seconds . . . . . . . . . . . . . . . . . . Path of the Harvester Predicted by the Simulation Model for Row with Step-Change and Model Modified to Turn the Wheel 0.4 for Each 2 cm of Tree Position Error Path of the Harvester Predicted by the Simulation Model foroCurved Row with Model Modified to Turn the Wheels 0.4 for Each 2 cm of Tree Position Error . . . . Path of the Harvester for Row with Test Number 1 O O O I O O O O I 0 Path of the Harvester for Row with TeSt Number 2 O O O O O O O O O 0 Distance Data from 25.5 c O O O O 0 Distance Data from 25.5 C O O O O 0 Distance Data from 25.5 C . . . . . Distance Data from 25.5 C O O O O 0 Distance Data from Sonar Sonar Sonar Sonar Sonar Unit-2 at Unit-3 at Unit-4 at Unit-5 at Unit-2 at Step-Change-- Step-Change-- Air Temperature Air Temperature Air Temperature Air Temperature Air Temperature of of of of of -100 c O O O O O O O O O O O O O O O O O O O O O O 0 Listing of the Program for the Harvester's Steering Control System Simulation Model Diagram of Steering Angle and Block Diagram of Vehicle Made] 0 O I C C O O O C O O O O O O O O O O O O O O 0 Diagram of Tire Side Slip Angle Page 170 172 174 177 179 180 182 183 207 208 209 210 211 Figure C.3 C.4 C.5 C.6 C.7 C.8 C.9 0.1 Typical Tire Characteristic of Side Force with Respect to Side Slip An91e O O O O O O O O O O O O O O O O O O 0 Dimensions for Vehicle . . . . . . . . . . . . . . . . . . Body Centered Axes Model for Vehicle Equations of Motion 0 O O O O O O O O O O O I O O O O O O O O O O O 0 Basic Ackermann Steering Geometry . . . . . . . . . . . Free-Body Diagram of Forces for Steady State Turn with Allowable Tire Friction Force Equal to 2,200 N (500 lbf) . . . . . . . . . . . . . . . . . . . . Plot of Steering Angle and Yaw Rate . . . . . . . . . . . Computer Program used to Compute the Vehicle Response Due to a Specified Steering Angle Input . . . . . . . . Assembly Language and Object Code Listing of the Program Used by the Steering Controller's Microprocessor . . . . xi 233 236 238 243 1.0 INTRODUCTION For several years, the United States Department of Agrhuflture (USDA) has been developing an over-the-row apple harvester. This harvester which is being developed by Tennes gt_al; (1976) is an experi- 1nental prototype harvester. They also are doing research and develop- Inent work to design spraying and pruning equipment. By installing the appropriate equipment on to the harvester's main frame, this machine can be used for harvesting, spraying and pruning of apple trees. Figure 1.1 is an illustration of the USDA apple harvester. 'The harvester is equipped with a hydrostatic drive system used to drive all four wheels and hydraulic cylinders are used to steer the front and rear wheels. The hydraulic system also supplies power for other auxillary equipment. Tennes and Brown (1981) have reported on the development of the shaker bar system and the basic dimensions of this system are shown in Figures 1.2 and 1.3. For the harvesting operation, the harvester is being designed to Operate in a high density orchard which contains semi-dwarf trees that are spaced in either a straight or curved row at a minimum distance of 305 :5cm (120 :2 in) between the tree trunk centerlines along the row. These rows are to be spaced apart at a nominal distance of 487 cm (192 inches). During the harvesting operation, the harvester travels at a constant speed of 0.8 km/h (0.5 mph). Figure 1.1 Perspective View of the USDA Over-The-Row Apple Harvester 1.1 The Need for Automatic Steering Systems In order to harvest apples effectively, the USDA apple harvester should drive over the tree row so that the tree trunk centerline follows along the harvester's longitudinal centerline within an allowable toler- ance. Tennes and Brown (1981) have reported that the harvester's shaker system performance may be improved if the harvester is centered over the tree tuna when the shaker bars are operating. During preliminary tests, they found that the harvester should be centered over the tree such that the harvester's centerline is within :22 cm (9 in) from the tree trunk centerline. Accurate steering may also prevent damage to the trees and the harvester (hue to inadvertent collisions with an apple tree. But, manual steering to keep the harvester centered on a tree is a difficult task. To accurately steer the harvester, it is necessary that the operator see the tree trunk,tnn;CMe to the operator's position on top of the harvester, the field of view to the tree trunk is blocked by tree foliage. Therefore, since manual steering is difficult and since accurate centering of the harvester over the tree may improve the effectiveness of the harvester shaker system, there is a need for an automatic steering control system which will perform the task of auto- matically steering the harvester over the tree row. An automatic steering control system may be able to solve other problems associated with steering agricultural vehicles. Busse Eli]; (1970) have reported that for a corn combine, operators appear to be SHAKER BAR I DIRECTION ASSEMBLY \x OF TRAVEL 1.26 m (4.1 ft) \\ \ FLYWHEELS _____...'/ Z/{Z/a /, R A\ 7s \/ 7% / R -. I: / \ é. E 212642) .c. . (. DRRIB/ENG :E: ("-1 -3; g - -.\ DRIVING LECL H 4.09 m (13.4 ft) E Figure 1.2. Top View of the USDA Over-The-Row Apple Harvester. ' ' ‘0 — k '1 \ DRIVE MOTORS ] J 1| §_ SYNCHRONIZING CHAIN _4 PENDULUM 381 cm (150 in) ,. PENDULUMS AND CRANK HOUSINGS _ 0 0 o o O 0 o O c O o O SHAKER BARS 4 256.5 cm t (101.0 in) ‘/ ‘thfi §/ I, 365.8 cm 7 n440inI I Figure 1.3. Front View of the USDA Over-The-Row Apple Harvester. physically limited to operating at speeds of 5 to 6 km/h for long periods, yet the automatic steering system that was develOped by Busse (1970) Operates effectively at speeds of 8 to 9 km/h. Busse also report- ed that when an automatic steering system was used for a combine, the operator could put his full attention on regulating the forward speed as required by the crop conditions and thus the operator was able to in- crease the output of the harvester up to 15 percent. In addition to a limit on the forward speed with manual steering, the accuracy of manual steering may be affected by the complexity of the steering task and by operator fatigue. Kirk g£_al; (1975) reported that long hours of steer- ing a self-propelled swather is very taxing and monotonous. The Opera- tor must continuously make steering corrections to accurately follow the edge of the standing grain. Also, due to the sensitive steering system of a swather, Kirk et_al; have reported that even an experienced opera- tor can have difficulty accurately steering tOIWHfiflfize the amount of overlap onto the previously cut swath. Therefore, it may be possible to use an automatic steering system on an agrhnntural wflficle to: (1) help reduce the fatigue due to long hours of steering, (2) increase the forward speed, (3) allow the Operator time to control other machhue functions which could improve the machine output, and (4) help reduce the overlap that sometimes results with manual steering. 1.2 Comparison of Sensing Methods The development of an automatic steering control system can be divided into two major components, a controller and sensing system. The function of the controller is to: (1) receive signals from the sensing system; (2) process these signals and; (3) transmit signals that will control a machine or system. Controllers may be divided into two types; (1) analog electronic controller which processes analog signals, and; (2) a digital electronic controller which processes digital signals and which may use a microprocessor. In order for a controller to perform well, a sensing system must be available to send a signal containing appropriate information to the controller. Sensing systems may also be divided into two categories; contact and non-contact sensing systems. Since the development of a control system requires the selection of a sensing system, these two types of sensing systems will be compared. 1.2.1 Contact Sensing Systems There are generalized problems associated with use of a contact type sensing system and these problems result because the contact type sensor must make physical contact with the object that is being sensed. A typical example of a contact sensor is a mechanical feeler arm assembly that is used to sense the position of an object. The mechani- cal feeler arm assembly is usually held in a undeflected position during its operation. As the feeler arm moves to a position where it contacts the object that is being sensed, the object applies a force to the arm and this causes the arm to be deflected to a new position. The feeler arm can be connected to an electrical transducer which converts the deflection of the arm to a voltage signal that represents the position of the arm. The position of the arm also represents the position of the object that is being sensed. Because the contact type sensor contacts the sensed objects, several problems can occur. One problem with contact sensing is damage of the sensing assembly. During the sensing operation, when the vehicle is traveling at high ground speeds, the sensing assembly can strike or impact the sensed object with enough force to damage the sensing assembly. Also, this impact with the sensed object can cause the sensing assembly to rebound or bounce off the sensed object and this rebounding can result in erro- neous signals that are sent to a controller. Also, the sensed object can become damaged by the impact with the sensing assembly. Another problem is that a contact assembly can become fouled with debris. This usually results because the sensing assembly is placed in an area where it can sense an object and this is typically at a location where the sensing assembly is unprotected from fouling by debris. There is also a problem of sensing system failures that are due to excessive wear of moving parts. Contact type sensors typically have several moving parts such as bearings or gears and, thus, a sensing system can fail periodically due to the excessive wear of these moving parts. Busse _e_i_:__ 2.1.; (1970) also have explained that there is a problem with sensing fragile objects. Busse gt__a_l_._ (1970) have reported that a contact type sensor has not been developed to effectively sense the position of stems of small grain plants because these stems are too weak to activate a contact type sensor . 1.2.2 Non-Contact Sensing Systems To begin the discussion of how a non—contact sensing system may be advantageous when compared to a contact sensing system, we must first consider the definition and understand the general operation of a non- contact sensing system. Non-contact sensing is usually accomplished by transmitting a signal such as light, radio waves, or ultrasonic sound pulses toward the object that is being sensed. The signal travels to the object and reflects from the object back to the sensing system. Generally, the reflected signal allows the sensor to detect the presence of an object in a specific sensing zone, although some sensing systems can sense more information about the object such as, color, shape, or range of the object from the sensor. The use of light, radio waves and sonar are only some of the ways that non-contact sensing can be accom- plished. This subject of non-contact sensing (or sometimes called remote sensing) is a broad area that covers many different sensing systems. In order to give a description of how a non-contact sensing system operates, two examples of non-contact sensing systems will be briefy explained. Two examples of non-contact sensing systems are radar and sonar (airborne). With a radar system, an object's distance from the radar unit can be measured. The radar transmits a pulse of electromagnetic waves (radio waves) toward an object and for some objects the electro- magnetic waves are reflected back to the radar unit. The time is measured for the pulse of waves to travel out to the object and back to the radar unit. Then the distance to the object is computed using the measured time interval. In this computation, the total distance traveled by the pulse of waves is equal to the measured time interval multiplied by the velocity of the pulse. The pulse velocity is the speed of light. The airborne sonar system works in a similar manner as the radar system, altough the sonar system transmits a pulse of ultrasonic sound waves. This pulse of waves is also reflected from the surface of an object and returned to the sonar system. Then, like the radar system, the object's distance is determined by computation using 10 the time of travel of the pulse and the velocity of pulse. The pulse travels at the speed of sound. when comparing a non-contact sensing system to contact sensing, the non-contact sensing system appears to e able to solve some of the problems associated with the contact systems. First, non-contact sensors are typically not damaged by the object that is sensed because the object does not contact the sensor. Also, since the object does not strike or impact the non-contact sensor, there is not a problem with the sensor's rebounding or bouncing from the surface of the object as it sometimes occurs with cntact sensors. Second, a non-contact sensing system can sometimes be positioned on a machine in a remote location where it can be protected and this protection usually prevents the non- contact sensing system from being fouled by debris. Third, a non- contact sensing system can generally be built with few moving parts, thus, there is a lower probability of failure due to the excessive wear of the moving parts. And, fourth, a non-contact system can be used to sense weak or fragile objects which is a problem for contact type sensors, because they must apply a force to the object. 'Therefore, it appears that a non-contact system may be able to solve the problems associated with contact sensing systems. Since the non-contact sensing system has the potential of providing good performance without the problems of a contact senshu;systan,it is worthwhile to develop a non-contact sensing system that can be used on an automatic steering control system for an agricultural vehicle. The develOpment of a non-contact sensing system and digital electronic con- troller was selected as the subject of this research. 11 1.3 Objective The objectives are: I. Design and test a non-contact sensing system which can measure the distance from the sensing system to a tree trunk. 2. Design a micrOprocessor-based automatic steering control system for the USDA apple harvester using a non-contact sensing system. 3. Design the steering control system to control the steering of the harvester's front wheels. 4. Design the steering control system to steer the harvester so that each tree passing through the inside space of the harvester stays within an allowable zone. The allowable zone is 45 cm wide, 409 cm long and is centered on the harvester centerline. This zone extends from the harvester's front wheels to the rear wheels (Figure 3.1). 5. Design the steering control system to satisfy the performance objective of item 4 above with the harvester traveling at a forward velocity of 0.8 km/h (0.5 mph) and with either a straight or curved tree row. A curved tree row shall have trees positioned on a smooth continuous curve with minimwn radius of curvature of 121.9 m (400 ft). 6. Design the control system to interface with the existing USDA apple harvester steering system and to have a total cost of less than $2,000 for the electronic equipment used in the steering control system. 12 7. Test the steering control system using simulated conditions. These tests shall determine if the steering control system satisfies the control system performance objectives which are listed above. These objectives were used to develop a set of design requirements which are shown in Chapter 3. 2.0 LITERATURE REVIEW A review has been done of reseach work on automatic steering sys- tems that are used in agriculture. The steering systems that were reviewed can be divided into four categories and the categories are automatic steering systems using: (1) contact type sensors, (2) optical type sensors, (3) buried cable technique and (4) special position sensing. Research work on the development of automatic steering systems has been published for many years. Grovum and Zoerb (1970) in their work to develOp an automatic steering system have reviewed many different types of automatic guidance systems that were developed outside the field of agriculture. They concluded that sensing systems such as laser and radar have been used successfully in military appli- cations for guidance control but are too expensive to be used on an agricultural field machine. Also, several references were reviewed on the subject of sonar sensing systems. 2.1 Automatic Steering Control Systems Using Mechanical Contact Type Sensors Grovum and Zoerb (1970) have designed an automatic steering control system to steer a tractor during the plowing operation. A sensing system and an electronic analog controller were designed such that the tractor follows a furrow that has been previously plowed. This sensing system was a displacement sensor which consists of two mechanical feeler arms that contact both sides of the furrow in order to sense the position of the furrow. When the tractor is not accurately following 13 14 the furrow, one of these feeler arms will move away from a neutral position. The feeler arm is connected to variable resistors and when the feeler arm moves to a new position the variable resistor circuit produces an output voltage signal which is proportional to the amount of displacement of the feeler arms. This steering control system by Grovum and Zoerb (1970), also included a directional gyroscope. The gyroscope was used in the control system for two purposes. First, the signal from the gyroscope was used to keep the tractor heading in a straight line on the first pass of the tractor through the field because on the first pass there was no furrow for the tractor to follow. Second, the gyroscope was used as a feedback signal in the control system so that this signal would attenate an error signal from the displacement sensor. This results in a damping effect and the tractor has a slower response to the error signal from the displacement sensor. The voltage signal from the gyroscope is proportional to the change in azimuth of the tractor's direction of travel. Grovum and Zoerb (1970) also utilized an analog electronic controller in this steering control system. The controller processes the analog voltage signals from the directional gyroscope and the displacement sensor, and these signals are feedback signals in the closed loop automatic control system. Grovum's simulation model of the control system indicated that the system behaved like a linear second order system. The test results of the control system showed that the system was unstable with the displacement sensor mounted next to the rear of the tractor. With the displacement sensor mounted near the front wheels the control system provided effective control for speeds up 15 to 6.8 km/h (4.2 mph). The tests also showed that at speeds less than 6.8 km/h the signal from the direction gyroscope was not neded, but at speeds greater than 6.8 km/h the gyroscope signal was required to achieve stable control of the tractor. Shukla et al. (1970) have reported on the analysis of a vehicle steering control system. In this analysis a model was developed that simulated the kinematic motion of a vehicle with front and rear wheel steering. The model was developed for an automatic steering control system that used an ideal prOportional control algorithm. In this control algorithm, the controller sends out a signal to turn the wheels to an angular position corresponding to 1.0 radian per 30.4 cm (12.0 in) of error detected by a path detection sensor. This path detection sensor was tested at various locations in order to find the best sensor position. They also assumed that the wheels turn (for steering) at a high enough rate that negligible time is required for the wheels to arrive at the desired wheel position. The wheel turning rate is defined as the angular velocity of the wheel where the wheel rotates about its vertical centerline. The model was used to investigateeffects on the tractor tracking error that was caused by changing the major dimensional parameters that define the vehicle steering system. To verify the model, a sinuflified tractor'steering control system was used to determine if a good correlation existed between the path traveled by the experimental test vehicle and the path predicted by the computer model. The agreement between the predicted and measured results were so closely correlated that they concluded there were no gross errors in the simula- tion model. Some of the major conclusions from their study are: (1) the path detection sensor's optimum position is approximately 30.4 cm 16 (12.0 in) in front of the front wheels; (2) to achieve stable operation at least one path detection sensor must be near or in front of the front wheels, and (3) effective steering control can be achieved by using the error signal from a path detection sensor in front of the front wheels to control the steering of both the front and rear wheels. These con- clusions were based on results of the computer simulation with the vehicle velocity set at 8.0 km/h (5.0 mph). Another steering control system, reported by Busse et__a_]_._ (1970), is the development of automatic steering control for a combine and forage harvester, which was done by the Claas Company. This machine when used to harvest corn, uses a pair of feeler arms to sense the position of corn stalks. When the harvester is not centered on the row, the corn stalks will contact the feeler arms and move the feeler arms from a normal positon to a displaced position. The feeler arm is connected to a variable resistor circuit which converts the position of the feeler arm to a voltage signal from the feeler arm circuit is sent to an electronic analog controller. When steering corrections are needed, the controller sends a voltage signal to a hydraulic valve and this causes the wheels to turn. An analog voltage signal is also received by the controller to determine the angular position of the wheels. Busse et al. have tested the steering control system and have reported that the harvester deviated from the row centerline 2.5 cm (1.0 in) at speeds less than or equal to 6.0 km/h (3.7 mph) and at speeds up to 9.7 km/h (6.0 mph) the deviation did not exceed 5.1 cm (2.0 in). Upchurch _e_t_:__a_l_._ (1980) have reported on the design and development of an automatic steering control system for the USDA apple harvester. 17 The control system uses a sensor arm system which contacts the tree trwnfl< to detect tree position. At the front of the apple harvester one sensor arm system is mounted on the left side of the apple harvester and one on the right side. The sensor arms are located near the harvester centerline so that if the harvester is not centered over the tree, then one of the sensor arms is pushed by the tree trunk. The sensor arm system is designed as a 4-bar linkage and is spring loaded to hold the arm in a neutral position. As the tree pushes on one of the sensor arms the sensor arm moves to a new position. A shaft encoder is mechanically connected to the sensor arm and the shaft on the shaft encoder turns as the sensor arm moves. The shaft encoder produces a 4 bit binary output number which is proportional to sensor arm position, and this binary number also represents tree position. The binary code from the shaft encoder is sent to the microprocessor based digital electronic controller and the controller sends signals to electrical relays which control the Operation of solenoid operated hydraulic valves. These valves are used to steer the front and rear wheels. Shaft encoders are also used to determine the angular position or steering angle of the front and rear wheels. Upchurch et al. have tested the steering control system and they reported that the control system kept the harvester centered on a curved path which had a minimum radius of curvature of 22 m (75 ft) within a tolerance of _+_7.6 cm (3.0 in) with a forward speed of 0.8 km/h (0.5 mph). These automatic steering control systems that have been reviewed are significantly different from the objectives of this research project for the development of an apple harvester steering control system. Most 18 of the control systems that were reviewed had a position-error signal that was continuous, whereas the apple harvester may only have a noncontinuous error signal because the trees are 304 cm (120 in) apart in the row. The review of these control systems found that all but one system used an electronic analog controller whereas the apple harvester is being develOped for a microprocessor based controller. The reviewed control system used contact sensors and the apple harvester is being developed for the non-contact sensors. A control system that was designed by Upchurch et al. (1980) was an effective steering control system for following a row of apple trees using contact sensors. But, the steering control system for this research should use a non-contact sensing system and will require the development of: (1) a non-contact sensing system; (2) computer control algorithm; (3) computer software, and (4) hardware to interface the sensing system with a controller. 2.2 Automatic Steering Control Using Optical Type Sensor Kirk and Krause (1970) have designed a steering control system for a swather which uses an infrared proximity sensor to measure the distance from the sensor to the stems of small grain plants. This sensor measures the intensity of the reflected light and then converts it to a voltage signal. The voltage signal is processed by a linearizing circuit to produce an output voltage that is linearly proportional to the distance measured. The range of the sensor is 10 to 56 cm (4 to 22 in). An analog electronic controller receives the voltage signal from the sensing system and the controller sends signals to hydraulic valves to steer the wheels. The voltage signal is used by the controller to determine the error in position of the swather. The 19 wheels are then turned to a steering angle that is proportional to the position error. Kirk and Krause have reported that on a single pass, the swather will follow the row within :_12.2 cm (4.8 in) at speeds in the range Of 3.2 to 12.1 km/h (2 to 7.5 mph). For the case where the swather made more than four passes, the edge of the crOp which was cut by the swather developed an oscillating pattern with 30 cm (12 in) amplitude with wave length of 12.2 m (40.0 ft). For a larger number of passes, amplitudes greater than 30 cm were observed. Ambler et;§fl;_(1980) have developed a prototype tractor which does not require a driver. This tractor uses optical sensing to follow a plowed furrow and an electronic distance measuring device to measure the distance to reference targets for controlling the turn that is required at the end of each furrow. This electronic distance measuring device can be described as an Optical ranging system. The range measurement from the ranging system is used by a microprocessor based controller for locating the end of the furrow and controlling the turn-around of the tractor at the end of the furrow. This control system also requires a measurement of the bearing which is the direction of travel of the tractoru. 'This bearing measurement is made by using the ranging system and a precision potentiometer. The ranging system rotates at a rate of one revolution per second and the potentiometer setting changes as the ranging system rotates. When the ranging system is pointing toward the pole or target at the end of the furrow a voltage measurement is taken from the potentiometer and this voltage signal is translated into an angle. This angle is the direction of travel of the tractor relative to the target. The targets are posts which are wrapped with reflective tape 20 and these posts are located near the end of furrows with a spacing between posts of 15 m. Note, that the steering control system for the USDA apple harvester is not being designed for an automatic turn around at the end of a tree row. This Optical ranging system that was developed by Ambel e_t__a_l_._ measures the distance to the posts by processing a modulated infrared light signal that is transmitted to the post and reflected back to the ranging system. The infrared light beam is produced by energizing and de-energizing an infrared light emitting diode at a frequency of SMHz. In order to make a distance measurement, the light beam is directed toward the target and the beam is reflected back to the ranging system. The ranging system then receives the reflected modulated light signal and measures the phase shift between: (1) the transmitted beam, and (2) the received (or reflected) beam. This measured phase shift is converted to the target's range. This distance measured by the ranging system is usually called electronic distance measurement. Smith (1980) reported on a similar ranging system developed by Hewelett Packard Cor- poration which is called an industrial distance meter (Model 80850 A).1 This system is accurate to within :8 mm plus 1 mm per kilometer of range and the reading rate is 9 measurements per second. This steering control system that was developed by Ambler et al. (1980) used another optical sensor to determine if the tractor was lTrade names are used in this paper solely to provide specific information. Mention of a product name does not constitute an endorse- ment of the product by the author to the exclusion of other products not mentioned. 21 accurately following the edge of the furrow. This system uses a light source that is directed downward at a furrow wall. The light reflected from the furrow goes through a lens that directs the light onto two arrays of photo cells. If each of the two arrays receive light covering equal area on the photocell arrays then the tractor sensor is centered on the furrow edge. Each photocell array produces an output voltage which is prOportional to the number of photocells in the array which receive the reflected light. The voltage signal from each array is summed to produce a single voltage signal which is linearly proportional to the error of the tractor position. The polarity of the voltage corresponds to the direction of the error hiinactor position. 'This error signal is sent to an analog controller circuit to execute the required steering of the tractor's front wheels. .After reviewing these references on steering control systems, it appears that these systems are not directly applicable to the apple harvester steering control system for the following reasons. All of these control systems used sensors to measure a continuous signal for vehicle position-error and these signals were processed by an analog controller which controlled the vehicle steering. However, the apple harvester steering control system will probably have to use a noncon- tinuous position-error signal due to the space between trees. Also the design of the apple harvester steering control system has the objective to utilize a micrOprocessor based controller. The electronic ranging system that was described, may be effective at measuring the range to apple trees, but this systehIis too expensive for the apple harvester control system. The optical sensor that was designed by Ambler gt_al; to follow the furrow is specially designed to sense the edge of a furrow 22 and it would not function as a sensor to detect apple trees. This concept of using an Optical sensor appears feasible to sense the position of an apple tree trunk, but a new lens system and electronic cirwnIit woul<1 be required. The infrared optical sensor that was developed by Kirk gt_al; may be applicable for apple tree trunk sensing for the apple harvester's non-contact ranging system, but there is the disadvantage that the sensor functions only within a small range. The sensor's maximum range is 56 cm (22 in). 2.3 Automatic Steering Control System Using Buried Cable The buried cable technique is a concept that has been investigated for many years. Schafer and Young (1980) have develOped a digital electronic controller for a tractor which utilizes the buried cable technique. The tractor's control system uses a pair of antennas which develop a signal that indicates if the tractor is either to the right or to the left of the cable. The steering control system uses this signal so that the tractor follows the buried cable. The buried cable transmits an electromagnetic wave due to low frequency (3 kHz) alternating current flowing in the cable. The antenna circuit develops a logic: signal that corresponds to the tractor's position. This system has only two logic states which are: state (1), the tractor is Off the course to the left of the buried cable, or state (2), Off the course to the right of the buried cable. The controller for the steering control system uses a 8085 micrOprocessor which sends out an eight bit binary code number to control the direction of wheel turning (for steering) and the rate of wheel turning. Six bits of this number are used to control the rate of wheel turning. This is accomplished by sending the six bit 23 numbers to a digital-tO-analog converter and then sending the analog voltage signal to a proportional control valve in the hydraulic circuit. ‘The computer algorithm that is used to control the steering, increments the binary number for the turning rate at discrete time intervals during the time period when the error logic signal remains unchanged. At the time when the error logic signal changes to the other state, the binary code for the rate of turn is reset to zero and the binary code for the directicni of turvI is changed to the other direction. Then the control process repeats. Preliminary test results of the tractor's steering control system indicated that the tractor's deviation from the cable is within _+_7.0 cm at speeds up to 20 km/h (12 mph) for a straight line course without towing an implement. The buried cable method is too expensive for use in an apple orchard and therefore this technique was not selected for development with the apple harvester's automatic steering system. 2.4 Automatic Steering Control Using Spacial Position Sensing Smith et al. (1979) have investigated the feasibility of con- trolling the steering of a tractor using a technique called non- contact spacial position sensing. This technique requires that the coordinate position of two points on the tractor be measured relative to a fixed reference point. One method to accomplish this is to use two radar units which are at two fixed locations in a large field. The radar units could then measure the distance to two points on the tractor, and by using triangulation, the X and Y coordinate position of the two points on the tractor could be computed. The X-Y coordinate system is fixed to the field. 24 The work by Smith et al. investigated the feasibility of a| I-«i——44.7cm 12° 12° LEFT SONAR "t~ggfln' AR SENSING ZONE SENSING ZONE SONAR UNIT X /—— SONAR UNIT j J W \\\\\\\ // // \X\ I; <<< :1 J (Q 5 s ALLOWABLE TREE TRUNK ZONE Figure 4.3 Method 3 - Tree Sensing Using Two Sonar Units to Detect the Tree Presence in Zones 45 centerline because the tree would not be in any of the sensing zones and if the tree is not detected in a sensing zone then a steering correction will not occur. 'This method of tree sensing is infeasible for the following reasons. First, the sonar units must be located close to the harvester centerline where there is the potential that a tree will collide with one of the sonar units and damage the sonar system. A movable frame could be designed so that the sonar sensor would move out of the way of each tree but, a mechanism Of this type could be complicated and expensive. Second, when the tree row is curved, there is the potential for the sonar unit to detect another tree in the curved row instead of the tree which is directly in front of the harvester. This could occur because the sonar sensing zone is 10.6 m long and thus if a wrong tree is detected then the steering controller would execute an erroneous steering correction. 4.4 Method 4 - Two Optical Sensors and Two Angle Measurements Used to Determine Tree Position by Triangulation This tree sensing concept uses two optical sensors and two angle measurements to determine the tree position. The optical sensors are used to locate the tree and a shaft encoder is used to measure the bearing of azimuth of the sonar beam. This azimuth can also represent the angular position coordinate of the tree. IMNNng the operation of this sensing system the optical sensor sweeps from side to side to locate the tree which is ahead of the harvester. As the optical sensor is sweeping it sends out a modulated light beam. The tree has a retro- reflective target mounted on the tree trunk and when the optical sensor 46 is pointing toward the tree, the modulated light reflects back from the tree. At the moment when the optical sensor receives the reflected light beam an output signal goes to the TRUE state. The optical sensor can be mechanically connected to a shaft encoder and when the output signal from the optical sensor goes to TRUE, the angular position of the sonar unit is measured by the shaft encoder. This angular position represents the angular position coordinate of the tree. Figure 4.4 is a diagram that shows the harvester, the two optical sensors, and the two angle measurements that are needed to define the position of the tree. These two angle measurements define a triangle, and thus triangulation can be used to compute the tree lateral offset from the harvester centerline. An optical sensor was purchased and preliminary tests indicated the beam angle of this sensor was 4.0 degrees when a rectangular target was used.3 This target has the dimensions of 1.8 cm (0.7 in) by 15.2 cm (60 in). This sensor operated effectively when the target was in the range of 0.0 m to 9.1 m. The tree position can be computed using the distance between the two sonar units and the two angle measurements. To check the accuracy of this tree sensing method, the tree lateraloffset was computed for a tree that is offset 50 cm from the harvester centerline and the tree is ahead of the harvester within a range of 100 to 450 cm. The computed lateral tree offset were in the range of 57 to 59 cm and thus the error between the actual tree lateral offset and computed Offset is in the 3Photoelectric transducer was purchased from Banner Engineering Corp., 9714 10th Ave. NO., Minneapolis, MN 55441; Model MULTIBEAM with internal components; scanner block SBLX; logic block LM 3; power block PST and reflector No.2 BRT-L. 47 METHOD 4 TREE TRUNK Q RETROREFLECTIVE ‘77‘~TARGET ANGLE MEASURED TREE TRUNK BY SHAFT ENCODER AS THE SENSOR +— , , RETROREFLECTIVE ””7“ H \‘\\ TARGET \ OPTICAL SENSOR No.1 INFRARED TRANSMITTER/RECEIVER ‘ OPTICAL SENSOR No.2 \\ ll/ \\« / * INFRARED _____ _Q__ _ - _ TRANSMITTER / RECEIVER I"? \SI 7 \'| IQ) “(2, $8.: DIET-E ' :11 I7/ R‘II l7??? /.§Ii I%\\\11 Figure 4.4. Method 4 - Tree Sensing Using Two Optical Sensors and Two Angle Measurements Used to Determine Tree Position 48 range of 7 to 9 cm. This error is acceptable, although there is a high degree of uncertainty in the assumption that the tree angular position can be measured within i} degree. For this sensory system to operate, a reflector must be attached to each tree, thus there would be a large cost to purchase and attach these reflectors to the trees, and there would be an additional cost to replace those reflectors which are damaged by vandalism or inadvertent collisions with orchard machinery. Therefore a more detailed economic analysis must be done in order to complete the evaluation of the feasibility of this sensing method. 4.5 Method 5 -- Two Optical Sensors used to Detect Tree Presence in Zones This sensing technique used two optical sensors to determine if a tree is present in one of the two sensing zones. Figure 4.5 is a diagram whicNI'Tllustrates the harvester with two optical sensors that are used to detect trees in one of two zones. These zones are on the left and right sides Of the harvester's centerline. The zOnes extend into the area ahead of the harvester and are located close to the harvester centerline so that this sensing system can detect if a tree is Offset an excessive distance from the harvester's centerline. For this tree sensing concept, if one of the optical sensors detects a tree in the sensing zone, then a steering correction is required so that the harvester will be aligned with the tree as it drives over the tree. Note, that this sensing system cannot detect a tree which is too far to the left or right of the harvester centerline because the tree would not be in any of the sensing zones and if the tree is not detected in a sensing zone then a steering correction will not occur. 49 METHOD 5 —b( (<— 45.7 cm 40 TREE TRUNK RETROREFLECTIVE 9J4". TARGET MAx. DETECTION SESSITNG RANGE ZONE LEFT SENSING 9- ZONE LEFT RIGHT OPT'CAL OPTICAL * SENSOR :: a / SENSOR I '\ x» . \ \ ‘ \ ‘. \ / . I ; / I \ _ .. u I I l I I-v—-v< m h— --... /\ ”7:3“? /<§§l IJZEQSH Figure 4.5. Method 5 - Tree Sensing Using Two Optical Sensors to Detect Tree Presence in Zones 50 This method of tree sensing is infeasible for the following reasons. First, these optical sensors are located 22 cm from the harvester's centerline which is a location where a tree may collide with the sensor and damage the sensor. Second, there is a potential that if the tree row is curved, then the optical sensor may detect another tree instead of the tree which is directly in front of the harvester. This could occur because the Optical sensor which was described in the previous section has a sensing zone with a range of 9.1 m and if the tree row curves into one of the sensing zones, then a wrong tree may be detected. This would cause the steering controller to execute an erroneous steering correction. 4.6 Method 6 — Multiple Sonars Used to Measure Tree Position Method 6 uses several sonar units to measure the position of the tree. Illustrated in Figures 4.6 and 4.7 is a tree sensing system which uses several sonar units that are mounted in an array in front of the left front wheel and these sonar units are pointed toward the harvester's centerline. These sonar units are positiooned such that the centerlines through the sensing zones of each sonar unit are perpendicular to the harvester centerline. As the harvester drives over the tree row, each sonar will move past each tree in the row. When the harvester moves forward to a position such that at a tree trunk is in one of the sonar unit sensing zones, then that sonar unit measures the distance to the tree trunk. This measurement of tree position can be used by a steering controller to compute the tree lateral offset from the harvester centerline. Then, the steering controller can execute a steering correction as needed if the tree offset is too large. The 51 METHOD 6 SONAR UNn' I A SET OF 5 I 35.5 cm SONAR UNITS\‘ I" 80 cm EOUA £§AC$$Y f 3 F“ 35,5 cm _- SHAKER BAR CUT AWAY APART 1524a“ —' TO SHOW SONAR UNIT _L 0% F: %\ . \MI I22§§I \\' ZZSSI IE ‘ 1-11 \\ A (2 SI I/é?‘ ::‘| I r \ TOP VIEW Figure 4.6. Method 6 - Tree Sensing Using Multiple Sonar Units to Determine Tree Position (Top View) 52 METHOD 6 L l A. 0% DIID 0/44’ 000 000 000 000 _SONAR UNn' I ‘\\/’ Sfiv' 45cm §§5;. $$§23 nan» FRONT VIEW Figure 4.7. Method 6 - Tree Sensing Using Multiple Sonar Units to Determine Tree Position (Front View) 53 distance of tree offset from the harvester centerline can be computed by knowing the sonar's position from the harvester centerline and the sonar measurement of range from the sonar unit to the tree. To compute the tree offset, subtract the sonar measurement of tree range from the sonar's position (distance from harvester centerline). The accuracy of this computation can also be estimated. If the sonar measurement is accurate to within :3 cm and if the sonar's position is accurate to within :1 cm then the tree lateral offset could be computed with an accuracy of :4 COL. 'This amount of inaccuracy is not excessive and is not expected to cause any problem with development of an automatic steering control system for the USDA apple harvester. As shown in Figure 4.6 this concept of tree sensing uses five sonar units. These sonar units are arranged so that the array of sonar units is 142.2 cm long and when operating at a ground speed of 0.8 km/h (0.5 nuni) the tree will move past all five sonar units in 6.4 seconds. Thus, by using five sonar units there will be 6.4 seconds of time available to make any needed steering corrections to keep the harvester aligned with the tree. And, based on data reported by Upchurch _e_t__a_1_2_ (1980), it is estimated that the harvester steering system and con- troller would have a short enough response time that any needed steering corrections can be made within the 6.4 second time span, using five dis- crete tree position measurements from the sonar sensing system. Thus, based on this preliminary evaluation, Method 6 is a feasible tree sensing system for the USDA apple harvester. To make a more complete study of the feasibility of this tree sensing method, a dynamic analysis of the complete steering control system is needed to verify that the 54 system re5ponse time is short enough to make any needed corrections within the available time of 6.4 seconds. 4.7 Final Selection of Tree Sensing Concept After reviewing the description of each of the six tree sensing concepts, Method 6, the system which uses the array of five sonar units, was selected for develOpment. This method was selected because based on a preliminary evaluation it is a feasible system. Also this sensing system does not have any moving parts which could fail due to excessive wear or due to the moving parts being fouled by debris. Another factor considered is that the sonar's position is 80 cm from the harvester centerline and this reduces the probability of the sonars being damaged by colliding into a tree. The electronic circuitry for the sonar sensing system is expected to have high reliability and adequate accuracy. When considering the computer software required to compute the tree position, the automatic steering controller needs to use the sonar's measurement of tree range and perform simple arithmetic operations. Thus, complex arithmetic algorithms or large table-look-up algorithms should not be required to compute tree position. 5.0 APPLE HARVESTER STEERING CONTROL SYSTEM The automatic steering control system for the USDA apple harvester has been developed with the tree sensing system that was described as Method 6 (Section 4.6). This tree sensing system uses an array of five sonar units mounted on the front of the harvester. Figure 5.1 is a diagram which shows the major components of the steering control system. This control system is designed to operate according to the following description. The steering control process begins with one of five sonar units detecting a tree. The sonar unit then sends a timing signal to the interface circuitry. The interface circuit is designed to convert the timing signal from the sonar unit into an eight bit binary number that repesents the measured distance to the tree. This binary number is an input number for the controller. The controller uses the binary number to determine if the tree trunk which is ahead of the harvester is aligned on the harvester centerline within an allowed tolerance. Another component of the steering control sytem is an absolute shaft encoder, (Model 76-OClO-4-E-1, Litton Inc.) which uses a ten bit gray code output. This shaft encoder is used to measure the angular position or steering angle of the front wheels of the harvester. Only Six bits of the ten bit binary code from the shaft encoder are used. The two least significant bits and the two most significant bits from the shaft encoder are not used. When using the six bits from the encoder, each output code represents 1.4 degrees of rotation of the encoder's shaft. 55 ONE OF FIVE SONAR CIRCUITS ”57> I EACH SONAR UNIT IS EOUALLY SPACED 35.5 cm APART (+80 cm-h 56 FORWARD P- “ TREE TRUNK I O<— B I E] I I! I .5 s BIT /' NUMBER ’ FRONT WHEEL “MING STEERING SIGNALS ANGLE sHAFT 20 mA ENCODER CURRENT LOOP :9 INTERFACE MODEL CIRCUITRY 76--GC10-E--1 +12 VDC LITTON INC. AUTO N BATTERv ESTATAR 1° 3'73 _ GREY CODE D" 7:31;: we +5voc CONTROLLER t 3 ‘— (F:- CONVERTERI— I RCA 3 ' COMPUTER -- CDP188691 -- l/O BOARD CDP188660 1 BIT 1 BIT . L I) SOLID SOLID STATE STATE BOTH RELAYS RELAY RELAY MODEL 630—2 ' TELEDYNE _ ’ DC CONVERTER MODEL UMC12~55200 SEMICONDUCTOR CIRCUITS INC. Ui—t—(nz VDC FROM ENGINE ELECTRICAL ‘—" HYDRAULIC SYSTEM » SOLENOID VALVE FOR LEFT AND RIGHT TURN Figure 5.1. Diagram of Major Components of the Automatic Steering Control System. 57 In order for the shaft encoder to measure the front wheel steering angle, the encoder is attached to a vertical shaft which turns as the wheels are turned (or steered) to a new steering angle. Turning of a wheel refers to wheel rotation about the wheel's vertical centerline for the purpose of steering the apple harvester. The encoder is mechanically connected to a shaft on the harvester steering system such that for each degree of rotation of the harvester's front wheels, the encoder shaft rotates one degree. The shaft encoder is calibrated so that when the harvester's front wheels are pointing straight ahead, the shaft encoder output is a gray code binary number of magnitude equal to 32, and this represents a steering angle of zero degrees. Then, when the front wheels steer left the encoder output number increases and when the front wheels steer right the encoder output number decreases. The next major component of steering control system is the controller. The controller is a microprocessor based computer (model CDP185691 with I/O board no. CDP185660, made by RCA). This is a CMOS type computer which uses less than 1 amp at +5 VDC and has high noise hmnunity. The computer is in a compact card case with dimensions of 13 cm by 4 cm by 10 cm. This computer also has a special switch that allows the computer to operate in a single-step mode and one of the computer cards contains six hexidecimial display digits for the computer address bus and data bus which are very helpful for debugging software and hardware. During operation of the steering control system, the controller executes any needed steering correction. When a steering correction is needed the controller sends a CMOS level signal to one of two solid state relays (Model 630-2 made by Teledyne). One relay is used to steer 58 right and the other relay is used to steer left. When a solid state relay is ON, a +12 VDC signal is sent to one of two hydraulic solenoid valves which controls the position of two linear actuators. The motion Of the linear actuators steer the harvester's front wheels. The hydraulic circuit diagram is shown in Figure 3.3. A proportional control algorithm was selected for the automatic steering control system. Basically, this proportional control algorithm computes a tree position error and then directs the wheels to turn to a steering angle or wheel position that is proportional to the tree position error. The following is a more detailed description of how this control algorithm operates. First, a sonar unit measures the range to a tree trunk. The controller receives a binary number from the interface circuitry that represents the range or distance from the sonar unit to the tree. The controller then computes the distance from the tree to the harvester centerline. Figure 5.2 shows a diagram of a sonar unit and a tree trunk. This diagram shows that the sonar unit is positioned 80 cm from the harvester centerline. Also shown is a tree trunk with a diameter of 4 cm and if the tree trunk is centered on the harvester centerline then the left edge of the tree trunk is 78 cm from the sonar unit. This tree is labeled TREE A in Figure 5.2 and also shown in this Figure is TREE B which is not on the harvester centerline. Therefore, the position error or tree offset for TREE B (in Figure 5.2) can be computed by: E = 78 - S (5.1) where E = tree position error (cm) S = sonar mesurement (cm) 59 TREE TRUNK CROSS SECTION \ /TREE A ‘ SONAR ‘ ...... ‘3 32“.?! .......... UNn' , —"’I k— 4 cm 4 78 cm —.. ‘7 80 cm + HARVESTER CENTERLINE /© ‘- 2 cm TREE TRUNK ‘\\\ OFFSET AWAY “‘55 3 FROM HARVESTER CENTERLINE e S a ‘1 78 cm a E = 78 - S E=TREE POSITION ERROR (cm) S= SONAR MEASUREMENT (cm) Figure 5.2. Diagram of Tree Position Error. 60 Note that tree trunk diameters will vary between 4>cm and 14 cm and therefore an average value of 75 may be used in equation 5.1 to compute error for a production version control algorithm that could be used in orchards containing trees which have the full range of tree trunk diameters. After the tree position error is computed, the control algorithm checks to see if the error is within an allowed range Of tree position error. For example the allowable tree position error may be :2 an although the best value must be determined by either a dynmnh: simulation of the control system or by experimental tests. If the tree position error exceeds the allowable tree position error, then the front wheels must be turned so that the error is reduced to a value that is within the allowed range of error. When the controller has determined that a steering correction is needed, the controller executes a turn so that the wheels turn to a position which is proportional to the tree position error. The wheels are turned in the direction which causes the harvester to move in the direction needed to reduce the tree position error» 'The controller continuously is checking the tree position error and as the tree position error decreases, the controller directs the wheels to turn to a position of smaller steering angle. Note that when the steering angle is zero, the wheels are pointed straight ahead and the shaft encoder code for the front wheel position is equal to 32. 61 Thus, a wheel position code for the new steering angle for the front wheels is computed by the controller using the equation: 0 = k (78 - S) + 32 (5.2) where D = Desired wheel position code K = Proportional gain factor S = Sonar measurement (cm) The desired wheel position code, 0, is a value that is scaled so that the controller can periodically compare the value of D directly with the value of the shaft encoder code to determine if the wheels are turned to the correct wheel position. The value for the gain factor, K, must be determined by analysis or tests. The tree sensing system has five sonar units and the one controller will determine one valid tree position measurement for each sonar unit. Each time a valid sonar measurement is made, a new value for D is computed. As soon as D is computed the controller will direct the wheels to turn in the approprite direction until the wheel shaft encoder code equals the value of D. The wheels remain at this steering angle until the next value of D is computed. When the value for D is computed using the tree position measurement from sonar number five, a time delay routine is used because sonar number five is the last sonar in the array. And, the harvester must then travel about 152 cm before the next value for D can be computed using the tree position measurement from sonar unit number one. For this time delay routine, one second is allowed for the front wheels to turn to the wheel position equal to the valIN: of D and the wheels are held at that position for the duration of 62 the time delay. Next, the wheels are turned back to the straight ahead position. The wheels stay in the straight ahead position until the next value of D is computed. The one second magnitude for the time delay was selected as an approximate value, because this value is close to the 1.5 second time duration required for a tree to move from one sonar unit to ‘the next sonar in the sonar array as the harvester travels at 0.8 km/h (0.5 mph). One Objective of this research was to design the steering control system for a cost less than $2,000. The major components of the control system and their costs were: RCA computer ($850); I/O board ($375); wheel position shaft encoder ($350); two solid state relays for steering the front wheels ($60); five solid state relays for activating the sonar units ($70); interface circuit board ($100); five sonar units ($300); and DC to DC converter ($100). Note that each sonarIufit:includes a Polaroid transducer and Polaroid circuit board which have a cost of $17 and $25 respectively. The cost to manufacture one complete sonar unit is about $60. The interface circuit has discrete electronic components and integrated circuits which cost about $40 and the estimated total cost to manufacture one complete interface circuit is $100. Therefore the total cost of these components of the steering control system is $2,205. This is $205 over the cost objective of $2,000. Note, that the steering control system as described in Chapter 8 includes a video display monitor and keyboard and these components were not included in the cost of the control system because they were used only for the development of the control system. The cost of the video display monitor and keyboard were $200 and $360 respectively. Chapter 8 also explains why the five solid relays were used to activate the sonar units. 63 There are two ways to reduce the cost of the control system. First, the I/O board could be eliminated and a special circuit board could be designed to do all of the I/O functions. The cost to build such an I/O circuit would be about $75. Second, the computer could be replaced by one single board computer (DP183601 by RCA) which has a cost of $325. Thus, if these two changes were made, the total cost of the control system would be $1,430. Therefore, this data indicates that the control system can be designed using electronic components which have a total cost less than $2,000. 5.1 Dynamic Considerations of the Automatic Steering Control System The automatic steering control system for the apple harvester has three major components that affect the dynamic behavior of the control system and these areithe kinematic motions of the apple harvester, the Inicroprocessor based controller and the wheel steering motor (hydraulic linear actuator). This type of control system may be called a regulator because the reference signal is set at a constant valuecfl 78. The reference signal of 78 was used because when the harvester's centerline is perfectly aligned with a straight tree row then the distance measured by the sonar unit is 78 cm. Thus, this control system controls the position of the harvester until the sonar measured value is 78 cm. Figure 5.3 shows a block diagram of the apple harvester's automatic steering control system. The controller used a proportional control element (Figure 5.3) where the harvester's wheels were turned to an angle which was proportional to the harvester's lateral position from a tree trunk. Note that the sonar sensing system was positioned ahead of the harvester's front wheels so that the harvester would be aligned on the tree trunk before the tree trunk passes through the inside space of 64 ZOFUmED |> NIP Z. mwhwm>m<1 m1... ”.0 20:50.. wh<25m000 In| I mK .vfi .MK mchdzdm 203.05. U_h<$_w2_x mmhmw>mgmx DFDD< any CD Emgmmwo xuopm m.m mgszu AEDV xzzm... mum... < O... .223 m._._QO._m> » Dm<>>m0u mmhmm>mm 0252mm EDE 324.55 I m m<20m 1 IIIIIIIIIIIIIIIIIIIIIIIIIII IIIU _ . D _ _ n so... _ _ mDoD _ DUMBO“. ON u 3.4m _ 20.:mOn— _ zmapqmmzz “ I mex3 I " $585 :35 _ ,_ DmEmmD I D 3. I D _ 29:8,. #5:; LT $10528 + D I w + NDDD Dz< cogs. “ to \ 20 " ZO_._..WO& 02—mwwhw AwwI>> _ PZMENJw JOEFZOU _ flux; _ ._ ['1 *— ASTB I TO couPUTER PORT .5 “m" r O 0 I 9 D O 13 A’ 2 ; _ J I II v D .3 '— D .0 R S 7. D 5 To a COUNTER '7 0" '5 , “'7 INOII 2 II. .5 "m OI r > p . , 3 “II. DELAY I’U‘- ; v 4 T34 ‘ 5 I?" RESET 6 TI 7 To —> n ‘— 8 I l CLOCK' CLOCK 16.066 5 COUNTR U Interface Circuit Diagram 80 until the sound pulse echo is received by the sonar circuit. An inter- face circuit is designed to count clock pulses only during the time that COO—NTR is at logic LOW. A clock frequency is then selected so that the clock signal produces one pulse during the time the ultrasonic sound pulse travels 2 cm. Since the sound pulse must travel out to the object and back to the transducer the sound pulse travels twice the magnitude of the distance to the object. Thus, for every centimeter of range to an object, the sound pulse must travel two centimeters. To determine the clock frequency the velocity of the speed of sound in air must be known. As explained by Gross (1978), the sound pulse travels at the speed of sound and is a function of air temperature according to the equation: V 331.5 + (0.607) T (m/sec) where T air temperature (0C) For air at a temperature of zero degrees Celsius, the velocity of an ultrasonic sound pulse is 331.4 m/sec. For the hypothetical case where the range from the sonar transducer to an object is 1 cm, the sound pulse must travel 2 cm and the time required for the sound pulse to travel 2 cm is 6.033 X 10'55. If a clock signal in an interface circuit is designed to produce one pulse every 6.033 X 10"5 s, then one clock pulse will occur for each centimeter of range to the Object. Using 6.033 X 10'55 as the period of the clock signal, the corresponding frequency is 16.575 kHz. The clock frequency selected for the interface was 16.67 kHz because it could be easily designed using a 1 MHz crystal to develop a lMHz clock signal and then dividing this clock signal by six hundred. Using this frequency of 16.67 KHz should only cause an error in the sonar measurement of less 81 than 1% for sonar measurement less than 150 cm with air temperature of 0°C. A block diagram of this circuit to produce the clock signal is shown in Figure 6.8. In order to develOp an 8 bit number that represents the distance to the object, the COL—INN? signal is used to start and stop the sending of the clock signal to a binary counter. The interface Circuit is designed such that when COUNTR is at logic LOW, the 16.67 kHz clock signal is sent to the counter. The clock signal is stopped from being sent to the counter when CENT—R is at logic HIGH. Refer to Figure 6.9 for the timing diagram of these signals in the interface circuit. Four monostables or one-shots are used in the interface circuit to produce time delay pulses and control signal pulses. These pulses have a duration of one millisecond each. The time delays are used to control the timing of several signals in the interface circuit. The first delay occurs when WIN—TR changes from LOW to HIGH. This gives the counter some time to settle because a ripple counter is used and it requires about five microseconds for the output of the counter to become valid. After the first delay, the signal ATB, is a one millisecond control signal pulse which is used to indicate that the 8 bit number on the counter is ready or valid. This ASTB signal is a control signal used to latch data into a steering controller input port. The ASTB pulse will latch an 8 bit number that represents the sonar measurement. After the ASTB signal pulse, another one millisecond delay pulse occurs. This second delay pulse is followed by a RESET signal pulse which is used to reset the counter. Note, that when five interface circuits are grouped together and built onto one circuit board then only one 16.66 kH clock signal is needed. This clock signal can be used by each of the five 82 —>l I6“ .I... _jmmmuuuuuuuuuwmuummmmmmm 18.566 KHz awn—TR \# J cw“. WWW ——I I——w NOTE: LARGE TIME SCALE BELOW COUNTR I DELAY 1 *1 ms ASTB :4 I- OELAv 2 ‘ _"I_I‘;_l m RESET Figure 6.9. Timing Diagram for Interface Circuit 83 interface circuits. This technique of using one common clock signal was used to develop one large interface circuit board that sends five binary numbers to the steering controller. 6.2 Sonar System Testing_and Results A sonar system was built and tests were performed to evaluate the accuracy of the range measurements of the sonar system. Tests were done at 25.5% and -1.0°C to check the effect that temperature has on the accuracy of the sonar measurements. Tests were also performed to check 'the approximate beam angle of the Polaroid ultrasonic transducer. The beam angle basically represents the sensing zone which is the area where an object can be detected by a given sonar unit. The objectives of the accuracy tests were to determine the accuracy of a sonar unit within the temperature range of 0.0°C and 32.2°C because these are the limits of the Operating temperatures for the sonar unit which are specified in the design requirements. Since a large temperature controlled chamber was not easily accessible, one test was performed outdoors where the outside air temperature was -1.0°C which was close to the operating temperature limit of 0.0°C. Another test was done indoors where the highest available air temperature was 25.5°C. This temperature was close enough to the upper operating temperature limit of 32.2°C that the sonar reading could be estimated using the assumption that the sonar measurement will change linearly with air temperature. The sonar reading should change linearly because the sound pulse that is transmitted by the sonar unit travels.at the speed of sound and the speed of sound changes linearly with the change in air temperature. Also it is assumed that there are negligible temperature effects on the sonar measurements that are caused by changES'hItMe 84 performance of the digital electronic circuitry. The manufacturer's operating temperature range for the integrated circuits used in the sonar system is -40°C to 85°C. The effect on the sonar accuracy due to the air relative humidity was not studied because a report by Gross (1978) indicated that the speed of sound changes less than 0.5% due to a change in the relative humidity from 0 to 100%. 6.2.1 Sonar System Accuracy Tests and Results Five sonar circuits and one interface circuit were constructed according to the circuit diagrams in Figures 6.5 and 6.8. These circuits were then tested to determine the accuracy of the sonar distance measurements. The test procedure was performed as follows. First, a flat metal target was positioned such that the target was centered on the sonar beam centerline at a specific distance or range from the ultrasonic transducer. The target was a piece of sheet metal that had a square surface area with dimensions of 12.7 cm by 12.7 cm. The air temperature was 25.50C. The size of the target was selected because it was assumed that the performance of the sonar system may be affected by the size of the surface area of the target, therefore this target size that was selected approximately represents the size of objects the sonar system may be used to detect. Preliminary tests have indicated that the sonar readings are more consistent at detecting an object with a large surface area. The target also has a hard flat surface and this type of surface was selected because it was assumed that this surface would be the best type surface for reflecting a sound pulse. Thus, this hard flat surface should result in sonar readings with the highest possible accuracy for a target of the same size and this accuracy should represent a base line 85 for the best accuracy for this sonar system. If other targets are used, the sonar accuracy may vary from the accuracy obtained in these tests. In order to collect the sonar data for this test, the sonar system was connected to a computer (RCA Model COP18S694) which uses an interpreted BASIC language and assembly language for the 1802 microprocessor. The computer was programmed to read four consecutive sonar measurements and store these measurements. These four measurements were for one specific position of the target. The computer program used assembly language subroutines so that the sonar data could be quickly read from the interface circuit and the values stored in the computer memory. The computer program also directed the sonar data to be printed onto a teletype and stored on cassette tape for further data analysis. Four sonar readings were collected for each position of the target and the range to the target was changed in 10 cm increments frmn a range of 30 cm to 150 cm. When the range to the target was between 75 cm and 85 cm the target was moved in 1 cm increments to obtain additional data points in this zone of target distances. Typical measurements by the apple harvester sonar sensing system will be in this range of 75 to 85 cm. The 150 cm upper limit on the target position was selected because this represents the maximum distance measurement required for the apple harvester sonar system. The lower limit of 30 cm for the target position was used because this was near the value of 28 cm which is the minimum reading that the sonar system is capable of making. Preliminary tests on the sonar system also indicated that for a given target position the sonar reading would oscillate between hm: values. For example if the sonar reading was initially 30 cm, then 86 sometimes the sonar reading would change back and forth between the values 30 and 31. Therefore, to show this variability in the sonar readings, the test was done by taking four sonar readings at each target position. The sonar data was plotted and Figure 6.10 shows this plot of sonar data versus the actual target range for sonar unit one. The data plots for the outer four sonar units are in Appendix A. The data for sonar unit one are in Table 6.1 and the data for the other four sonar units are in Tables in Appendix A. Table 6.1. Data from Accuracy Test for Sonar Unit One at 25.50C. Target Four Consecutive Average Actual Sonar Measurements Sonar Range Measurement cm cm cm 30 31 30 31 31 30.75 40 40 40 40 40 40.40 50 50 50 50 50 50.00 60 60 60 60 60 60.00 70 69 69 69 69 69.00 75 74 75 75 74 74.50 76 75 75 75 75 75.00 77 76 76 76 76 76.00 78 77 77 77 77 77.00 79 78 78 78 78 78.00 80 79 79 79 79 79.00 81 80 80 79 80 79.75 82 81 81 81 81 81.00‘ 83 82 82 82 82 82.00 84 82 83 83 83 82.75 85 84 84 84 84 84.00 90 89 89 89 89 89.00 100 98 98 98 98 98.00 110 109 108 108 108 108.25 120 118 117 117 117 117.25 130 127 127 127 127 127.00 140 137 137 136 137 136.75 150 147 147 147 146 146.75 87 ERROR - 4 cm 140.00 100.00 I SONAR UNn'NO.1 TEMPERATURE - 25.5°c * 120.00 L 1 100.00 LINE REPRESENTS THEORETICAL EXACT DATA 80.00 SONRR HERSUREI‘IENT (CHI 60.00 1 4p .00 N20.00 0.00 36.0 Mb ‘ Ibo. 00 Ibo. 00 140. 00 100.00 DISTRNCE BETIIJEEN TFIRoGET RND SONFIR UNIT (CH) Figure 6.10. Distance Data for Sonar Number One Accuracy Test at 25. 5 (I. 88 'The plot of the sonar data for sonar unit number one is typical of the plots for the other four sonar units. Also, each of the sonar data plots shows that the data points lay close to a straight line. Note, that in the data plot of Figure 6.10, at each target position the values of four sonar readings were plotted, therefore, when only one point is shown on the graph for a given target position, then that data point on the graph represents four sonar readings of equal value. For the case where two data points are shown on the graph for a given target position, then the four sonar readings are distributed between these two plotted sonar reading values. By inspection of Table 651it.can be determined how many sonar readings are represented by each plotted data point in Figure 6.10. Note that in Figure 6.10 the data show that the Inaximum sonar measurement error is 4 cm when the actual distance to the target was 150 cm. A linear regression analysis was done for the data from each of the five sonar units. The form of the equation of the straight line that best fits the data is: S = mx + b where S = predicted sonar reading (cm) m = lepe of the line x = actual target range (cm) b = intercept with S axis 89 The regression equations and correlations coefficient for each of the sonar units are: for Sonar 1 S (0.96521)X + 1.75116 (6.1) R 0.99992 = correlation coefficient for Sonar 2 S = (0.96237)X + 2.13578 (6.2) R = 0.99993 for Sonar 3 S = (0.96419)X + 1.86059 (6.3) R = 0.99990 for Sonar 4 S = (O.96609)X + 1.67628 (6.4) R = 0.99992 for Sonar 5 S = (0.96313)X + 2.09320 (6.5) R = 0.99991. This same test procedure to determine sonar range accuracy was also done for sonar unit number one and two at an air temperature of -1.0.:0.5°C. Table 6.2 shows the collected data for sonar unit one and Figure 6.11 shows a plot of this sonar distance data for sonar unit number one. See Appendix A for the data from sonar unit number 2. 90 Table 6.2 Data from accuracy test for sonar unit one at -1.0°C. Target Four Consecutive Average Sonar Actual Sonar Measurements Measurement Range cm cm cm 30 32 32 32 32 32.00 40 42 42 42 42 42.00 50 52 53 53 53 52.75 60 63 63 63 63 63.00 70 73 73 73 73 73.00 75 78 78 78 78 78.00 76 79 79 79 78 78.75 77 80 80 80 80 80.00 79 82 82 82 82 82.00 80 83 83 84 83 83.25 81 85 84 85 85 84.75 82 85 85 85 85 85.00 83 86 86 86 86 86.00 84 87 87 87 87 87.00 85 88 89 88 88 88.25 90 94 93 93 94 93.50 100 104 103 104 104 103.75 110 113 114 114 114 113.75 120 124 124 124 124 124.00 130 134 133 134 133 133.50 140 144 143 144 143 143.50 150 154 154 154 154 154.00 A linear regression was also dOne on this sonar data and the straight line equations that best fit the data are: Sonar 1 S = (1.0399)X + 1.98690 (6.6) R = 0.99990 Sonar 2 S = (1.01567)X + 1.69050 (6.7) R = 0.99993 91 J 180.00 + SONAR UNn'NO.1 t TEMPERATURE - ~1.o°c 4, ERROR - 4 cm 140.00 120.00 .9. 1 100.00 1 40' LINE REPRESENTS THEORETICAL EXACT DATA SONRR HERSUREHENT (CH) qLoo NLoo 40.00 1 ,gpdm 0.00 COCO Tea. 00 120.00 140. 00 160.00 DISTRNCE BETWEEN THROGET RND SONRR UNIT (CHI Figure 6.11. Distancg Data for Sonar Number One Accuracy Test at -1. 0 C. 92 The data of Figure 6.11 show that the maximum sonar measurement error was 4 cm although Table 6.2 shows that this maximum error occured at a target distance of 150 cm, and at several other target distances. The inaccuracy of the sonar data can be seen qualitatively in Figures 6.10 and 6.11. By inspection of the graph of sonar data in Figure 6.10 where the air temperature was 25.5%, it can be seen that the data points form a line that has a greater slope than the line which represents the theoretical exact data. This line for theoretical exact data is the line which represents sonar distance measurements which are equal to the actual distance from the sonar unit to the target. There- fore, any sonar data points which are on the theoretical exact lime shown in Figure 6.10 or 6.11 have zero error. Note, the regression ‘line through the data was not drawn through the data points because the line would make it difficult to see the small variations between data points. Therefore, since the data points of Figure 6.10 diverge from the line for theoretical exact data, the error of the sonar data increases as the range increases. This increasing error results because the 16.67 kHz clock frequency in the interface circuit is not the optimum frequency for the air temperature of 25.50C. 0n inspection of the sonar data of Figure 6.11 where the test air temperature was -1°C, it can be seen that the data points form a line that is nearly parallel to the line that represents the theoretical exact data. The sonar data is offset slightly from the theoretical line and this offset or error in the sonar data may be partially caused by using a clock frequency of 16.67 kHz instead of using a clock frequency of 16.55 kHz which would be the computed frequency for sonar measurement at -1°C 811‘ 'temperature. This computation of clock frequency was explained in section 6.1.2. Another possibility for the 93 offset is signal delays in the electronic circuitry. Further research is required to determine the cause for this error in this sonar data. On inspection of the data it is seen that the devefloped linear regression equations have a correlation coefficient almost equal to one, and therefore it can be concluded that the equations for the lines that were developed are good predictors of the sonar values for a given target position assuming the sonar readings were taken with the same conditions as used during this accuracy test. A value for R near one also indicates that most all of the data points lay near or on the regressicni line. It is also assumed that the accuracy for each of the five sonaT'IuIits is equivalent. This is shown by equations 6.1 to 6.5 being nearly equal, and by inspection of the data which show that there are negligible deviations between each of the five sonar unit's readings at each target position. Since the sonar data for all five sonar units was very similar for air temperature of 25.5%, only two sonar units were used for accuracy tests at air temperature of -1°C. An important point about the data in Figure 6.10 and 6.11, which are the sonar data at air temperatures of 25.50C and 1.0°C respectively, is that the maximum deviation of the sonar reading from the theoretical exact value is 4 cm. This deviation or error can be computed by: E=S-T (6.8) where E = error (cm) S = sonar reading (cm) T theoretical exact sonar reading (cm) Therefore it can be concluded that the tests indicate that the sonar units are accurate between :4 cm for the target distances of 30 cm to 94 150 cm and for air temperatures between -1.00 and 25.50C. Using linear extrapolation an error of 6 cm (error defined by equation 6.8) was computed for sonar measurements with air temperature of 32.2%. 'Therefore, the sonar system accuracy is expected to be within :6 cm for the temperatures between 0.0°C and 32.2°C, for sonar system readings in the range of 30 cm to 150 cm, and for the same conditions as used in this test. This error or inaccuracy of the sonar system is acceptable for use in the apple harvester steering control system because it is believed that with this sonar inaccuracy the apple harvester steering control system will be effective at controlling the harvester so that each tree trunk stays within the defined allowable zone as specified by the design requirements. This allowable zone is 45.7 cm wide and if the maximum tree trunk diameter is 14 cm then there would still be 31.7 cm of width available in the allowable zone that the tree must stay within. There- fore, this :6 cm inaccuracy in the sonar reading is not expected to pose a serious problem for controlling the apple harvester steering. Note, also that this 6 cm error is the maximum expected error at 32.20C and at sonar readings of 150 cm and that the expected error is less at lower temperatures and at smaller sonar reading values. 6.2.2 Sonar Beam Angle Tests and Results A test was done to check the beam angle of a sonar unit. This beam angle is important because it indicates the area in front of the sonar 'transducer'where an object may be detected. Also, the test was done to determine if there is a large error in a sonar reading when the object is near the outer edge of the sonar beam. The test procedure for measuring the beam angle was made based on the assumption that two lines 95 extending from the sonar transducer approximately define the sonar beam angle. The beam angle defines an area where objects that are inside the beam angle are detected by the sonar unit and objects which are outside the sonar beam angle are not detected. A cylindrical tube was selected as a target for this test because it was assumed that this type of target would be the best for reflecting the sound pulse back in the direction of the transducer even when the target is offset from the sonar beam centerline. This assumption was based in preliminary tests which indicated that a flat surface can reflect a sound pulse in a direction away from the sonar transducer and then the object would not be detected beause the transducer did not receive a strong echo. Therefore, it was believed that a cylindrical shaped target would provide data that represent the largest beam angle of the transducer. The size of the target for this test was selected to approximately represent an apple tree trunk. The basic procedure of the test was to begin the test with the target initially at a position offset from the beam centerline so that the sonar did not detect the target. Then the target was moved toward the beam centerline until the sonar unit detected the target. This was done at several positions on each side of the sonar beam so that data could be collected to define two lines which describe the beam angle. The test was performed as follows. A piece of steel tubing 3.8 cm in diameter and 61 cm long was used as a target. A guide line was positioned above the sonar transducer to be used as a reference line so that the position coordinates of the target could be measured from the reference line. The reference line was positioned so that it was approximately at the centerline of the transducer's beam. Figure 6.12 96 ° 0 152.25 mo 0 155.25 152.25 150.00 mo 1:10 1.10 O . ‘ o . O 1151” 111.00 11325 125.25 10000 100 LINE DRAWN 0v INSPECTION THROUGH TI-IE DATA POINTS so TRANSDUCER BEAM ANGLE 22.: DEGREES O D .0 . . 0200 f. .925 05.50 0000 5300 10 LINE DRAwN TEST REEERENCE LINE 011 INSPECTION THROUGH THE DATA POINTS 50 5200 50 O O 5500 f. 5400 5000 5100 00 O O O .5. m: r” I... 3175 ans LEGEND II NUMBERS UNDER DATA POINTS EOUALS AVERAGE SONAR VALUE 3—*’—3 E 2) ° - FIRST DETECTED TARGET LOCATION TARGET COORDINATE (uni SI 0 ' LOCATION WHERE SONAR VALUE HAS RELATIVE ERROR BETWEEN 1 AND 3 I I T T I 1 I V 10 30 20 10 SONAR UNIT 10 20 30 40 YL -. LEET SIDE 3 YR _. RIGwT SIDE TARGET COORDINATE (Om) TARGET COORDINATE (cm) Figure 6.12. Plot of Sonar Data to Determine Transducer Beam Angle. Target used was a steel cylinder of 3.8 cm. diameter and 70 cm. long. 97 shows the sonar unit and in this figure the x-axis represents the reference line. The target was positioned with the long axis of the tube held vertically and the target was initially placed 30 cm ahead of the transducer. The target was offset to the right, perpendicular to the reference line. This 30 cm distance and the offset were designated as the X and YR coordinates of the target respectively. When the target was on the left side of the reference line, the target coordinates were X and YL. The coordinates were measured from the reference line to the target's vertical centerline that was scribed on the surface of the steel tube that was used as the target. The target was initially at a YR coordinate position that was out of the sonar beam angle, thus the target was undetected by the sonar unit. When the sonar did not detect an object in the beam angle (or sensing zone) the sonar reading value was between 213 and 215 cm. Next, the target's offset was decreased by moving the target toward the reference line in increments of one centi- meter. These small increments were used to accurately locate the edge of the sonar beam. Then at each coordinate position of the target four consecutive sonar measurement values were read by a small computer that was connected to the interface circuit of the sonar system. This computer was the same one as described in section 6.2.1. The target was moved on this YR coordinate axis until the target was directly in front of the sonar transducer. The target was then moved to a new X coordinate position and the procedure was repeated. This test was done on both the right and left side of the reference line to determine the locations where the target would be detected by the sonar transducer. This procedure was performed for X coordinates of values of 30, 50, 80, 98 110, and 150 cm. These X coordinate values were selected to obtain data points through the full range of 30 to 150 cm which would define two lines that represents the boundaries of the sonar unit sensing zone. The sonar data was printed on a teletype and stored on cassette tape. The purpose of performing the test was to determine the beam angle, or the zone of sensing, for the Polaroid ultrasonic transducer. This beam angle was determined by plotting the coordinate positions of the target where the target was first detected. A plot of this data is shown in Figure 6.12. These coordinate positions where the target was first detected indicate the outer edge of the beam angle. A line was drawn through these data points and the beam angle was determined graphically to be 22 degrees. Also, shown in this plot is the average of the four sonar values for each target coordinate position and the plot shows the average sonar value when the target is near the center of the zone. Notice that there is a major deviation between the average sonar value where the target was first detected and the average sonar value where the target was at the reference line. This indicates that there is additional error in the sonar readings when the target is at the outer edge of the sonar beam angle. It was observed during this test that as the target was moved away from the sonar beam edge, toward the reference line, this additional error rapidly decreases. To illustrate how this additional error decreases as the target position was changed, another set of target positions were plotted and are shown in Figure 6.12. This set of target positions were those target locations where the average sonar value was greater than the sonar value near the reference line by 1 to 3 cm. 0n inspecting this plotted data, there are only a few centimeters 99 separating these target positions and the positions that represent the outer edge of the sonar beam angle. Notice, that at the target location at the outer edge of the sonar beam the measured sonar value is not very accurate and thus to obtain a more accurate sonar value, the target should be well inside the sensing zone. Since these beam angle data indicate that the sonar readings are not accurate when the object being sensed is on the edge of the sonar beam angle, then some method must be used to compensate for this in order to achieve effective operation of the apple harvester steering control system. During the operation of the sonar sensing system the apple harvester will be moving down a tree row at a constant 0.8 km/h (0.5 mph) and an apple tree trunk will move across the sonar beam in a direction approximately perpendicular to the sonar beam centerline. This means that if the steering controller is continuously sampling the sonar readings from a particular sonar unit then the tree will be detected at the edge of the sonar beam and these may be erroneous sonar readings. One technique that may solve this problem is to delay using any sonar readings until the tree has moved into the sonar beam angle and away from the edge of the sonar beam angle. This may be accomplished by sampling the sonar readings until the sonar value is less than 150 cm. When this value is obtained it is assumed that the tree has entered the beam angle and the sonar is Close to the edge of the sonar beam angle. After the tree is detected to be in the beam then the steering controller can execute a time delay by waiting until five sonar values have been sent to the steering controller and this would take 0.23 s because a sonar measurement is made every 0.046 s. Four of the five sonar values are not used because they may be erroneous. The 100 fifth sonar reading is assumed to be a valid sonar reading. Then once a valid sonar reading has been made, no further readings are made by that particular sonar unit and this would allow the tree to move out of the sonar beam angle without taking any sonar readings near the rear edge of the sonar beam angle where the sonar readings may also have a large error. This process repeats with the next sonar unit in the array of five sonar-IHIits. The next sonar unit waits until it detects the tree trunk entering the sonar beam angle and then this sensing process begins again. 7.0 SIMULATION MODEL WITH INTERACTIVE COMPUTER GRAPHICS A simulation model with interactive computer graphics was applied to design an algorithm for the harvester's automatic steering control system. Computer facilities of the Case Center for Computer Aided Design at Michigan State University were used to simulate the harvester's motion. A computer model was developed to simulate the motion of the apple harvester and the control tasks of a microprocessor based steering control system. The simulated motion of the harvester, as it traveled over a row of apple trees, was displayed on a graphics terminal. The use of computer graphics allowed the vehicle's response to be quickly displayed and analyzed. The influence of the system variables on the vehicle response was easily simulated and displayed. The most important design consideration for the performance of the steering control algorithm was the alignment of the harvester with respect to the apple tree trunks. Figure 3.1 is a diagram of the allowable zone which is 45 cm wide and extends from the front wheels to the rear wheels. A design requirement for the steering control system was that the tree trunk must stay within this allowable zone as the harvester passed over the tree. 1The Albert H. Case Center for Computer-Aided Design is a college wide facility. Dr. James Bernard, Professor of Mechanical Engineering, is the director. Software was developed at the Case Center by Mark Zykin, Case Center computer technician. 101 102 The computer facility used to develop the simulation model was the Case Center for Computer Aided Design which was established in the College of Engineering for research and teaching in the areas of Computer-Aided Design/Computer Aided Manufacturing (CAD/CAM). The CAD System is built around a PRIME 750 computer. The PRIME 750 computer has two 80 MB disk drives and services a color graphics terminal, a number of graphics and alphanumeric terminals, four dial-in lines, a digitizer and a Printronix printer. The PRIME 750 had compilers for both Fortran IV and Fortran‘V. Fortran IV was used for the automatic steering control simulation. The main software packages that were used in the steering simulation were the PLOT 10, AGII, UTILV, and IMSLS. The PLOT 10 package UM‘TEKV Library), developed by the Tektronix Corporation, does the basic drawing 11f graphics data. The AGII program, also by Tektronix, calls PLOT 10, scales the graphics screen, draws the frame on the graphics screen and draws the labels. The program UTILV, which was developed at the Case Center, is a special program which allows other standard software library to run on the PRIME 750. The IMSLS software package, which was created by the IMSL Corporation, is a set of single precision mathematical subroutines. Subroutine DVERK, from the IMSLS package, was used for solving differential equations which were used to model the harvester motion. 7. 1 Model Requirements” A simulation model was needed to predict the motion of the harvester as it moved over a tree row with the front wheels steered by a microprocessor based steering control system. This simulation model was needed to reduce the time required to design the steering control 103 algorithm for the harvester's steering control systan. lTfis control algorithm was converted to a computer program for the harvester's micro- processor based steering controller. The design of the steering control algorithm is affected by many parameters, therefore, the following parameters were selected as variables in the simulation model. The selected variables were harvester velocity, number of trees, X - Y position coordinates of trees, geometry of steering system, wheel steering rate (angular velocity), parameters of the control algorithm, measurement rate (cycles per second) of tree sensing system and position of the sensors relative to the harvester. To easily determine if the control system had satisfied the harvester's performance requirements, a graphical display of the harvester motion was needed. Therefore, two graphical display routines were selected for development. The first graphical display routine, called Mode 1, was designed to use the harvester's allowable zone (as defined by Figure 3.1) as a frame of reference. This display routine uses the allowable zone of the harvester as a moving frame of reference. As the harvester moved along the tree row, the graphical routine plotted the position of the tree trunks relative to the allowable zone. The scaling of the graphical output is the same in both the X and Y directions. Thus, there was no distortion due to unequal scaling and a Mode 1 output made it easy to determine if the tree was within the allowable zone as the harvester moved along a tree row. Figure 7.1 is an example of a Mode 1 output. The graphical output from the model was observed on a graphics terminal and a printer was used to make c0pies of the graphical output. 104 .Ampmm>cm; Asp Co Come mgu op Cmpmm>gmg Us» Co ucogy Us» EoCC mmmmma mega 05p mm umupopa m? Cmumm>gw5 059 D» m>wum_mg cowawmoa m.mmgp m .umummamg mw _oDEAm cm>wm m m< o.H I x EAL; eowAepzeAm UEA sate Asapeo _EUCEARLA A was: H.A mesmmm =29 mmhmw>m<1 ”.0 m_._.<._wm_ SI. 5. e... . e... 53...... 20:50.. mwpmm>mtoo._m> W DIE D» DEERE DDCO: I. 3. I < w m v.22: mum» NEEDED < m5 ADE. IO 3 I v. m w. 3535? .6225 :05 3.. DD”: I A . u .N. . MAR mum... .II 0 3 . . ; «A. much I + m w m o E DDS I x . w M. o . Dzjmmtho 55m>mmos memm>gms ms» mm mcwpgmucmu Cmumm>gmg asp ucmmmgamg mucmsmmm mew, mmmzp o.H I g EDP; eowpe_=ewm 0;» SOLE usauso PRUAEAALA N ace: N.A mesmwm =20. ZOFOmEOIX 2. 30m— mwmh UZOA< wavy—(#20 .xxn" new. -.x«:. Dav. Ann. .xmz. Dow ADD .xx. new a . Ea NNI < c3822: A SEE Dz» “_O $3553. a .20.:«2 much. m4m<>>044< "4 m .2 28. 88.: . DE 3585: .532» w. 226 #20353”... I 3 IV. #523.» m2... =93 “.0 L w. A . , 20.555 . n . DD<¢._. cwhmm>¢<1 DA 00.”. oo.mA oo.mA Doug oo.m oo.. oo.~ I? h LI 00:96” OO'SZ I 9 OO' 1 OO‘S <.-.m30: mm»; OZOJ< moz.:oo._m> 223:... mm:.-. d éésa ._ P d 4m>m<1 “—0 P20": m.o u x EOE; H aeoz - compm_=swm co “amuse _muwgaacu OH.“ aL=m_u =29 mmhwm>m<1 m0 m_._.<._m¢ can can _8— 33883 b P D b I b P I d d I4 d J 1 4 I! 4 4 q 1 1 1 d d d d d d a d d d d 1 1 d I d d d A d a 4“ d EOE m.9 .55. Q: n >._._UO..w> “ ON u < m md u v. u w U . «N ”.2 . K K 0.x. OX 0: r + Ix I“ ~ I fill? .. .._ + k LI+.L. SILIK 3...». LDIX 0 K0 0 JMV D . . D x I u . 4+ I 4A; I «a.-. C ~...L Q I H.144. 0 rd + A + A + A + A... A... Am.o :- DP L cu L 4w>tood> o~u< . m... u v. $32"... .3 29535 A E Q8 u 3.04m L 26: 3250 XZDm—P mum; M08 3381. :IO ESLNBO WOHd NOILOBUIO-A NI BONVLSIO (“19) 130 m.o I x EDS; H «no: - cowpm_=swm co Ozapzo quwcaoau NH.“ «Lamwa =29 Ewhmm>m<1 no m_._.<._wm sea 2.3 2.5 Bu 2.: bmmefifiop a EOE m.9 .55. ad u >._._UO.._m> O.Nu< mduv. O L A D mmpmm>ml m" O Figure 9.1 Diagram of Simulated Tree Stand Used in Steering Control System Tests 151 value of tree position error. This reduced the amplitude of oscilation of the front of the harvester. The change was made in the circuit where the wheel position data lines are connected to the wheel shaft encoder. The data lines were disconnected and then reconnected in a different order. 0f the ten data lines available from the shaft encoder, lines numbered 2 through 7 were used as data inputs to the steering con- troller. Line number 1 represented the shaft encoder's least signifi- cant data bit. By making this change, the shaft on the shaft encoder rotated approximately 0.7 degree for each change in the value of the encoder's output number. Before this change was made the shaft rotated approximately 1.4 degrees for each change of the encoder's output code. Note, that the control algorithm was written to turn the wheel by one wheel position code for each 2 cm of tree position error. Thus, after the modification was completed the wheel rotated about 0.7 degree for each 2 cm of tree position error that was computed by the steering controller. Another change was made to improve the reliability of making valid sonar measurements. A change was made to the software which provided a longer time delay for the microprocessor's measurement sequence for each sonar unit. This longer delay was used to allow sufficient time so that the tree stand would be more near the center of the sensing zone when the last sonar measurement in the program sequence was made. The computer program achieved a time delay initially by waiting until three sonar values were read in successive order. To make a longer time delay the progam was changed so that the microprocessor read six sonar values in successive order. The last sonar reading in this sequence was used to compute the tree position error. This change was made because a 152 sonar value of 109 cm was measured intermittently during the tests with a straight row. These sonar values of 109 cm were erroneous and they caused deviations from the typical path travel by the harvester. This software change was used for all tests except the straight row tests. Also, when the alternator was disconnected so that it was not charging the harvester's battery, the sonar measurements appeared to be more reliable. The alternator was disconnected during all of the curved row tests. 9.2 Test Procedure To evaluate the alignment of the harvester with each tree stand as the harvester moved over the row, data was needed that described the path traveled by the harvester. This data was obtained by attaching felt-tip pens to the harvester's frame and these pens drew lines on a long paper strip to record the path of the harvester. These pens were attached using a mechanical linkage as shown in Figure 9.2. This linkage used a hinge so that as the harvester frame moved up or down the pen would stay in contact with the paper. A small caster wheel was used to support the weight of the linkage arm. Each pen was supported in a steel tube and a small spring pressed on the top of the pen so that there was a small force downward on the pen to keep the pen in contact with the paper. Two pens were attached to the frame of the harvester so that data could be obtained for the path of the front and rear of the harvester. Figure 9.3 shows a plan view of the harvester and shows that the pens were located at each of the wheel centerlines. The pens were positioned so that they were offset 60 cm from the harvester's centerline to provide clearance so the tree stands could pass through the center space of the harvester. 153 FORWARD FRONT WHEEL SONAR = S r210 (h-h--- ) I . 25‘ cm ' I100m) PEN HOLDER _/ ///f/ CASTER FRONT PIANO HINGE WHEEL 53 cm REAR 72 cm I FRONT WHEEL PLAN VIEW ------.--‘--- I 51.5 cm ‘- o.5 50"“ = ‘ 07.5 n) PIANO HINGE PMER STRIP Figure 9.2 Diagram of Linkage that Supports the Front Pen \\ . ’/ ’/ Z FRONT PEN REAR PEN \///A 154 OF DIRECTION TRAVE L I\ g R 2 @\ /A 409 m (13.4ft.I 60 cm-) | 2.59 m I&5ftI I PENS WERE USED TO DRAW LINES ON A STRIP OF PAPER TO RECORD THE PATH TRAVELED BY THE HARVESTER Figure 9.3 Plan View of Harvester Showing Pen Locations for Performance Tests 155 The harvester was tested on a straight row as shown in Figure 9.4. A wooden fence 15.2 m (50.0 ft) in length was used at the beginning of each test to get the harvester into alignment with the row. Since the fence was a continuous surface, the sonar system was able to bring the harvester into accurate alignment with the fence in a short amount of travel. Thus, the fence was used to obtain consistency in the procedure. The data from the three test runs using a straight row of tree stands were evaluated to determine if the control system performance was consistent. A straight row test was done using tree stands that were set on a concrete driveway. A straight seam in the concrete was used as a line that represented the X-axis. The tree stands were positioned in the Y-direction by measuring from this straight line in the concrete. The paper was rolled along the row so that the pens were approximately centered on the paper. The paper was held down by laying strips of steel on both edges of the paper along the whole length of the row to prevent the paper from blowing away. Each pen on the harvester drew a continuous line as the harvester traveled along the row. Data to describe the path of the harvester were obtained from the paper which had a set of two lines for each run. The two pens on the harvester were different colors so that the path of the front and rear could be easily analyzed. Data points consisting of X-Y coordinate values were obtained by manually measuring to specific points on the line drawn by the front pen on the harvester. The distance along the X-axis between the measured points varied when the path curvature varied. A total of approximately seventy data points were measured for each test run and approximately 4 hours were required to measure and 156 _ x = 2743 o \ TREE STAND # 9 x = 2438 o x = 2133 O x = 1828 o x = 1524 . X =1219 . x = 914 , x = 610 . x = 304 . 7\ TREE STAND #1 Y ‘—__ CONHNUOUS WOODEN PAPER ___S.-'—__> FENCE mrvaE L. Figure 9.4 Plan View of Test Configuration Used For Straight Row Test 157 record the data points for a group of three test runs. The specific data points were selected such that a straight line drawn through two adjacent data points approximated the actual line drawn by the pen on the front of the harvester. The felt tip pens drew lines which were about 0.5 cm wide and the coordinates of each data point were measured to the approximate center of the lines drawn by the pens. Ifiw'each front data point a specific point was marked on the test paper that represented the position of the rear pen. This was done by using a long straight stick that had a length of 409 cm which was the distance between the harvester's two pens. One end of the stick was placed on the front data point and the other end was placed so that it was on the line drawn by the rear pen. The back end of the stick thus indicated the position of the rear pen that corresponds to the front data point which was made by the front pen. Thus, these two data points define the position of the harvester. For the rear data point, the Y coordinate was measured and the X coordinate was computed by subtracting 409 cm ownn the front data point X-coordinate. Three test runs were made with a straight row and data were recorded. The harvester was also tested using a curved row on the same concrete driveway and the same procedure was used to obtain the data points. For the test on a curved path the X-coordinate for the harvester's rear-pen data point was computed. The curved row for this test had a radius of curvature of 121.9 m (400 ft) which was the design requirement for the harvester. For this test the wooden fence was used in the same way as the test for the straight row. The trees were positioned so that they curved to the right and the X-Y coordinates for these trees are shown in Table 9.1. 158 Table 9.1 The X-Y Coordinate Position of Tree Stands For The Curved Row Tests Tree Stand X Y Number (cm) (cm) 1 304 -0 2 609 0 3 914 -4 4 1218 -15 5 1523 -34 6 1826 -61 7 2129 -95 8 2431 -137 9 2732 -186 Another test was done on a campus lawn with the harvester traveling over a curved row. A string was used to make a straight reference line and the tree stands were positioned by measuring from this reference line. The procedure for obtaining data points was the same as described for the other tests. The tree stands for this test were positioned at the coordinates of Table 9.1. The wooden fence was also used as described in the straight row test. The last test was done with a course that had a step change in the positicni of the tree stands. The row consisted of the wooden fence (as was used in other tests) and nine tree stands. The first three tree stands were on the X-axis and thus their Y-coordinate value was zero. The other six tree stands in the row had a Y-coordinate value of 8 cm. The X-coordinate values for the tree stands were the same as used in the straight row test. 159 9.3 Results of Straight Row Tests In order to determine the alignment of the harvestercnufing the straight row tests, the data from each test were used to plot two lines as shown in Figures 9.5, 9.6, and 9.7. In each figure one of the lines represents the path traveled by a point on the harvester's centerline at the position between the front wheels (point A, Figure 7.5) and the other line represents the path traveled by a point on the harvester's centerline at the position between the rear wheels (point 8, Figure 7.5). For each test the paths traveled by points A and B were plotted by the Calcomp plotter. A computer program was used to compute the X-Y coordinates for points A and B using test data and these coordinates were used to plot the two lines that represent the paths traveled by the front and the rear of the harvester. To verify that the computer program computed the coordinates correctly, a full scale drawing was drawn manually using a set of X-Y coordinates for one harvester position using data from a curved row test. This drawing demonstrated that the computed position coordinates were within 1 cm of the positions of points A and B which were determined graphically. This indicated that the computer program was correct. Inspection of Figure 9.5 shows that the harvester deviated at most 5 cm to the right of the row centerline and 0.5 cm to the left. Due to inaccuracy in collecting the test data and due to round-off error in the computations, the plotted data has a tolerance on the accuracy of :1 cm in the X and Y direction. This data was also plotted with unequal scale factors in the X and Y direction so that the harvester's travel in the Y direction could be easily seen in the data plot. Notice that since the scaling factors for the data plot are unequal, the paths shown in the 16C) oo.m»w H Canasz pmmh Zomuugmwmgum so» Emgmm>gmz ecu Lo new; m.m wgamwu .o—1 .:0. zo-uu¢_onx z— mozc~m~o oo.om~ ao.mmw oo.om« oo.u». oo.omL oo.mmwl oo.om_ oo.»m 99.5» oo.m~ oo.o P Dzgoz 9;» $9 spam o.m 9L9mwm ~o 9. 9 99.m~ 99.9 99.9w- 99.9m r b .o—l .zu. zo-uu¢~oix z~ uuzchm 99.9»9 99.9»9 99.9w“ 99.9mm 99.9»m 99.9mm 99.9mm. 99.9m. 99.»9 9 mhwmozoonmo291 9;“ Lo cum; N.m mgsmwm .o_- .zu. zo_bum¢_oix z. uuchm_o 99.9mml 99.9w“ 99.9mm 99.9». 99.9». 99.9m_ 99.9p. 99.»p 99.m9 99.m~ 99.9 99.9w» 99.9m D mhumozooumo ......... VP I......I.III1-1%.).--“me Dm<3¢0g I 00' 0'- oo-oéF oo’o I I T 00' 08 oo-os-' (W3) NOIIDEHIO-A 163 plot are distorted and do not represent the true shape of the lines for the actual paths for points A and B. To visualize the scaling factors used in these data plots, notice that the round symbols shown in the data plot (Figure 9.5) have dimensions of 1.6 cm in the Y-direction and 40 cm in the X-direction. Figure 9.6 shows the harvester's path for the second straight row test and the paths traveled by the front and the rear of the harvester did not exceed y=2.5 cm and y=-3cm. For the third straight row test, Figure 9.7 shows that the front and rear points of the harvester did not exceed y=1.0 cm and y=-3.0 cm. The data from these three tests shows that the automatic steering control system was effective at keeping the harvester aligned with the row centerline within the tolerance specified by the design requirements of Chapter 3. The required tolerance for alignment was :20 cm. 9.4 Results of the Curved Row Tests The alignment of the harvester for the curved row tests was analyzed by plotting the data using the same methods as described for the straight row tests. The paths of the front and the rear of the harvester are plotted for each of the three tests and these plots are shown in Figure 9.8, 9.9 and 9.10. These tests were done on a concrete driveway. Notice in these plots that the X and Y scaling factors are not equal. The round symbols in these plots have a dimension of 6.4 cm in the Y-direction and 40 cm in the X-direction. The data plot of Figure 9.8 is the first test with the curved row and this plot shows the front of the harvester had a maximum deviation in the Y direction from the center of a tree stand of 7.5 cm. This value of 7.5 cm was obtained by measuring the data plot. Note that due to the distortion in this plotted data, the value measured on the data 1(54 x939>wgo mumgocou 9 :9 H Cmnssz «99p 39m 99>Lau Com Cmpmm>gmz 99p $9 EH99 w.m mgsmwu .o—n Azu. zo~huuz_oix z— muzchm_o 99.9»9 99.9mm 99.9w“ 99.9n9 99.9». 99.9». 99.9». 99.93. 99.»9 99.59 99.m~ 99.9 99.9wwl 99.9. vaMuWAIII. :3.c¢ oo-ooa-' 5 ~§——N d oo 03?; oo'osf- __> I 00.06- C(mm I 00'07- a 9 ozfim mm...» .59.". Sm... _Eums I 9 9. ///vnv K /PIII EOE. _ALS 99L T 00‘0 I 00'09 Dmwgo mumgocou 9 :o N E99592 p99H 39m 99>Lau Com Cmpmw>g9z 9;» m9 :u99 m.m mgamwu .o—l .zu. zo_humz~on z— muzchw_o 99.9»9 99.9mm 99.9wu 99.9au 99.9». 99.9mm 99.9m— 99.9m. 99.m9 99.mu 99.m~ 99u9 99.9ww 99.9 J w. o e 9 I. w a m I. .Eomé a SIR 9 r. M W a 0 .llll'IVJ TW . .. m939>_Co mumgucou 9 :9 m E99592 ummp zom 99>Lzu EoL Emu99>g9z 9;» L9 9999 oH.m mgzmwu n:- 38. 22:6":on 99.9.99 99.39 99.99.“... 99.9%.. 99.9... . 99.9». 99.9w. 99.99F. z._.9.szS “mm. 99.»... 99._9 99.9w. 99.9.-. e m w. r. w e .m T. m .a. m .9 am“ .1. mm m299 292 29999929: 999 29 9999 NH.m 929mw9 .9.. .29. 29.2999.9-x.z. 992929.9 co.m#w co.o&w ca.mmu co.cmw oo.m#— OO.O&WII OO.@W— oo.om— oc.mr oa.mm co.mN oouo co.me Oo.omJ a 0 o 99 . w 222.. 992.25 ”995.999 925.99 I. m 9 m r- m o 9 3 nu Jrnu Eu OK ml .1. mm... WW 9 2<92 .wm .9 mm“ .1. Eu N 92099 a x .. HIB 3 III Ewan. I I 08 9235.202 _AEo 999 v— mm mm N m“ u V 00'0? T 00'00 T 00'021 170 9299 999599 9 99 N 299592 9999 299 99>299 299 29999>292 999 99 9999 mH.m 929999 .99.. :89 onpuumHn—Ix z~ muzcbeo 99.9.29 99.999 99.9999 9999.9 99.9... 99.99. 99.99t. 99.99F. 99.99. 99.99 99.99 8.9 99.999. 99.99-. 9 9 o 2>><._ mans—<0 "wo999 99» F999: 99999pzewm 9;» >9 999999999 99999>99I 999 $9 9999 99.9 999999 95. zoEEEoux 993:: 995922: 999. 999. 98 .. - n p a d u u d u u d E QPNanDR—(cuinuc Om>CDO 20:34:25 0 ¢999 999 $9 9999 99.9 999999 .093 _zu. zo.~um¢~cux z— wuzchm—o 99. 2.9 99.999 8.99.9 99.999 8.2—F 99.9m— 99.9fl 99.99. 9999. 99.9% 99.9r9 99.p9 99.9w. 99.99-. M .9 v. m 9 r. m .9 4". am” 9.... B N .u. 3 m3 cu> m WM Eoan> Eoopu> W A NW... 99<3999 m :9 r 0 175 Y-direction. This data of Figure 9.15 shows that the steering control system was able to drive over the tree row with an 8 cm step change while maintaining the proper alignment as specified by the design requirement. The results from the simulation model for a row with a step change were compared with the harvester's actual performance. Figure 9.16 shows the path of the harvester as predicted by the simulation model with a time constant of 2.0 seconds as described in Chapter 7. Figure 9.16 shows that the motion of the harvester in the Y direction has a faster response than the harvester's actual response shown in Figure 9.15. The simulation model showed that as the front of the harvester moved past the~step in the row that the maximum deviation of the front of the harvester from the row centerline was 3 cm and then the front of the harvester oscilated between y=10 cm and y=4 cm. After the simulation had run for 8 trees past the step-change, the simulation shows that the oscilation continued. This amplitude of oscilation predicted by the simulation model varied between 2 cm and 3 cm. Figure 9.16 shows that the rear of the harvester stayed within 2 cm of the row centerline as the rear point of the harvester moved past the step change in the row. Figure 9.16 shows that the response of the front and the rear of the harvester as predicted by the simulation mode, was faster than the response of the actual harvester. The actual harvester may have responded slower due to misalignment between the front and rear wheels. If the front and rear wheels were perfectly aligned then the centerlines through the left front and left rear wheels would lie in a single plane and the same would occur for the wheels on the right side. 176 The simulation model assumes that when the front wheel steering angle is zero, then the wheels are perfectly aligned and thus the paths of the front and rear points on the harvester centerline would lie on the same straight line. For the actual harvester, the task of aligning both the front and the rear wheels was difficult. The technique used to align the wheel was to allow the steering control system to followrthe wooden fence described previously. The harvester's pens were used to observe the path of the front and rear points on the harvester. Adjustments were made to change the specified zero position of either the front or the rear wheels. Thor the front wheels the shaft encoder connection to the front wheel shaft was used to change the centered position of the wheels. For the rear wheels a limit switch assembly was used to keep the wheels at the centered position and wheel position was changed by moving the limit switch assembly. After a wheel position change was made, the harvester was tested to determine if the harvester traveled along the fence such that the lines drawn by the two pens were close together. Also, it was observed if the steering controller needed to turn the front wheels in order to follow the fence. If the two lines drawn by the harvester pens were not close together (within 6 cm) then the process was repeated. Also, if the steering controller had to consistently turn the front wheels to keep the front centered on the fence then the front wheel center position was adjusted. The factors that may contribute to the harvester's not responding as fast as the model are deformation of the tires when turning; loose- ness or "play" in the steering linkage; and misalignment between the front and rear wheels. Note that the steering linkage was a well 177 9999999 9.9 99 99999 99999999 9599 9993 999999-9999 9993 399 999 9999: 9999999599 999 99 999999999 99999>99z 999 99 9999 99.9 999999 .53 20_h0m:_ol> 4m>¢ .meod I 5&8 wisp. >SUO dwwzi 432 202.5323 1 v 0. 5.2089. (W9) NOILOBHIO‘X 13AVUJ. UBLSBAUVH 178 designed system and little improvement would be obtained by redesigning the steering linkage. Thus the model was modified so that it would show a slower response for the harvester's motion in the Y-direction. The simulation model was modified such that the model used a steering angle which was about 50 percent less than the actual harvester steering angle. For the modification the first order time constant was not used because it only decreased the steering angle by 10 percent. The actual harvester turns the wheels 0.7 degree for each 2 cm of tree position error computed. The model before it was modified was programmed to turn the wheels 0.7 degree for each 2 cm of error computed. To modify the model this value of 0.7 was changed to 0.4. This value was selected because the simulation showed a harvester motion response similar to the actual harvester response for the row with a step change. Figure 9.l7 shows the results from the modified simulation model for the row with a step change. From this figure it can be seen that the model predicts that the front of the harvester is 2.5 cm from the row centerline as the front of the harvester moves past the step change in the row. The rear is 4 cm from the row centerline as it moves past the step change in the row. For the actual harvester performance for the step row (Figure 9.15) the front was within l cm and the rear was within 3 cm as these points on the harvester moved past the step in the row. The modified model was also used to simulate the harvester's travel over a curved row with radius of curvature of 121.9 cm (400 ft.) and the simulation results are shown in Figure 9.18. The simulation model predicted that the front of the harvester is within 9 cm as it passes by the tree and the actual harvester was within 8.7 cm as shown in Figure 9.13. The simulation model predicted the rear of the harvester is within 4 cm as 179 99999 99999999 9999 99 59 N 9999 999 99.9 99993 999 9995 99 9999999: 9999: 999 999999 9999 9993 399 999 99999 9999999999 999 99 999999999 99999>999 999 99 9999 99.9 999999 9.5. 29599.9... $3.99 9999922: P F ’ F Idi d 1 q ..§ h L F +11 J cl J J nesulwl. mu "- Il‘l""""‘ fl D¢<3C0u Imp. in. 13AVUJ. UBLSBAHVH (“”1 NOILOBUIO‘ A 180 9999a 99999999 999k $9 59 N :99m 999 99.9 mp9993 9:9 99:» 9» 99wmwnoz 99992 :99; 39¢ 99>9=9 999 _9992 cowpmpzewm 9:9 99 999999999 999m9>9m= 9:9 99 :9»; 99.9 993999 .50. 20:03:09. 4m><¢h thmw>¢<1 D¢<3¢Ou @— WSAVUL HSLSBABVH (W9) NOILOBHIO" 181 it passes by the tree and the actual harvester was within 3.2 cm as shown in Figure 9.8. Thus, these results indicate the model reasonably predicts the motion of the harvester when the model is modified to decrease the magnitude of the steering angle. 9.5.1 Reliability Problems During Test During the two successive tests with the step-change row, sonar unit 1 read a value of 109 cm for the first tree stand in the row, and this value was erroneous. The sonar value of 109 was observed on the video monitor and actual sonar value may have been larger than 109 as explained in Chapter 8. This tree was observed to be leaning slightly in the positive Y direction (about 3°). The path of the harvester for these two tests is shown in Figure 9.19 and 9.20. The test was repeated after the tree stand was straightened and the erroneous sonar value was not obtained. These results show that the front of the harvester deviated from the row centerline by 15.5 cm. Another test was run with tree stand leaning 120 in the positive Y direction and both sonar 1 and sonar 2 read a value of 109 cm when these sonar units moved past this leaning tree stand. Tests were done in the laboratory to investigate the cause of the sonar units making a measurement with a large error. A test was done with a tree stand and a large apple tree limb. The limb diameter varied between 3 cm and 7 cm and the limb length was 70 cm. For this test, an oscillosc0pe was connected to the PROCESSED ECHO signal on the Polaroid ultrasonic circuit (Figure 6.1) which was a component of a sonar circuit (Figure 6.5). The amplitude of this signal was observed on the oscilloscope as the tree stand was tilted from the vertical toward the sonar transducer which was transmitting sound pulses toward the tree stand. The tree stand was 78 cm from the sonar 1132 99.9»9 H 999592 um9huu9mcmguuqmum saw: 399 999 99999>991 9:9 99 9999 99.9 999mwm .o_- .:0. zo—wuwm—nnx z— uuzc~m~a 99.9»9 99.9mm 99.939 99.9». 99.9». 99.9». 99.9m— 99.mm 99.m9 99.»~ 99.9 99.9». Dm<3¢0u m991 9:9 99 £999 om.m 999999 .o—n “to. zo_huu¢~onx z. muchm_a 99.9mm 99.999 99.999 99.9». 99.9». 99.9M9 99.99. 99.m9 99.99 99.99 99.9 99.99- 99.99 oo-os-' I Ti oo-ov- 00'06' Y OO‘OZ' (H3) NOILDHEIU-A NI BUNULSIO 529:9 m THE NUHIER OF TREES OEINO 00:0 1N THE SIHULATION. TREELO -> AN ARRAY cONTAxNzNO x EACH TREE THE AND V COORDINATES OP SNR -> AN ARRAY CONTAINING THE D: ISTANCE TO EACH SONA UNIT non THE POINT CENTERED BETHEEN THEOENONTENNEELS (CH) AND EggsspghglggTuszu THEDLINE SEEHEgENIER POINTT T0 VESTER (RADIANS. L "E OF THE - ANG -> HALF THE CONE ANGLE OF THE OONAN UNITS (RADIANS) TINCR.) TIHE INCREHENT FOR ITERATION 0F HARVESTER HOTION SIHULATIDN. NLOOP I) NUHDER OF TIHES THR HA . HHEN COHPLETED. U IN LOOP CAUSES PROGRAH TO STOP 214 Figure 8.1. (continued) AWPINC I) WHEEL TURN RATE (DEG/SEC) INCALC I) SCALE FACTOR FOR ERROR USED TO OBTAIN DWP: INCALC I IIK K I PROPORTIONAL GAIN AHP I) INITIAL ACTUAL WHEEL POSTIION IPR I) COUNTER FOR PRINTING IN VALUE DUHP HODE (GIO) : IE. IT DET- ERHINES THE AHOUNT OF DATA PRINTED. EALWNC I) ALLOWED ERROR IN SONAR READING (CH) BEFORE TURNING WHEELS VI) VELOCITY (CH/SEC) ICOUNT I) COUNTER FOR DRAWING HARVESTER CENTER LINE OR “EES IN GRAPHICS HODE: IE. DETERHI NES THE NUHDER OF ITEHS DRAWN. ICK I) THE NUHDER OF TIHE STEPS BETWEEN SONAR READINGS G I) HODE DETERHINING VARIABLE. G I 1 FIRST GRAPH HICS HODE. DRAHS TREES WITH RESPECT TO THE HAR ESTR RER INTERIOR. G I 2 SECOND GRAPHICS MODE. DRAWS HARVESTER CENTER LINE WITH RESPECT TO TREES. G I 0 PRINTS VALUES OF SELECTED VARIABLES EVERY IPR TIHES THRU THE HAIN LOOP. SCAL I) SIZE FOR SYHDOLS REPRESENTING TREES WITH RESPECT TO HARVESTER TI I) TIHE DELAY CONSTANT FOR FIRST ORDER DELAY IN DISCRETE FORH LENGTH I) LENGTH OF THE FIELD IN SECOND GRAPHICS HODE (CH)‘ SIIE I) SCALING FACTOR FOR TREES IN GRAPHICS HODE I WID I) WIDTH OF THE FIELD IN SECOND GRAPHICS HODE (CH) C 91.90.! C’I‘ROOO9.0.9."...O“...{CGIGOCOGOOQOGOQORQGQ §§OWCGSQ“C“WO"IIOGIIO C O'COQOC‘Pflii‘ROOO. DOS... D“... .04... i“ QMQOOQO‘I‘DO. 9.. Oil. GOGGGG§QOOOGQOQOCRQD 0000000000000000OGOOOOOOOOGOOOOOOOOOO ANG I ONE HALF THE ANGLE OF VISION FOR SONAR UN ITS (RADIANS) AWP I ACTUAL WHEEL POSITION IN 1. DEGREE INCREHE NTS) AHPINC I THE RATE OF CHANGE FOR ACTUAL UWHEEL POSITION (DEG/SEC) COUNT I LOOP COUNTER FOR GRAPHICS OU TP DELT I HARVESTER ANGLE OFF OF TE HORIZONTAL REFERENCE DELTD I DERIVATIVE 2F HARVSS;ER ANGLE FUNCTIW DIS I DISTANCE FROH FIRST TREE WITHIN HARVESTER TO CENTER POINT DIST I DISTANCE IFROH THE TREE. WITHIN A SONAR FIELD OF VISION. TO THE SONAR DISTIN I DISTANCE FROH THE FIRST TREE WITHIN HARVESTER TO THE NT OF THE MRVEST DISTIE I DISTANCE FROH SECOND TREE WITHIN HARVESTER TO FRONT OF HARVESTER. DIS? I DISTANCE FROH SECOND TREE WITHIN HARVESTER TO CENTER POINT DWP I DESIRED WHEEL POSITION (IN 1.4 DEGREE INCREHEN TS) TEA LHNC I ALLOWED ERROR WITH RESPECT TO IDEAL HARVESTER POSITION ERRDZDB HARVESTOtnPOSTRDN EIFOR GI IO THE PROGRAH DUHPS VARIABLE OF OUTP WHE E O - VALEES. IF G I 2 THE PROGRAH DRAWS1 POSITION OF HARVESTEaEWITH vwvvvvovvvvvvvvvvvvwvvvvv9vvvvvvvvwvwvvvvvvvvvvvvvvvvvvvvvv ~ ' Hr-uufluuo-InvnunuuuuuuuuuuuuuuuuaunnuuuuuuuHuuuuuunuuuuuuuflfluu “I5323883882882836338:003888888288»SSBSRDNBBBBSESES2802838388283 c c c c c c c c C C C c c C c c C c 1 1 8 I 1 c RESPECT TO A ROW OF TREES. IF 0- THE ROGRAH DRA . P TI OF THE TREES WITH RESPECT TO THE HARV I 5; 8 ESACTSNL EL P00 N IN 4000005 INCREHENTS WHERE 32 10 “W 8 m THENRDRBE’R‘DOgaTIHE STUEPS~SEIT5EEN some ammo . 020; C COUNT - INCREMENT FOR DRAWING HARVESTER TREE POSITIONS 109) C I R - ERROR PARAHETER F ERK n70) c IERR - ERROR P A ER FOR OPENCC 171) C INC - AHOUNT OF CHANGE FOR DESIRE ED WHEEL POSITION .721 c INCALC - SCAL' FACTOR FOR ERROR TO ODTAIN THE CHANGE IN DWP A73) 0 IE. I CALC - I/K. K - PROPORTIONAL CAIN FACTOR 713;; 2 III - 02312110;IIIIUIMDITIITIOII 3352* I 933; E ITLAP - £53§RD§E2§IBEFORE WHEELS RETURN TO CENTER AFTER LAST $93; 8 ITRE- THTNEEEENESENTETEDEINC VIEWED 0‘11 SONAnw I I00) g JINX - CguangsgAEAHETEN FOR DRAWING TREE POSITIONS WITH RESPECT 435* ° W".- 05.0..” was 0.2139: .TIIDWR We I I32; 5 LENGTH - HORI TAL LENGTH OF FI an VISIBLE WITHIN GRAPHICS FRAHE 10:) c LINEDS - TOSCENTERFLI FIRST TREE WITHIN THEM mvesTEN . 1 IE2) g LINED: - TOBCENTERFLI SECOND TREE WITHIN THE HARVESTER I33; 8 NL. NUHDEN OERDAEFTSEEEICLM ECUATIONO TO BE OOLVED av DVENN L OOP - NUHS T NUHBER OF TREESU0 BE muszn IN THE SIHULAT ION. A30; 8 NHREER ' DIRENDION OF THE MATRIX W EIACTLvm SPECIFIED IN THE .931 c DIHENSION STATEHENT 0:81 8 NS . EARDNTCDDERRETONRED FROH ONCHCC . "391 8 335318 " §0§§1¥35u”o%"033532n°¥§§n}"mw 7“ "‘“m” . 93:1 8 SIGF - PRONTSWSEEIREROLER(RADIENSTIVE To v: 4) - 200) C SIZE - SCALING FACTOR FOR TREES IN GRAPHICS HODE 1 201) c ONNID.3) - LOCATION OF SONAR 20;) g SONAR - TRUNCATEDflDISTANCE FROH THE TREE. WITHIN A SONAR FIELD OF 215 Figure 8.]. (continued) § §§§§§ “nun—puuuwo vvvvvvvvvvvvvvvvvvvvvvvvvvvvvuvvvvvvu OOVOUOUNO‘OO NUMUMDNIJN”MUMNNMMNMMUNMDN I)” 3¥8§888833333£8833 ‘n: u - 3% R) A? -8 II I vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv 000000000000 88888””83””” oovoa280~3 E ”SE 0‘ U EEEEI All)” VINO 0‘00 §§I§§§§§§§§§§§§ Maps ‘1 3 0000000 §§§§§33¥§§3§§§ VISION: TO SONAR UNIT THETA I ANGLE BETWEEN CENTER LINE OF HARVESTER AND LINE FROH FIRST TREE TO CENTER POINT THETE I ANGLE BETWEEN CEN TEF[ LINE OF HARVESTER AND LINE FRON SECOND TREE TO CENTER POINT TINCR I TIME INCREMENT FOR ITERATION OF HARVESTER NOTION SIHULATION TOL I TOLERANCE FOR ERROR NTROL IN DV R R I NUHB R OF THE TREE WITHIN HARVESTER CLOSEST TO THE FRONT TREELO I LOCAT ON OF TREE IN FIELD FRAHE OF REFERENCE TR? I LAST TREE SEEN BY S R UNITS . TWHICH I LAST AR UNIT T HAVE A TREE WITHIN ITS FIELD OF VISION TI I TIME AY CONSTANT OR FIRST ORDER DELAY IN DISCRETE FORH V I FORWARD VELOCITY (CM/SEC) VI I F ON WHEEL VELOCITY (CM/SEC) W(3o9) I WORKSPACE MATRIX DVERK WHICH I IDENTIFIES THE UNIT PRESENTLY READING TREE POSITION WID I VERTICAL WIDTH OF VISIBLE IN GRAPHICS FRAHE WPDEG I WHEEL POSITION IN DEGREES X I INDEPENDENT VARIABLE FOR DV ER K XDATA7 I X-CDORDINATE OF FRONT POINT TO BE PLOTTED XDATAS I X-COORDINATE OF REAR POINT TO BE PLOT TED XDIF I BIT::NE HARV gngEENX X-COORDINATES OF HARVESTER AND FIRST TREE XDS I BITHINENCR VESTEREN X-COORDINATES OF HARVESTER AND SECOND TREE XEND I THE EN D POINT OF EACH TINE INCREHENT XMAX I MAXIMUM X-VALUE IN GRAPHICS FRW XMIN I MINIMUM X-VALUE IN GRAPHICS FRA XI I FRONT X-COORDINATE SAVED AND PLOTTED FRO" FOR EACH TIME INCREMENT X13 I WHEEL BASE (CH XI7 I DISTANCE BETWEEN FRONT AXLE AND FRONT POINT BEING TRACED (CM) X2 I REAR X-COORDINATE SAVED AND PLOTTED FROMK FOR EACH TIME INCREMENT Y(3) I MATRIX FOR DEPENDENT VARIABLES IN DVERK YDATA7 I Y-CDORDINATE OF FRONT POINT TO BE PLOTTED YDATAB I Y-CDORDINATE OF REAR POINT TO BE PLOTTED YDIF I DIFEERENCINBEIREENW Y-COORDINATES OF HARVESTER AND FIRST YD2 I DIFFERENCE BETWEEN Y-COORDINATES OF HARVESTER AND SECOND TREE WITHIN HARVESTER YIDEAL I SONAR VALUE OF CENTERED TREE (I78 CH) YHAX I HAXIHUH Y-VALUE IN GRAPHICS FR AHE YMIN I HINIMUH Y-VALUE IN GRAPHICSF Y! I FRONT Y-COORDINATE SAVED AND PLOTTED FROH FOR EACH TIHE INCREMENT Y12 I DISTANCE BETWEEN MONT WHEELS (CH Y2 I REAR Y-COORDINATE SAVED AND PLOTTED FROH FOR EACH TIHE INCREMENT CCCCCCCCCCCCCCCACE...9C0.Coo.9909.CCCCCCCCCCCCCCCCCCOCCCC CCCCCCCCCCSCCCCCC.94...CCCCCCCACCCCACCCCCCCCCCCCCCCCCCCCC: FCNISUBROUTINE CONTAINING DIFFERENTIAL EQUATIONS TO BE SOLVED C IS T B ING USED IN THIS ASE YPRIHE(N)IDIFFERENTIAL EQUATION TO BE SOLVED CEN...ICCCCCCCCCCCCCCCQCQCCCCCSAC...9999909...CCCCCCCCCCCCCCCCCCC CCCCCCaCNCSCCCCCCCCCCCCCCCCCCCCQCCCCCCCCCCCCNCCCCCCCCCCCCCSCQCCCE cCNS...CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCNCCCCCCCCECCCCCCCCCS COQOiOCOOOOOOQOOOODEOINNINO OF PROGRAH GDISPOCCCCCCCCCCCCCCCCCC cCSCCCCCCCECCCCCCOCCR90.009...CCCCCSCCCCCCCCCCCCCCCCQC9.00.9990 00000000000000000000000000000000000000 INTEGER IERIINDIKININuOSAHPLTOD~c0 COUNT SONAR ITLAPTNLOOPII PR.WHICH.NTR EESTICOUNTICSoPOINT 1.!AWP.I X. .ITR.ITREEoICKokOUNT.IERR.O.TR2.JIN REAL C(24).TOL.H( .9).X.XEND.Y(3).SIOF.V. T.DELTD.YIDEAL.LAB(6).AWP.DWP.ERROR.AWPINC.TINCR.EALWNC. XVI.X17.YI2.X13.XHIN.XHAX.YHIN.YM X.X1.Y1.Y2.XDIF.YDIF x. (2.20).SNR(3.3).ANO.DIBT.DIS.T A.L1NED8.SIZE zcxoa.Y02.0132.THET2.LINE02.DISTINonlsTlaoLENOTH.INCALC XTERNAL FCN.8NRDBT.TREE _gx CRNAL TESTI? RE.NéDLDCK/xsmollfl'.DELLDCLTDAII.Y12.117.X13 L w 2o GINOERT SYSCOM)KEYS.F CALL SRCHG$(KCREAD0'PAULDF' boSrNoNI) C EQQQOOQOOINITIAL CONDITIONS FOR DIFFERENTIAL ECUATIONSOOOCQOCQ OIOIIIIOINITIAL CONDITIONS FOR DVERK SUBROUTINEOIIOOOOI NANDW RUST BE SET TO THE NUMBER OF (DIFFERENTIAL EQUATIONS BEING SOLVED) 33“:§§1"“£ IIIISBCI5ISS oo§o§§§”° " 000 A 216 Figure B.1. (continued) 304) TOL-0.01 305) C 306) C 38;; 2.9.009O91N1T1AL CONDXTXONB 99994900 309) C 310) [AMP-32 311) COUNT-O 312) C2-0 313) XDIF-O 314) YDIF-O 315) THPDEo-0.0 316) P1-3.1415927 317) ERROR-0.0 318) SAMPLT-O 319) DELTAT-O 01 320) PT-O. 321) ”TO-0.0 322) DEF-32. 323) PEP-0.0 324) HPDEOOO 0 325) R2-.04 326) YIDEAL-78.0 327) TR-1 328) 1TREE-1 329) 1TR-1 330) DIS-78. 331) LINED8-78.0 332) TR2-1 333) L1NED2-0 0 334) 0152-0. 335) XD2-0 0 336) YD2-O 0 337) INx-1 338) XDATA7-0 0 339) YDATA7-0 0 340) ICH-O 341) I78 342) MOUNT-0 3:2: 2 345; 8%”. READ VALUES Fm BMDBT CUISTANTS «no 347) C 348) DO 1126 101.20 349) TREELO11.1)--1 350) TREELO(2.1)-O 351) 1126 CONT! ‘ , 352) C READ DATA PRO" 'PAULDF' 353) READ(9.0) NTREES ' 354) D0 1127 1-1.NTREE8 355) READ (9.0) TREELO(1.1).TREELO(2.1) 356) 1127 CONTINUE 357) DO 1128 1-1.5 358) READ(9.0) 8NR(1.1).8NR(1.2) 359) 1128 CONTINUE 360) READ (9.9) AND 325: 2 363; 80499.. READ IN CONBTANTS FOR PROORAN CONTROL 000.. 365) C 366) READ (9. O) TINCR. NLOCP. AUP1NC. INCALC. Aw. IPR. EALWC. V 367) ICOUNT.ICK.0.8CAL)T1.LENOTH.SIZE.H1D.POINT 368) PRES 0-0. 369) YH!N-(-1OHID) 370) YflAx-HID 371) J- 372) v1-v 373) XHA!-LEN8TH 333; 8 37¢) c969.60.696.04409909999999.09.9609499.646.6.6.66609996669 377) C SET UP FRAHES FOR GRAPHICS 33:) C95409404666060966906669969.9969.06.666009966666096...66¢ ) 380) C 381) CODRAH LADEL FOR HAVEBTOR 81TH RESPECT TO TREE ROHOO 382) ) OOTO 1151 383) C 5) 384) (K!CLO8.0.0.5.N1.N2) 385) 120) 386) (‘0 DATAO'.7.IERR) 387) (-450.0.LENOTH.(-1iHID).HID) 388) 100.0a(HID/12. ).5) 389) 300 ) 390) 311 ) 391) 312 ) 392) 303 F ) 393) 0.2.0.1) 394) 1.LA8)XH1N 395) (LAB.4.0.8) 396) 0.1.(0.0-(UID-(HlD/50.0)))) 397) 2.LA8)YH1N 398) (LA8.4.0.8) 399) (LENOTH—200).O0.0) 400) 0.LA3)XNAX 401) CALL ANPLOT¢LAD.5.0.8) 402) ALL HOVEA¢0 1.(0.0+(HID-(HIDI25.0)))) 403) ENC (7.303.LAD)YHAX 404) CALL ANPLOT¢LA8.4.0.8) 217 Figure 8.]. (continued) 405) CALL HOVEA(0.0.0.0) 406) C 407) C 408) C04... DRAN TREES IN FIELD4044O 409) DO 1135 J-1.NTREE8 410) CALL TRESL1TREELO(I.J).TREELO(2.J).(LENOTH+500).(2OHID).SIZE) 411) 1135 CONTINUE 412) 1151 CONTINUE 2:23 8 :12; 5949*. DRAH ALLOHABLE TREE ZONE 999*****¢* 417) C 418) IF(O.NE.2) OOTO 1122 419) CALL SRCHflO¢K|CLOBo010o5oN1aN2) 420) CALL IN TT¢120) 421) CKL L! .l‘ 9‘ ’0 DAT.’06) IERR, 422) CALL DMZ DO(-4U9.0:0.0a-155.5.155.5) 423) CALL FIRAmE(10.0. . 0. 5 424) CALL NOVEA(-409. .22.86) 425) CALL DRAHA(0.0.22.86) 426) CALL HDVEA(0.0.-22.86) 427) CALL DRAHA(-409.0.-22.86) 428) CALL HOVEA(0.0.0. ) 431) CALL 8NRDBT ARRA Y2 BY 20 THE UP TO 19 TREES FROH T OF -I. THE SECOND RON SNRI> 5 BY 2 ARRAY. C POINT OF THE HARVESTOR O IN THE FIRST COL UMN. AND T THE HARVESTOR NND THE LINE IN THE SECOND COLUNN. PARAHETERS RST RON CONTAINS THE X DISTANCES OF UP TO 19 TREES N. FOLLOHED DY ASTOPPER NS THE Y DISTANCES FROH THE ORIGIN. TAINS THE DISTANCE FROM THE CA T THE 5 SONAR UNITS HE ANGLE SET HEEN THE CENTERLINE” FROM THE CENTER POINT TO THE SONAR UNIT 2 ni $2 3 14 % OOOOOOOOOOOOOOOOOOO 220 Figure 8.1. (continued) 783; g XDEAST I) X COORDINATE OF THE HARVESTOR IN THE FIELD. ;§?; 3 YBEAST I) Y COORDINATE OF THE HARVESTOR IN THE FIELD. 712) C DELTA I) THE ANGLE THE CENTER LINE OF THEM WVESTOR MAKES HITH ;:2; g X-AXIS IN THE NEGITIVEY YDIRECTION. (RADIANS) 715) c66.66666...66.66666666.6666666.I66666666666666.6666...66666666666666. 71¢) c6666666666666666.666.666.666..66666666666666.6666666666666666.6666666 3:3; S¥Et°¥$ézé $::22*;32§e$*.;2§3$*.?§#zA+:"ICH'gg:r:.s~°-~EXTRE’ 719) REAL DELTA.XSNR.YSM EELO' 720) INTEGER SNRPO.TEST.NEXTRE.HHICH 721) COHHON/BLOCX/XENO.SIOF.DELT.DELTO.V1.Y12.X17.XI3 722) 2.TREELO. SNR 723) C SNRP O I POINTER FOR HHICH SONAR UNIT IS BEING TESTED ' 724) C TEST I IF TEST I0 TR REE IS IN FRONT OF SONAR FIELD OF VISION 725) C IF TEST I 1 TREE IS IN SONAR VIEH 726) C IF EST I -1 TREE IS PAST SONARS FIELD OF VISION 727) C TEST IS RETURNED FROM SUB. 728) C NEXTRE I POINTER TO NEXT TREE TO BE OR BEING TESTED FOR gggg DIMENSION TREELO(2.20).SNR(5. 2) 731) XTELE.ITREELO(1.NEXTRE) 732) YIRE .ITREELO(2.NEXTRE) 733) SNRPOII 734) 100 IF (XTREE .LT. 0) GOTO 99 735) XSNR'XBEAST+SNR(SNRPO:IIGCOS( SNR(SNRPO.2)-DELTA) 736) YSNRIYBEASTISNR(SNRPO.1)6SIN( SNR(SNRPO.2)-DELTA A) 737) CALL AFIELD(XTREE.YTREE.XSNR.YSNR.DELTA.TEST.M 738) IF (TEST .E0. 0) GO TO 99 739) IF (TEST .E0. 1) GO TO 200 740) SNRPOISNRPO+I 741) IF (SNRPO .LT. 6) GO TO 100 742) NEXTREINEXTRE+1 743) 515; ITREELO(1.NEXTRE) 744) YTR; ITREELO(2.NEXTRE) 745) SNRPOI1 746) GO TO 100 747) 200 CONTINUE 748) HHICHISNRPO gggg BISTAI( no: 826; g Y EI4 TO N I) POLYGON HITH N-I SIDES S28) c.66666666666.6.6666.66.66.666.6666666.6.6666666666666666. S29) SUBROUTINE TREE(X.Y.SC.TYPE) 830) INTEGER TYPE.SIDES.JI 831) REAL X.Y.SC.X1.Y1.RAD8.PI ' 832) PII3.141592654 - 838) SIDES I TY E-I 834) IF(TYPE.LE.O) GOTO 940 835) IF(TYPE.GE.4) GOTO 990 SS6) GOTO (960.970.980).TYPE 837) 960 CONTINUE 838) C DRAH AN X AT X.Y 839) CALL HOVEA(X.Y) 840) CALL HOVER((1.0ISC).(1.068C)) S41) CALL DRANR((-2.OISC).(I2.OISC)) 842) ”AL HOVERt0.0.(2.06SC)) 843) CAL. DRAHR((2.OISC).(I2.068C)) 844) GOT 940 S45) CON INUE 846) C DRAH A PLUS SIGN AT X.Y S47) CALL HOVEA(X.Y S48) CALL HOVER(SC.0.0) S49) CALL DRAHR((-2.ISC).O.0) S50) CALL HOVER((1.0ISC).(1.06SC)) S51) CALL DRAHR(0.0.(-2.6SC)) 852) GOTO 94 853) 980 CONTINUE S54) C A BOX AT X.Y 855) CALL VEA(X.Y) S56) CALL NOVERtSC.SC) 857) CALL DRAHR( (-2. ). ) 858) CALL DRAHR(0.0.(-2.6SC)) S59) CALL DRAHR((2.ISC).O O) 860) CALL DRAHR(0.0.( ) S61) GOTO O 862) 990 CONT NUE 863) C DRAH A POLYGON HITH TYPE-1 SIDES 864) ALL HOVEAt .Y) S65) CALL HOVER(SC.0.0) S66) 1000 1I1.SIDES S67) RADSIJ1IPII2./SIDES S68) X1ISC6COS¢RADS)+X ) YIISCISIN(RADS)+Y ) CALL DRAHA(X1.Y1) ) 1000 CONTINUE ; 940 gngRN ) 6111113 1111111111111111111111111111111111111111111111111111111 ) - '22 7272222 22 ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) §§§§§2§3§§33 222 Figure 8.). (continued) SX). (I2. OOSY)) mmmmemmu mmmmmmmm mmmmmmm ALL ALL ALL CALI... CAILJL RETURN ’F)’,’,’ mmmmmmmm.. ), 23 161 APPENDIX C ANALYSIS OF THE HARVESTER'S INERTIAL EFFECTS DUE TO GROUND SPEED APPENDIX C ANALYSIS OF THE HARVESTER'S INERTIAL EFFECTS DUE TO GROUND SPEED The following analysis was done to determine what harvester speed would cause the harvester to deviate from pure kinematic motion. Model of Vehicle Motion A model of vehicle motion must accurately predict the position of the vehicle for a specified input. The input to a dynamic model is wheel steering angle, 5, and the output is the vehicle position in X-Y coordinates. The variables X and Y vary with time. Ffigure C.1 shows a diagram of wheel steering angle and shows a simplified block diagram of the vehicle model. Dynamic Model A kinetic model of a vehicle, considers the inertial effects of the vehicle. The inertial effects on the vehicle result due to the mass of the vehicle and the mass moment of inertia of the vehicle. The equations of the kinetic model were used to describe the motion of the vehicle. Note, that when inertia effects are not significant then kinematic equations can be developed to describe the vehicle motion. Kinematic equations do not consider any effects from the system's inertia. The vehicle kinetic model has an input which is steering angle and the vehicle mass is moved by forces from the vehicle tires. 223 224 REACTION FORCES FROM t VEHICLE TIRES )(lt), \!(t) ism—,1 “mm-5 a—-———, VEHICLEPOSITION STEERING MODEL x,v COORDINATES ANGLE ) FORWARD STEERING ANGLE Figure C.l Diagram of Steering Angle and Block Diagram of Vehicle Model 225 Tire Force Characteristics According to Ellis (1969) tires generate a side force when the wheel is turned to an angle away from the direction of motion and this is shown in Figure C.2. The angle, , is called the side slip angle and the lateral side load is a function of the side slip angle. Figure C.3 shows a plot of the tire side load with respect to the side slip angle for a typical automobile tire. This data was obtained from Collins and Wong (1974). An accurate characterization of the side force is complex because it depends on tire size, construction, vertical load, air pressure speed, and tractive effort. Good results have been achieved by describing the side force as a coefficient, Cf, called the side force coefficient which is the initial slope of the plot of side force versus side slip angle as shown in Figure C.3. Note, that in Figure 0.3, the slope is initially constant for slip angles as large as 6 degrees. Yet, Cf is reasonably accurate for side slip angles up to 10 degrees. Research by Ellis and by Collins and Wong have developed other ‘techniques to model the tire side load for larger side slip angles and other special conditions. As reported by Ellis, an approximate value for Cf is 698 N/degree (157 lbf/degree) which is for a typical automobile tire with vertical load of 4,448 lbf (1000 lbf). This value was used as a conservative value for the tires on the harvester because specific data for the harvester's tires were not available. Kinetic Equations of Motion Kinetic equations of motion were developed for a vehicle model as shown irtl’igure 0.4. These dimensions shown in Figure 0.4 are for the apple harvester. 226 + DIRECTION OF MOTION 5= STEERING ANGLE 0 SIDE SLIP ANGLE Fy=LATERALFORCE FORCE DISTRIBUTION : FORCE ON THE GROUND PRODUCED BY THE TIRE TIRE PLAN VIEW Figure C.2 Diagram of Tire Side Slip Angle 227 N (lbf) 4448 (10001.. 4004 (900) .. 3559 (BOOT _ ’ 3114 (700) _ ’ TIRE SIDE 2559 (600).. / FORCE / / 1780 (400) _ / 1335 Isoo)_ / I TIRE BELTED BIAS 890 (2001_ / PRESSURE-1.6 m (24 psi) / VELOCITY-32 km/h (20 mph) 445 (1001- I VERTICAL LOAD-=4448 N (1000 1be 1'11111lllllr 02 4 6 81012 TIRE SIDE SLIP ANGLE, DEGREE Figure C.3 Typical Tire Characteristic of Side Force with Respect to Side Slip Angle. This Data was Reported by Collins and Wong (1974). 228 204 cm 204 cm (80.5 in) (80.5 cm) I 182 cm (71in) 365 cm A g (144 in) 182 cm CG (72 in) L122: :22: WEIGHT = 88,960 N (20,000 lbf) MODEL OF VEHICLE Figure 0.4 Dimension for Vehicle Ellis (1969), developed equations of motion that describe a vehicle's motion for small steering angles and for the case of constant forward velocity. These equations were written with respect to axes on the vehicle at the center-of—gravity (CG) of the vehicle and these axes were moving with the vehicle. These kinetic equations of motion by Ellis are: MIV+urI-Ic,+c,1IV/UI+Iac,-bc,II'/u1—c,5 (A.l) sz-(acf-bcf)(V/Ul+(ach+bzc,1('IUl-acffi (A.2) ‘11., (A.3) y-Vampwsinxp (L4) (4.5) x=Ums‘II-Vsin‘ll 229 Cf = effective side force coefficient for both front wheels (N/deg) or (lbf/deg) C,= effective side force coefficient for bother rear wheels (N/deg) or (lbf/deg) l2= mass moment of inertia of vehicle about the verical axis through the vehicle CG (cm-N-secz) or (in-lbf-secz) a = location of CG (cm) or (in) 1): location of CG (cm) or (in) M= vehicle mass N-sec2 or lbf-sec2 cm in \I= lateral velocity of CG (cm/sec) or (in/sec) LI: forward velocity of CG (cm/sec) or (in/sec) r= yaw rate about the CG (rad/sec) qr= angular position of CG (rad) x= position of CG (cm) or (in) Y= position of CG (cm) or (in) These equations developed ’by Ellis were derived in the following manner. First, a four wheel model was used to write two second order equations. These equations were (1) summation of forces in the Y-direction equals to the vehicle's mass times the acceleration in the Y-direction (the forces come from the tires) and (2) summation of moments about the vehicle's CG equals the vehicle's moment-of—inertia times the angular acceleration. The next step was to transform these equations to a form where the axes move with the CG of the vehicle. The last step, to simplify the equations, was to assume that the tire forces were symetrical and the vehicle equations could be written for a vehicle represented by one effective front wheel and one effective rear wheel as 230 shown in Figure C.5. Ellis stated that these Equations C.l to C.5 are often used to study vehicle dynamics. Effects of Tire Friction Force for SteadyStateTurning If there is sufficient friction force, then during a steady state turn a vehicle will move in a circle according to the Ackermann geometry assuming the vehicle is designed to satisfy the requirements of the Ackermann geometry. Figure C.6 shows a diagram of the Ackermann geometry. Durstine (1965) explained that the Ackermann geometry is often referred to as "ideal" steering geometry because with this configuration the front wheels will roll without tire "scruff” or slipping. The vehicle turns about the turning center. If the centri- fugal force or normal force, N, exceeds the tire friction force then the ‘vehicle will slip and not track as described by the Ackermann geometry. The vehicle speed, that will produce a normal force equal to the tire friction force, can be calculated. Assume that, , the coefficient of friction can be as low as 0.1 for slippery conditions. The mathmI friction force is calculated as _%_p = F Friction = 2,200 N (500 lbf) where N equals vehicle weight of 89,000 N. A free-body diagram is shown in Figure C.7 for equal 20 degrees. The necessary conditions for steady state turns are (1) the summation of the moments about the CG must equal zero and (2) the summation of forces in the X and Y direction must equal zero. A normal force is calculated to satisfy these conditions and is shown in Figures C.7. From the normal force, N, the 231 9 - .BA_° ‘r‘ \ "fir ANGULAR VELOCITY ‘sEC CG \\ \\ 11 F ,,~ ‘ B-STEERINO . I ‘\ ANGLE \b\¢; ’\O\ ~\\\ \ \ a lTIRESIDE DIRECTIONOF TIRE sIDE F FORCE " FORCE MOTION 0, = SIDE SLIP ANGLE PERPENDICULAR TO DIRECTION OF MOTION F, - 10,- c,I BODY CENTERED AXES MODEL FOR VEHICLE EQUATION OF MOTION Figure C.5 Body Centered Axes Model for Vehicle Equation of Motion 232 I FORWARD BAfiflC ACKERMANN GEOMETRY ACKERMANN / TURNING CENTERING ; .. - Figure C.6 Basic Ackermann Steering Geometry 233 409 cm (161inl A '2086 N f 365 cm 469 lbf R'2224 n (500 (bf) (144 in) CG A A=2086 N (469 lbf) B= 760N (171Ibe E3 §§ A-2086N R8 5 - 20 deg. N 469 lbf A N' NORMAL FORCE N'CENTRIFIGAL FORCE N88496 N (1910 lbf) Figure C.7 Free-Body Diagram of Forces for Steady State Turn with Allowable Tire Friction Force Equal to 2,200 N (500 lbf) 234 speed that will cause the vehicle to slip can be calculated as follows: v = 89,000 N (20,000 lbf) M=WI9 g = 980.6 cm/s2 N = M V2/R R = 1,120 cm (442 in) V NW The calculated speed for steering angle equal to 20 degrees was 32 cm/s (7.25 mph). This same calculation was done for the case of vehicle weight equal to 22,200 N (5,000 lbf) and the same value for the vehicle speed was calculated. Dynamic Considerations The following values of the vehicle parameters were used in this study: cf = cr = -80,000 N/rad (~18 000 lbf/rad) a = b = 204.5 cm (80.5 in) IZ = i—MLZ N-sZ/cm (lbf-szlin) L = a+b = 408.9 cm (161.0 in) for w = 89,000 N (20,000 lbf) M = w_= 90.7 N—sZ/cm (51.7 lbf -32/in) 9 for w = 22,200 N (5000 lbf) M = 3,: 22.7 N-szlcm (12.9 lbf -sz/in) (Q When the inertial effects on a system becomes significant, the mass is too large for the driving force (tire lateral force), and the system 235 will accelerate slower in the Y-direction when a turn iS'hfitiated. Therefore when the inertia or mass is large compared to the magnitude of the driving force, the vehicle response time is less than that described by kinematic motion. The response time for this study is defined as the time required by the system to reach a steady state condition beginning Iwith the time of application of a specified input. The input selected for analysis was the truncated-ramp. The input is the steering angle and a plot of this input which changed with time is shown in Figure C.8. The slope of the initial portion of the curve corresponds to the vehicle's maximum wheel turning rate. The steady state value of the steering angle is 10°. Inertial Effects For a given forward speed, increasing the weight of the vehicle to a large enough value will result in a significant increase of the system's response time. The dynamic or kinetic equations of motion are used to evaluate the response time of the vehicle for an input of steering angle as shown in Figure C.8. The steering angle increases at 150/5 until it reaches 10°; then the steering angle is held at that value. When the vehicle reaches a steady state condition it will move along a circular path. Also, the vehicle yaw rate, r,vfill reach a constant value at the steady state condition. The equations of motion were integrated using a numerical method technique to determine the vehicle response to the specified input. The response was evaluated for a vehicle of weight 22,200 N (5,000 lbf) and 89,000 N (20,000 lbf) and a forward speed of S36 cm/s (12 mph). The response of the vehicle in terms of the yaw rate is shown in Figure C.8. IA copy of the computer program for integrating the equations of motion 236 0 201 TRUNCATED RAMP J 0.00.000.000.000000000000.0....OOOOOOOOOQOOOOOOOOOOOO.I. 010- ° I I a ‘ ° I IRADI 912‘ , : I . | SYSTEM INPUT 875mm ”3‘ I STEERING ANGLE ANGLE " . I INPUT 0-04 . I 002 . ' | I : 1.0 2.0 3.0 40 5.0 6.0 . TIME ISECI I STEADY STATE |.00000000000000000000000000to. Degoooooooooooooooo.000.. 0.20 d I - I , 0.101 . : (RAD/sec, «I l VEHICLE WEIGHT-22 200 N 0.12 4 . I (5 000 lbf) - I u-sas CM/SEC VII-22200 N (5000 lbf) 0.089 . I (211 111/ICC) (12 mph) . I YAW 0.04.. ' ' RATE 0.024 . : : 1.0 2.0 10 40 5.0 6.0 ' TIME (SEC) I STEADv STATE | ...00000.00.00.00000000000000.000.000.0000.coco .0 I . 0.20 - I - . : ° ' I (RAD/SEC) 0‘16 - . VEHICLE WEIGHT-89000 N 0-12- ' 120000 1in ”am... I mm 0.08 < (211 111/001:) (12 mph) YAW ‘ RATE 0.049 ' 0.024 . 1.0 20 3.0 40 5.0 0.0 TIME (SEC) Figure C.8 Plot of Steering Angle and Yaw Rate 237 is shown in Figure C.9. It can be seen from Figure C.8 that the vehicle of weight 22,200 N has a smaller response time than the 89,000 N vehicle. This analysis assumes that there is enough friction force such that the wheels do not slip due to the centrifigal force. The vehicle of weight 22,200 N is assumed to react like a system of small mass because the response time is small. Criteria for Vehicle Model Evaluation During the operation of a vehicle guidance control system there must be periodic measurements of vehicle position so that the vehicle can correct for errors in position. It is expected that the actual position of a farm vehicle could be measured periodically at intervals of time represented by 304 cm (120 in) of travel. Thus, the estimated allowed position error for a farm vehicle with a steering control system is 2.5 cm 1.0 (in), therefore the following criteria were selected for use to evaluate the inertia effects on vehicle motion. For the vehicle to have negligible inertial effects the vehicle with large inertia must have the same position coordinates as the vehicle with small inertia within a tolerance of 12.5 cm (11.0 in) when the vehicle had traveled 304 :10 cm1 (120 :4 in). The vehicle center of gravity is selected as the point used to specify vehicle position. This criterion was used to evaluate the effects on vehicle position due to the vehicle inertia. Evaluation of the Kinetic Model The dynamic model showed that the 89,000 N vehicle had a larger response time than the 22,200 N vehicle for the specified input. This larger response time will result in a different position with respect to time for these two vehicles with different weight and same forward 238 00006: INTEGER KIINDDIERIJ 00007: REAL XIY(5)IIIHICFICRIUITIBICCIDIEIFIG.XEND.C(24).TOL.H(5I9) 00009: EXTERNAL FCN 00011:C 00012: N=5 00013: NU=5 00014: X=0.0 00015: IND=1 00016: Y(1)=0.0 00017: Y(2)=0.0 00018: Y(3)=0.0 00019: Y(4)80.O 00020: Y(5)80.O 00021: J=1O 00022: TOL30.0001 00023: UT=100000.0 00024: I=(NT*161.*161.)/(32.2*12.*4.) 00025: M=UT*(1./(32.2*12.)) 00026: CF8-18000. 00027: ORB-18000.0 00028: U817.6 00029: PRINT 200 NTIU 00030 200 FORMAT (2F10.3) 00031: PIR3.1415627 00032: DO 50 K=1.680 00033: XENDBFLOAT(K)/100.O 00034: NP=XEND*(15.75*PI)/180. 00035: IF(NP.GT.O.17453) HP=O.17453 00036:C 00037: CALL DVERK (NIFCNIXIYIXENDITOLIINDICINHIN.IER) 00038: J-J-l 00039: IF(J.EG.O) GO TO 12 00040: GO TO 50 00041 12 F=Y(5) 00042: 08Y(4) 00043: A=Y(3) 00044: CC=G+(80.5*SIN(A) ) 00045: BCF+(80.5*COS(A) ) 00046: DBF-(BO.5*COS(A) ) 00047: E-C-(80.5*SIN(A) ) 00048: 03((F**2)+(G**2))**0.5 00050 100 FORMAT(11F10.3) 00051: J=1O 00052 50 CONTINUE 00053: STOP 00054: END 00055: SUBROUTINE FCN(N.XIYIYPRIHE) 00056: INTEGER N 00058: c0""ON/BLOCK/XENDIAOuPIIONICFOCRIUOV 00059: YPRIME(1)--(U*(Y(2)/H)) + ((1./H)*(CF+CR)*(Y(1)/U)) 00060: Z-((CF/H)*UP) 00061: YPRIHE(2)=(( (80.5**2)*CF)+((80.5**2)*CR))*(Y(2)/U)*(1.II) 00062: Z-(80.5*CF*NP*(1./I)) 30063: YPRIME(3)-Y(2) 30064: YPRIHE(4)=(Y(1)*COS(A))+(U*SIN(A)) 30065: YPRIHE(5)-(U§COS(A))-(Y(1)*SIN(A)) 30066: RETURN 300671’ END BOTTOM Figure C.9 Computer Program used to Compute the Vehicle Response Due to a Specified Steering Angle Input 239 velocity. The integration of the equations of motion was performed to determine the position of the vehicle CG with respect to time for forward speeds of 1.6, 3.2, 4.8, 6.4, 8.0 and 19.3 km/h (1,2,3,4,5 and 12 mph) for vehicle weight of 22,200 N and 89,000 N. Table C.l shows the Y-coordinate position of the CG when the x-coordinate was approx- imately 304 cm (120 in). The Y-coordinate position for the 22,200 N vehicle was used as the reference for vehicle position without inertial effects. 'The difference in the Y-coordinate values of these two vehicles was considered the error in position that will result when the effects of inertia are not considered for computing vehicle motion. This error is tabulated in Table C.l. Table C.1 Position Coordinates of CG from Dynamic Model for Vehicle of Weight 39,000 N and 22,200 N VEHICLE VEHICLE FORWARD 89 000 N 22 200 N VELOCITY COORDINATE COORDINATE COORDINATE ERROR U x Y1 Y2 I Y 2.. Y1) km/h' (mph) cm (in) cm (in) cm (in) cm (in) 19.3 12 321.795 126.691 7.579 2.984 14.150 5.571 6.57 2.58 3.0 5 311,337 122.790 28.042 11.040 31.651 12.461 3.60 1.42 6.4 4 301743 119.190 30.726 12.097 33.155 13.053 2.42 0.95 4.8 3 306.822 120.796 35.720 14.063 37.211 14.650 11.49 0.59 3.2 2 310.810 122.366 40.704 16.025 41.394 16.297 0.69 0.27 1.6 1 301.727 118.790 42.669 16.799 42.812 16.855 0.14 0.05 These data show that at a forward speed of 6.4 km/h (4.0 mph) the difference in position of the two vehicles is 2.45 cm (0.95 in) and this difference gets larger as the forward speed increases. 240 Conclusions It is calculated that for forward speeds less than 6.4 km/h (4 mph) the vehicle of weight 89,000 M (20,000 lbf) as described by the developed dynamic model can be modeled without inertial effects without causing excessive error in the calculated position. These results were based on motion equations developed for the apple harvester. The input of the motion equations was wheel steering angle. The wheel steering angle was set at zero and allowed to increase up to 20° at a rate of 15 deg/s. For ground speeds less than 6.4 km/h (4 mph) the results show that the vehicle can be described by a kinematic model. This conclusion is based on the assumption that there is enough friction so that the tires will not slip. For steady state turns with steering angle of 20 degrees and coefficient of friction equal to 0.1 it is calculated that the 89,000 M vehicle will slip at speeds greater than 11.6 km/h (7.25 mph). APPENDIX D COMPUTER PROGRAM FOR HARVESTER'S MICROPROCESSOR-BASED STEERING CONTROLLERS APPENDIX D COMPUTER PROGRAM FOR THE HARVESTER'S MICROPROCESSOR-BASED STEERING CONTROL SYSTEM Appendix 0 contains a list of the assembly language program (Figure 0.1) that was develOped for the 1802 micrOprocessor which was used in the steering control system. Also shown in Figure 0.1 is the object code that was used by the microprocessor. This listing shown in Figure 0.1 is the output from the Cross-Assembler (made by RCA) which is a FORTRAN proram on the Michigan State University's main computer (Cyber 750). The object code has a total of about 2,000 bytes. If the program is modified to delete the portions of the object code which pertain to displaying data on the video monitor, then the control system's object code has about 1,000 bytes. 242 243 Figure D.l. Assembly Language and Object Code Listing of the Program Used by the Steering Controller's Microprocessor PROGRAM NAME IS' ASHDELAY ' LIST. 100: FL LOC COSHAC CODE LNNO SOURCE LINE 1103 0000 1 oooopROGRAH 3 RCATREE NCHAHON/CLEHENS 120' 0000 2 INT'RO 130' 0000 3 ooR1 140a 0000 4 SP'R2 150' 0000 5 PC8R3 160' 0000 6 CALL3R4 170- 0000 7 RETN-RS 180' 0000 8 LINK-R6 190. 0000 9 DSUPR'R7 200- 0000 10 TEHP'R84044COUNTER ACCUH‘ 210' 0000 11 USONR-R9 ‘ ' 220. 0000 12 SNR'RA 230' 0000 13 FLAGR'RD 2403 0000 14 00Rc 250' 0000 15 NSONR'RD 2603 0000 16 BINR'RE 270' 0000 17 00RFO0POINTS TO LARGE ADDRESS FOR INPUT 28°. 0000 18 0000009LSO FOR PRINTING 290a 0000 19 000.8888888888888 300' 0000 20 TYPE‘X’8198’ 310I 0000 21 OSTRG-X’83F0’ 320' 0000 22 0.008888888888888 ' 33°. 0000 23 000088888888 340' 0000 24 4.04FORT ASSIGNMENT 350' 0000 25 0000PORT N03 SONAR 0 OP CODE 3608 0000 26 ooooGROUF .20-A 2 1 INF 486C 3708 0000 27 0000 20-8 1 INF 686E 380a 0000 ‘ 28 0000 10-A ‘ 4 ‘ INP 486C 390= 0000 29 0000 10-B 3 IMP 6.65 ‘00. 0000 30 0000 OB-A 5 INF 4'60 410' 0000 31 0000 08-8 RELAYS OUTPUT CODE 0 420- 0000 32 0000TURN RIGHT-OUT 6IDC X’01’ 430' 0000 33 ooooTURN LEFT'OUT 6IDC X’02’ 440' 0000 34 0005ET DSUP'32 UHICH IS CENTER POSITION 450a 0000 35 0000 DSUP-O IS FULL RIGHT TURN 460' 0000 36 0000 DSUPI64 13 FULL LEFT TURN 470' 0000 37 0000000000000000000000000000000000000000} 480* 0000 38 0000INPUT PORT 490= 0000 39 0008888888888888888888 500' 0000 40 000 REGISTER INITILIZATION 5103 1000 41 ORG X’1000’ 52 = 1000 71 42 DIS 530' 1001 00 43 DC 0 540a 1002 C4 44 NOP 550' F 1003 F800A3 45 INITILDI A00(START)3PLO PC 5603 F 1006 F80033 46 LDI A.1(START)IPHI PC 5702 F 1009 C00000 47 LDR ENTER2 580= F 100C F8008485 48 ENTER2: LDI A.l(CALLR13PHI CALLIFHI RETN 590= F 1010 F800A4 49 LDI A.0(CALLR)3?LO CALL ' 6008 F 1013 F800A5 50 LDI A.0(RETR)IPLO RETN 6108 F 1016 FSOOBS 51 LDI A.l(RETR16PHI RETN 620- F 1019 F800A2 52 LDI A00(TOPSTK)3FLO SP 630= F 101C F80082 53 LDI A.l(TOFSTK13PHI SF 6408 101F E2D3 54 SEX SPISEP PC 650' 1021 55 40.888888888888888888 6608 1021 56 000 DESCRIPTION! 670:3 1021 57 oooSTANDARD SEP CALL? A(SUBR.NAHE) 680' 1021 58 00.888888888888888888 6903 1021 59 o4oSTANDARD CALL 7003 1021 D3 60 EXITC: SEP PC 710- 1022 E2 *‘“T"’*““‘“"TGt‘CALthSEX'SP‘* Figure 7208 7308 7408 7508 7608 7708 7808 7908 8008 8108 8208 8308 8408 8508 8608 8708 8808 8908 9008 9108 9208 9308 9408 9508 9608 970= 9808 9908 10008 10108 10208 10308 10408 10508 10608 10708 10808 10908 11108 11208 11308 11408 11508 11608 11708 11808 11908 12008 12108 12208 12308 12408 12508 12608 12708 12808 12908 13008 13108 13208 13308 13408 13508 13608 D.l. (continued) 1!” flfiflfl fin 13708~ 8673 9673 9386 83A6 4683 46A3 C01021 D3 9683 86A3 E212 7286 FOA6 C01032 C4 C4 F80888 840000 C4C4C4 F83FAF8F E36108 6253 6600 F800A7 F8008? F82057 F800A9 F8008? FBOOAD FBOOBD ED ~ FBOOSD6O 5860 5860 5860 5860 SD FBOOAA F8008A FBOOAB F80088 E8 F80058 6OF80158 6OF80058 6058 6058 6058 - - c—uo 244 OLD LINKTSTXO GHI LINKiSTXD 6H1 POTPHI LINK GLO Pc; PLO LINK LDA LINK; PHI Pc LDA LINKIPLO Pc LBR EXITC ...xtxxxx ...STANDARD RETURN EXITR:9EP Pc RETR: GHI LINKTPHI Pc GLO LINK; PLO Pc SEX SPIINC SP Lnxas PHI LINK LOX; PLO LINK LBR EXITR NOP ....xxxxtxxxxxgxxxRthx .... START ....xttxxxtxxxxxtxttxxx START:NOP LOI 83PHI R8....FOR 1200 BAUD SEP CALL6.A(scRN> NOPTNOPTNOP LOI x'3P'TPLO RFTPHI RF ...txxxxx ...OUTPUT ZERO TO PORT 08-0 SEX PCTOUTIIOC 8 0UT23DCX’S3’ ... OUTPUT ZERO OUT 6600 0 L01 A.0(DSUP)8PLO LOI A.1TPHI LOI 32$STR DSUPR . ... SET UP PTR TO uson RNO NSON LOI A.0(USON)5PLO USONR, LOI A.l(USON)6PHI USONR- LOI A.0TPHI SNR ....INITIALIZE--READ ONLY sou UHEN FLAG(X)-1 DSUPR' DSUPR: 0000 0000FLAG(0)-0 0000FLA8(1).1 0000FLAG(2 THRU 5).0 LDI A00(FLAG)‘PLO FLAGR_' LDI A.l(FLAG)3PHI FLAGR. SEX FLAOR - LDI OTSTR FLAORoooO IRX‘LDI 1’STR FLAORoool. IRXCLDI O’STR FLAORooo2 IRXISTR FLAGR0003 a IRXfiSTR FLAOR...4 IRX'STR FLAGR...5 0000I"ITIALIZE" ’ Figure 0.]. (continued) 1380- 1097 1390- 1097 F800A8 1400- 109A P90438 1410- 1093 1420- 1093 04 1430- 1095 0404 1440- 1000 0404040404 1450- 1045 0404040404 1460- 1004 2898 1470- F 1040 3400 1480- F 104E 340000 1490- 1031 F8043e 1500- 1034 FBOOAB 1510- 1037 040404 1520- F 108A FBOOAD 1530- F 1033 FBOOBD 1540- 1000 1550- 1000 1560- F 1000 FBOOAB 1570- F 1003 P80033 1590- F 1006 340000 1590- 1009 1600- 1009 1610- 1009 1620- 1009 1630- 1009 1640- 1009 1650- 1009 1660- 1009 040404 1670- 1000 13 1680- 1003 13 1690- 100E 43 1700- F IOCF 020000 1710- 1032 FBOISA 1720- 1035 E3 1730- 1036 6101 1740- 1038 6301 1750- F 1034 340000 1760- 1033 04 1770- 1030 09 1780= IODF FF96 , 1790- F 10E1 3300 1800- 10E3 1810- 10E3 03F00153 1820- 10E7 04 1830- 10EO 1840- 1058 1850- 10E8 18608 10:8 1870- 10E8 13 1880= 1059 43 1990- F 1004 020000 1900- 10E3 F8025A 1910- 10FO E3 1920- 10F1 6101 1930- 10F3 6302 1940- F IOFS 340000 1950- 10F8 04 1960- 10P9 09 1970- 10P4 FF96 1930- F 10P0 3300 1990- IOFE 03P00153 2000- 1102 0404 2010- 1104 2020- 1104 2030---—1104w-~- ~ “ l 0' 245 128 129 130 131 132 133 134 135 136 137 . 138 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 191 192 - --~ ~— -—-193- '3‘ .‘ flaunt-- 0000D13PLAY DELAY LDI O’PLO R8 LDI 4CPHI R8 000888888888888888888888888888888 LOOP23NOP LOOP13NOP’NOP NOP‘NOP‘NOP’NOP’NOP NOP‘NOP‘NOP’NOP’NOP DEC R83OHI R8 8N2 BERT SEP CALL19A(DISPL) LDI 43PHI R8 LDI 03PLO R8 DERTSNOP‘NOP‘NOP LDI A00(NSON)3PLO NSONR LDI A01(NSON)‘PHI NSONR 0000SET PTR TO FLOR 0000USED TO TURN OFF SONAR(X) LDI A00(FLAG)1PLO FLAGR ' LD1 A01(FLAO)‘PHI FLAOR SEP CALL‘ !A(TURN) ' 000888READ SONAR1 1F READY 000888STORE VALUE IN VSONR 000888UALUE8X’FF’ NEANS SONAR NOT READY 0008888888 000SONAR 1 000888888 000088815 SONAR(1) LT 150 NOP‘NOP’NOP INC NSONR INC FLAGR0000FL61 LDA FLAGR0000PTR TO FLO2 L82 CONT2 LDI 18 STR SNR 000 STORE '1' SEX PC OUT 1130 x'01'. OUT 33DC X’01’ SEP CALL ’9A(READ) NOP LDN VSONR SHI 150 000 885-150 DPZ CONT2 000VSONR IS LT 150 LDN NSONR O ADI 1 ’ STR NSONR CONT2: NOP 00088888888 000SONAR 2 00088888888 000088815 USONAR(2) LT 150 INC NSONR LDA FLAOR0000PTR TO FLG3 LDZ CONT3 I LDI 2 1 STR SNR SEX PC OUT 1’DC X'01’ OUT 33DC X'02" SEP CALL 3 9 A(READ) NOP ' LDN VSONR SHI 150 DPZ CONT3 LDN NSONR C ADI 1 9 STR NSONR CONT33 NOP‘ NOP 00088888888 000SONAR 3 IN SN 000STORE ’2’ IN SN 246 Figure D.l. (continued) 2040‘ 1104 194 00015 USONR3 LT 150 2050‘ 1104 1D \ 195 INC NSONR 2060‘ 1105 48 196 LDA FLAGR000PTR TO FLA84 2070‘ F 1106 C20000 197 L82 CONT4 2080‘ 1109 F8035A 198 LDI 33 STR SNR 2090‘ 110C E3 199 SEX PC 2100‘ 1100 6101 200 OUT 138C X’01’ 2110‘ 110F 6304 201 OUT 310C X’04’ 2120‘ F 1111 D40000 202 SE? CALL‘ 9A(READ) 2130‘ 1114 C4 203 NO? 2140‘ 1115 09 204 LDN USONR 2150‘ 1116 FF96 205 SHI 150 2160‘ F 1118 C30000 206 LFDF CONT4 2170‘ 1118 0DFC015D 207 LDN NSONR 3 A011 3 STR NSONR 2180‘ 111F C4C4 208 CONT43 NOP‘ NOP 2190‘ 1121 209 00088888888 2210‘ 1121 211 00088888888 2220‘ 1121 212 00015 V50NR4 LT 150 2230‘ 1121 1D 213 INC NSONR 2240‘ 1122 48 214 LDA FLAGR0000PTR TO FLAGS 2250‘ F 1123 C20000 215 L82 CONT5 2260‘ 112 F8045A 216 LDI 4 3 STR SNR 2270‘ 1129 E3 217 SEX PC 2280‘ 112A 6101 218 OUT 110C X’OI’ 2290‘ 112C 6308 219 OUT 33DC X’08' 2300‘ F 112E D40000 220 SEP CALL 3 9A(READ) 2310‘ 1131 C4 221 NOP 2320‘ 1132 O9 222 LDN USONR 2330‘ 1133 FF96 223 SHI 150 2340‘ F 1135 3300 224 882 CONT5 2350‘ 1137 0DFC01SD 225 LDN NSONR 3 ADI 1 1 STR NSONR 2360‘ 1138 C4C4 226 CONTS: NOP‘ NOP 2370‘ 113D 227 00088888888 2380‘ 113D 228 00050NAR 5 2400‘ 113D 230 00015 VSONRS LT 150 2410‘ 113D 10 231 INC NSONR 2420‘ 113E OF 232 LDN FLAGR 2430‘ F 113F C20000 233 L82 CONTND 2440‘ 1142 F8055A 234 LDI 5 1 STR SNR 2450- 1145 83 235 sax PC ‘ . 2460‘ 1146 6101 236 OUT 130C X’01’ 2470‘ 1148 6310 237 OUT 33DC X’10’ 2480‘ F 114A D40000 238 SEP CALL 1 9A(READ) 2490‘ 114D C4 239 NO? 2500‘ 114E 09 240 LDN VSONR 2510‘ 114F FF96 241 5H1 150 2520‘ F 1151 3300 242 DPZ CONTND 2530‘ 1153 0DFC015D 243 LDN NSONR 3 ADI 1 3 STR NSONR 2540‘ F 1157 C00000 244 CONTND3LBR CHKNSO00008RANCH TO CHKNSO 2550‘ 115A 245 0000888888888888888888888888888888888 2560‘ 115A 246 000SUBROUTINE: READ 2570‘ 115A 247 0000888888888888888888888888888888888 2580‘ 115A 248 000READ SONAR "HEN READY 2590‘ 115A 249 000SNR‘RC 000 AND IT MUST CONTAIN 01-5 2600‘ 115A C4C4C4 250 READ: NOPINOP‘NOP 2610‘ 115D 0A 251 LDN SNR 2620‘ 115E FFO1 252 SHI 1 000 D‘D-1 2630‘ F 1160 C20000 253 L82 CALL1 2640‘ F 1163 C00000 254 LBR CON1 2650‘ 1166 255 CALLI: 2660‘ F 1166 D40000 256 SE? CALL ‘9 A(READ1) 2670‘ F 1169 C00000 257 LDR CONS 2680‘ 116C 258 CON1: 2690"= -116C 0A “ ' ‘ 259'LDN SNR Figure 2700‘ 2710‘ 2720‘ 2730‘ 2740‘ 2750= 2760‘ 2770‘ 2780‘ 2790‘ 2800‘ 2810= 2820‘ 2830‘ 2840‘ 2850‘ 2860‘ 2870‘ 2880‘ 2890‘ 2900‘ 2910‘ 2920‘ 2930‘ 2940‘ 2950‘ 2960‘ 2970‘ 2980‘ 2990‘ 3000‘ 3010= 3020‘ 3030‘ 3040‘ 3050‘ 3060‘ 3070‘ 3080‘ 3090‘ 3100‘ 3110‘ 3120‘ 3130‘ 3140‘ 3150‘ 3160‘ 3170‘ 3180‘ 3190‘ 3200‘ 3210‘ 3220‘ 3230‘ 3240‘ 3250‘ 3260‘ 3270‘ 3280‘ 3290‘ 3300‘ 3310‘ 3320‘ 3330‘ 3340‘ D.1. (continued) W'W’IW “‘“71W ..0 0.0 V U fl'fi‘Ifl 0.0 H 0 N ”Ifl'“ H H 0 u 11EB ‘3350‘*'”‘11EE ( \ FF02 C20000 COOOOO 040000 C00000 0A FFO3 C20000 C00000 040000 C00000 0A FF04 C20000 COOOOO 040000 COOOOO 0A FFOS C20000 C00000 040000 05 C4C4 E36120 E9 3500 COOOOO C4 C4 6ECOOOOO F8FF59 05 C4C4 E36120 E9 3400 COOOOO C4C4C4 6CC00000 F8FF59 05 C4C4C4 E36110 E9 3500 COOOOO C4C4C4 6EC00000 F8FF59 05 C4C4C4 E36110 -£9 ——~---~—~ 247 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 _--— 325~88x—v80NR-_~_ SHI 2 L82 CALL2 LDR CON2 CALL23 SEP CALL 3 LDR CONS CON23 LDN SNR SHI 3 L82 CALL3 LBR CON3 CALL33 SEP CALL 3 LBR CON5 CON33 LDN SNR 5H1 4 L82 CALL4 LDR CON4 CALL43 SEP CALL3 LDR CON5 CON43 LDN SNR 5H1 5 L82 CALL5 LFR CONS CALL53 SEP CALL3 CON53 SEP RETN 00008888888888888888888888888888888888888 0000SUBROUTINES TO READ A SET OF 5 SONAR 00008888888888888888888888888888888888888 0000SUBROUTINE TO READ SONAR1 0000 STORE VALUE INTO VSON 0000 SET NR=1 0000 CALL READ1 TO READ SONRI 0000PUT FF INTO VSONR IF NOT READY 00008888888888888888888.' 9A(READ2) 9A(READ3) 9A(READ4) 0A(REA05) READl 3 NOP3 NOP SEX PC3 OUT 1 3 DC X’20’ SEX USONR 000PTR TO VSON B2 BETSDl ’ LBR CCONTI GETSfll 3 NOP NOP ' INP63LBR 810000STORE USONR BY REGX' CCONT1 3 LDI'X’FF’ 3 STR VSONR 813SEP RETN .0088888888888888888888888888 READ2 3 NOP3 NOP ' SEX PC 3 OUT1 30C X'20" SEX USONR 000 PTR TO USON 81 GETSD2 LBR CCONT2 GETSD23 N083 NOP3NOP INF 43LBR B2 000STORE BY REGX - USON CCONT23 LDI X'FF’ 3 STR USONR B23SEP RETN 000888888888888888888888888888 READ3 3 NOP3 NOP3 NOP SEX PC3 OUT 13 DC X’10’: SEX USONR B2 8ETSD3 LBR CCONT3 GETSD3 3 NOP3 NOP3 NOP-- INF 63LBR B3 ' CCONT3 3 3 STR USONR L31 x'rr' B33$EP RETN = ...xxxxxxxxxxtxx1000-10:00::x READ4 : NOP3 NOP3 NOP sex 80: our 1130 x'10' .— -- - .4-.—..31_. 248 Figure D.1. (continued) 3360‘ F 11EF 3400 326 Bl GETSD4 3370= F 11F1 COOOOO 327 LBR CCONT4 3380‘ 11F4 C4C4C4 328 GETSD4 3 NOP3 NOF3 NOP 3390= F 11F7 6CC00000 329 IN? 43LBR B4 000 STORE BY REGX-VSON 3400‘ 11FB F8FF59 330 CCONT43 LDI X’FF’ 3 STR USONR 3410= 11FE D5 331 B43SEP RETN 3420= 11FF 332 0008888888888888888888888888888 3430‘ 11FF C4C4C4 333 READS 3 NOP3 NOP3 NOP 3440‘ 1202 E36108 334 SEX PC3 OUT 13 DC X’08" 3450‘ 1205 E9 335 SEX USONR ' 3460= F 1206 3400 336 Bl GETSDS 3470‘ F 1208 000000 337 L88 CCONT5 3480= 12GB C4C4C4 338 GETSD5 3 N083 NOP3 N08 3490‘ F 120E 6C3000 339 INP 438R 85 3500‘ 1211 F8FF59 340 CCONT5 3 LDI X’FF’3STR USONR 3510‘ 1214 05 341 BS3SEP RETN ' 3520‘ 1215 342 00008888END OF SUBROUTINE88888888 3530‘ 1215 343 0088888818 NSON(X)91-53‘2? 3540‘ 1215 344 000088888888888888888888888888888 3550‘ 1215 345 00008888CHKNSO3 3560‘ F 1215 FSOOAD 346 CHKNSO3LDI A00(NSON)3PLO NSONR00RESTORE D 3570‘ 1218 C4 347 NOP 3580‘ F 1219 FBOOBD 348 LDI A01(NSON)3PHI NSONR- 3590‘ 121C 1D 349 INC NSONR00TO POINT TO P080 1 NOT 0 3600= 121D F8015A 350 LDI 13STR SNR0000PUT UNIT 0 INTO SNR 3610‘ 1220 OD 351 LDN NSONR 00INTO D ‘ 3620‘ 1221 FF02 352 SHI 200D‘D-2 ' 3630‘ F 1223 C30000 353 LBDF A500JUHP IF D‘O 3640‘ 1226 ID 354 A13INC NSONR000POINT TO P05 2 3650‘ 1227 F8025A 355 LDI 23STR SNR 3660‘ 122A OD 356 LDN NSONR 3670‘ 1228 FF02 357 SHI 2 3680‘ F 122D C30000 358 LBDF A5 - 3690‘ 1230 1D 359 A23INC NSONR 3700‘ 1231 F8035A 360 LDI 33STR SNR 3710‘ 1234 OD 361 LDN NSONR 3720‘ 1235 FF02 362 SHI 2 3730‘ F 1237 C30000 363 LBDF A5 3740‘ 123A ID 364 A33INC NSONR 3750‘ 123B F8045A 365 LDI 43STR SNR 3760‘ 123E 0D - 366 LDN NSONR 3770‘ 123F FF02 ‘ 367 SHI 2 3780‘ F 1241 C30000 368 LBDF A5 L 3790‘ 1244 1D 369 A43INC NSONR 3800‘ 1245 F8055A 370 LDI 53STR SNR 3810‘ 1248 OD 371 LDN NSONR 3820‘ 1249 FF02 372 SHI 2 3830‘ F 124B C30000 373 LBDF A5 3840‘ 124E COlO9E 374 LBR LOOP100GO BACK IF NOT 3850‘ F 1251 FBOOAD 375 A53LDI A00(NSON)3PLO NSONR 3860‘ F 1254 FBOOBD 376 LDI A01(NSON)3PHI NSONR: 3870‘ 1257 F800 377 LDI 000RESET UNIT ACCUHULATORS 3880‘ 1259 ED 378 SEX NSONR 3890‘ 125A SD 379 STR NSONR 3900‘ 125B 60 380 IRX 3910‘ 125C SD 381 STR NSONR 3920‘ 125D 60 382 IRX 3930‘ 125E SD 383 STR NSONR 3940‘ 125F 60 384 IRX ' 3950‘ 1260 5D 385 STR NSONR 3960‘ 1261 60 386 IRX 3970‘ 1262 SD 387 STR NSDNR 3980‘ 1263 605D 388 IRX3STR NSONR 3990‘ F 1265 F800AD 389 LDI A00(NSON)3PL0 NSONR 4000‘ F 1268 FSOODD 390 LDI A01(NSON)3PHI NSONR 4010“"""1‘2'68 " """ ' "' §91 ‘0000ENI' OF “SETTING " _ 3 Figure D 1 (continued) 4020‘ 1268 4030‘ 1268 4040= 1268 4050‘ 1268 4060‘ 126B 4070‘ 1268 040404 4080‘ 126E 040404 4090‘ 1271 F800AD 4100‘ 1274 F8008D 4110‘ 1277 F8005D 4120‘ 127A 040404 4130‘ 127D 4140‘ 127D 040404 4150‘ 1280 5D 4160‘ 1281 4170‘ 1281 D4115A 4180‘ 1284 040404 4190‘ 1287 4200‘ 1287 09FFFF 4210‘ 128A 020000 4220‘ 128D 09 4230‘ 128E FF96 4240‘ 1290 030000 4250‘ 1293 4260‘ 1293 4270‘ '1293 0DF0015D 4280‘ 1297 4290‘ 1297 D40000 4300‘ 129A C4 4310‘ 1298 0D 4320‘ 1290 FFO3 4330‘ 129E CA1281 4340‘ 12A1 4350‘ 12A1 4360‘ 12A1 4370‘ 12A1 09 4380‘ 12A2 040404 4390‘ 12A5 FD4E 4400‘ 12A? A8 4410‘ 12A8 030000 4420‘ 12AB . 4430‘ 12AB 4440‘ 12AB FBFFFCOl 4450‘ 12AF A8 4460‘ 1280 4470‘ 1280 88FF03 4480‘ 1283 4490‘ 12B3 CBOOOO 4500‘ 1286 4510‘ 1286 09 4520‘ 1287 FF6D 45 30‘ 1289 4540‘ 1289 080000 4550‘ 1280 4560‘ 1280 4570‘ 1280 F86D59 4580‘ 12BF 4590‘ 12BF 09 4600‘ 1200 FD4E 4610‘ 1202 030000 4620‘ 1205 FBFFFCO1 4630‘ 1209 F6 4640‘ 120A FBFFFCOI 4650‘ 120E 000000 4660‘ 12D1 F6 4670‘ 44202~F020—--~ 249 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 ~~~——-~—~—4577 00 04 .0 0088888888 0088888888 NOP3NOP3NOP NOP3NOP3NOP LDI A00(NSON)3 PLO NSONR LDI A01(NSON)3 PHI NSONR LDI 03 STR NSONR NOP3NOP3NOP 0005ET N‘03LET N‘H(NSONR) NOP3NOP3NOP3 STR NSONR 0000READ SONAR VALUE 0 SNR- REPEAT3 SEP CALL39A(READ) NOP3NOP3NOP3 0000IS VSONR‘FF LDN VSONR3SHI X’ FF’ LBZ TU000SKIP IF VSOngF LDN VSONR 5H1 150000D‘S'150 LDDF JUH150000DRANCH IF (+) 000015 SONAR0LT0150 ' 000N-N+1 LDN NSONR3 ADI 13STR NSONR 00000TURN "HEELS TU3SEP CALL33A(TURN) NOP LDN NSONR ' SHI 30000D-NSONR‘3 LBNZ REPEAT 0000HERE 3 GOOD VALUES HERE READ 0000CHECK ERROR ?-2+2 ' 0000 5.78-030” LDN VSONR 000D‘VSONR NOP3NOP3NOP SDI 78000D-78-VSON PLO R8000TEHP STORE ERROR LDDF EPOS0008RANCH TO EPOS 000HERE VSON IS NEG ' 000CONVERT TO POSITIVE ‘ XRI X’FF’3ADI 1000D‘+E ' PLO R8000TENP STORE ERROR 000015 ERROR GE 3 ‘ EPOS3OLO R898"! 3000-3 0000BRANCH IF D'NEG LDNF SETT32 7 0000SET DSUP LDN VSONR SHI 10900E‘VSON-109 MAGIC SAFETY NUHDER 0000TO PREVENT ADDITION OVER FLOU LDNF ERROR00DRANCH IF NEG 0000VSON IS TO LARGE 0000VSON IS 109 LDI 1093STR VSONR ....txtCOHPUTE DSHP ERROR3LDN VSONR SDI 7800 E‘78-VSONR LBDF E20000JUNP IF E.(+) XRI X’FF’3ADI 10000E IS NO“ (+3 8HR4000E.E/2 XRI X’FF’3ADI 10000E/2 IS NO" (NEG) LDR ADD32 E2 SHR0000 SHIFT THE POS E 'ADDB23ADI‘32- Figure D.l. (continued) 4680‘ 4690‘ 4700‘ 4710‘ 4720‘ 4730‘ 4740‘ 4750‘ 4760‘ 4770‘ 4780‘ 4790‘ 4800‘ 4810‘ 4820‘ 4830‘ 4840‘ 4850‘ 4860‘ 4870‘ 4880‘ 4890‘ 4900‘ 4910‘ 4920‘ 4930‘ 4940‘ 4950‘ 4960‘ 4970‘ 4980‘ 4990‘ 5000‘ 5010‘ 5020‘ 5030‘ 5040‘ 5050‘ 5060‘ 5070‘ 5080‘ 5090‘ 5100‘ 5110‘ 5120‘ 5130‘ 5140‘ 5150‘ 5160‘ 5170‘ 5180‘ 5190‘ 5200‘ 5210‘ 5220‘ 5230‘ 5240‘ 5250‘ 5260‘ 5270‘ 5280‘ 5290‘ 5300‘ 5310‘ 5320‘ 5330‘ 12D4 12D5 12D8 12DB 12DB 12DB 12DB 12DB 12DB 12DE 12E1 12E3 12E4 12E6 12E8 12EA 12EC 12EE 12EF 12EF 12FO 12F2 12F5 12F5 _12F5 12F6 12F8 12FB 12FE 1301 1304 1307 1307 1307 130A 1300 130E 130E 130E 130E 13°F 1312 1314 1317 1317 1317 131C 131F 1322 1325 1328 1329 132A 132D 132D 1330 1333 1333 1333 1336 1339 1330 133E 133F 1340 1341 57 000000 F8205? FBOOAB FBOOBB F800 EB 5860 5860 5B6O 5860 5860 SB 0A FF05 020000 0A F000 FCOIAB F8015B FBOOAB FBOOBB 040404 000000 00109D 04 0A 040404 FF05 CAOOOO F800F001AB F8015B F805A8 F80088 D40000 28 88 CA1325 C4C4C4 F8205? C4C4C4 FBOOAD F800BD 095D 1D 0A SD F800AD 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 250 STR DSUPR000DONE LBR AAA SETT323LDI 323STR DSUPR 0000(1) INITIALIZE FLAOR TO 0000 TOP OF STACK 0000(2) PUT ZERO IN ALL FLAG(X) 0000(3) PUT '1' IN FLAG(X) 0000 FOR X‘SNR AAA3LDI A00(FLAG)3PLO FLAGR LDI A01(FLAO)3PHI FLAGR LDI 0 ‘ SEX FLAOR STR FLAOR3IRX STR FLAOR3IRX STR FLAOR3IRX STR FLAOR3IRX STR FLAOR3IRX STR FLAOR 000015 SNR'S LDN SNR SHI 5 F L82 5N5 0000PUT A 1 IN 0000NEXT SNR LDN SNR . ADI A00(FLAO) ' ADI 13PLO FLAOR000A00(FLAGR)‘FLAGR + 1 LDI 13STR FLAGR LDI A00(FLAG)3PLO FLAOR.' LDI A01(FLAG)3PHI FLAGR" NOP3NOP3NOP ' 00008888888888888888888888888888888888888 00008888888888888888888888888888888888888' LDR A ’ JUH1503LDR LOOP2 A3NOP0000CONTINUE 00008888888888888888888888888 000015 SNR‘S 0000888888888888888888888 ' SN53LDN SNR NOP3NOP3NOP SHI 5 LDNZ DLOOP ' 00008EOIN DELAY 02 0000PUT A 1 IN FLAG(1) LDI A00(FLAG)3ADI 13PLO'FLA8R LDI 13STR FLAGR0000FLAG(1)‘1 LDI 53PLO R8 LDI 03PHI R8 DELAY 23SEP CALL3!A(TURN) DEC R8 OLO R8 ‘ LDNZ DELAY 2 00008ET ALL FL“G(X"O NOP3NOP3NOP3 LDI 323STR DSUPR 0000NOP3NOP3NOP ' 0000LOOP BACK TO TOP BLOOP3NOP3NOP3NOP LDI A00(SONNUH)3 PLO NSONR LDI A01(SONNUH)3 PHI NSONR LDN VSONR3 STR NSONR ' INC NSONR0000PTR TO DOXNUN LDN SNR000OET SN STR NSONR000 PUT AT BOXNUN LDI A001NSON)3'PLO NSONR " 251 Figure D.l. (continued) 5340- F 1344 PeooBO 524 L81 A.1(NSON)T PHI NSONR 5350- F 1347 840000 525 888 CALLT0A<818PL) 5360‘r 134A P80488 526 LOI 43PH1 R8 * 5370- 1348 P800A8 527 L81 oTPLO R8 5380‘ 1350 co1o9n 528 LBR LOOP2 5390- 1353 529 00.0HERE LOOP BACK 5400- 1353 530 ....TO TOP OF PR8 541o= 1353 531 0000ltttttttttttttttttttttttt! 5420: 1353 532 ....8888:8481xxx-8844888800488 5430- 1353 533 ....SUBROUTINE TO TURN UHEELs 5440- 1353 534 ....8:888:88:4008-048888088884 5450- 1353 535 ....8884 GET P0 P08 5460‘ 1353 536 0000tt¥¥AND TURN UNLS ~ 5470- 1353 O4c4c4 537 TURN:NOPTNOPTNOP 5480- 1356 E36101 538 sex PCtOUT1$DC 1 5490- 1359 EP6FAE 539 sex RPTINP 7TPLO BINR 5500- 1350 540 ...CONVERT TO BINARY 5510- F 1350 FBOOBE 541 LDI A.1 <81N>TPNI BINR - 5520- 135F 17 542 INc OSUPR ... PRT TO UPBIN 5530: 1360 543 ...:thxxx:UPBIN Is AOR BELOH THE AOR FOR 5540- 1360 08 544 LON BINR ... BET BINARY cone 5550- 1361 57 545 STR OSUPR ... 5560‘ 1362 27 546 DEC OSUPR ... POINT BACK TO OUSPR 5570- 1363 547 00.31tEXE0UTE UNEEL TURN - 5580- 1363 OE 548 LDN BINR ... D‘BINARY cone 5590- 1364 E7 549 sex OSUPR 5610- 1366 551 ...Pos-TURN R 5620- 1366 552 000NEG‘TURN L 5630- 1366 553 ...ZERO‘STOP 5640‘ F 1366 czoooo 554 L82 FOKAY 5650‘ F 1369 OBoooo 555 LBNF TFL 5660‘ 1360 556 ...NERE RESULT-PoszTURN R 5670‘ 1360 £3 557 sex PO 5680‘ 1368 558 ...OUTPUT PORT 08-8 5690‘ 1368 6108 559 OUT 1108 8...8ROUP NO.- 5700- 136F 6601 560 OUT 6IOC 1 5710- F 1371 cooooo 561 LBR OUNTRN 5720- 1374 E36108 562 TPL: sex POIOUT 1:88 8 ' 5730- 1377 6602 563 OUT 6Tnc2 5740- F 1379 cooooo 564 L8R OUNTRN 5750- 1370 E3 565 POKAYISEX Pc 5760- 137D 6108 566 OUT 1088 8 * 5770- 137F 6600 567 OUT 6188 o ' 5780- 1381 O4 568 OUNTRNINOP 5790- 1382 O5 569 SEP RETN - 5800- 1383 570 ....xxxxxxxxxxtxxtxxxt . 5810- 1383 571 ....8U8ROUTINE - 5820- 1383 572 ....xxxtxttxtxxtxxxtxx , 5830- 1383 573 00CONVERT BINARY TO Bcn. 5840- 1383 574 ..NUN8ER IN TEHP. 1 - 5850- 1383 575 ..INTO STALK POINTEO BY 8NR AT NUNB 5860- 1383 576 .. 5870- 1383 577 ..INITIALIZATION 5880- F 1383 PaooAA 578 8I8cn:LnI A00(NUHB)8PLO SNR 5890- F 1386 P8008A 579 L81 A.1IPHI 8NR 5900- 1389 F800 580 LnI o 5910- 1388 5A1A ' 581 STR 8NRIINc 8NR 5920- 138D 5A1A 582 STR SNR81NC SNR 5930- 138F 5A 583 STR SNR 5940- 1390 2A 584 888 8NR 5950- 1391 2A 585 888 8NR 5970- 1392 98 587 GM! TENP..O 5980- 1393 FF64 588 OIooxsnI 100..O-O-100 5990‘ 4' A -1395- CDOOOOHH ...--......___._ W-flEflfiTrffi 'D NEG?"— » Figure D.l. (continued) 6000‘ 6010‘ 6020‘ 6030‘ 6040‘ 6050‘ 6060‘ 6070‘ 6080‘ 6090‘ 6100‘ 6110‘ 6120‘ 6130‘ 6140‘ 6150‘ 6160‘ 6170‘ 6180‘ 6190‘ 6200‘ 6210‘ 6220‘ 6230‘ 6240‘ 6250‘ 6260‘ 6270‘ 6280‘ 6290‘ 6300‘ 6310‘ 6320‘ 6330‘ 6340‘ 6350‘ 6360‘ 6370‘ 6380‘ 6390‘ 6400‘ 6410‘ 6420‘ 6430‘ 6440‘ 6450‘ 6470‘ 6480‘ 6490‘ 6500‘ 6510‘ 6520‘ 6530‘ 6540‘ 6550‘ 6560‘ 6570‘ 6580‘ 6590‘ 6600‘ 6610‘ 6620‘ 6630‘ 6640‘ 6650‘ W'HVIW'HWI 1398 1399 139A 1390 139D 139E 13A1 13A2 13A4 13A6 13A9 13AA 13AB 13AD 13AE 13AF 1382 1383 1385 1386 1388 1389 1389 1389 1389 1309 1309 1309 1309 1309 1305 1301 1304 1307 13CA 1300 1300 1303 1303 1303 1304 1305 1308 1300 130E 1381 1356 13E9 13EC 13E0 13F0 13F2 13F3 13F6 13F? 13FB 13FC 13FF 1400 1402 1403 1406 1407 1409 140A 1400 88 0013A4 1A FCOA 5A 1A1A D5 0404040404 FBOBBE F800AA FBOOBA F800A9 F80089 F800A7 FBOOB7 09 88 F800A9 F80089 D41383 D483F0 1859262300 F800AA FBOODA 4A 0A0000 F820 8F D48198 000000 F030 8F D48198 252 590 592 620 621 622 623 624 625 626 627 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 1 ._ -...-_. 655 PLO LDN ADI STR . 8L0 TEHP00TEHP0OINTO D LDR D100 NEX1031N0 SNR 00N1 ADI 10000D‘D+100 D1038HI 10 LBNF NEX1 PLO TEHP LDN SNR ADI 100N1‘N1+1 STR SNR GLO TEHP LDR D10 NEX13INO SNR00N2 ADI 10 STR SNR00N2‘D (ONES PLACE) INC SNR‘INC SNR SEP RETN T 00FINISHED 0003883331littttttttttttttttl TEHP00D INTO TEHP00 SNR 00N(D) 10 0N'N4‘1 SNR 00 RCA DISPLAY 00tiltklttlltfittttlilll$113! DISPL3NOPONOPONOP§NOP8NOP LDI‘B’PHI RE LDI A00(NUHD)3PLO SNR LDI A01(NUHD)3PHI SNR LDI A0O(SONNUH)3PLO VSONR LDI A01(SONNUH)3PHI VSONR LDI A00(DSHP)3PLO DSUPR' LDI A01(DSUP)IPHI DSUPR: 0000TYPE'XIBI98’ 0 0 0 0OSTRG'X’83FO’ LDN VSONR PHI TEHP ‘ LDI A00(USON)8PLO VSONR" LDI A01(USON)3PHI VSONRr' SEP CALL‘0A(BI80D) SEP CALL19A(OSTR8) D0 X’1859262300’00PUTS CURSOR AT SPOT ON LDI A00(NUH8)1 PLO SNR - LDI A01(NUH8)3 PHI SNR ‘ LDA SNR LDNZ NZERO LDI X’20’ PHI RF SEP 0ALL’0A(TYPE) LDR SEO NZER03ADI 48 PHI RF ' 'SEP 0ALLOPA(TYPE) SEC3LDA SNR ” ADI 48 PHI RF SEP 0ALL30A(TYPE) LDN SNR ‘ ADI 48 ‘ PHI RF SEP 0ALL$0A(TYPE) DEC SNR, DEG‘SNRu- 253 Figure D.1. (continued) 6660‘ 140F 656 00 6670: 14°F 657 00 6680‘ 14°F 17 658 INC DSUPR 6690‘ 1410 07 659 LDN DSUPR 6700‘ 1411 27 660 DEC DSUPR 6710‘ 1412 88 661 PHI TEHP 6720‘ 1413 841383 662 SEP CALL‘9A(DIDCD) 6730‘ 1416 D483F0 663 SEP CALL39A(OSTRG) 6740‘ 1419 1859262000 664 DC X'1859262000’0000CURSOR TO NEH SPOT 6750‘ F 141E FBOOAA 665 LDI A00(NUH8)CPLO SNR 6760‘ F 1421 F8008A ' 666 LDI A01(NUH8)3PHI SNR 6770‘ 1424 4A 667 LDA SNR 6780‘ F 1425 CAOOOO 668 LDNZ NZERI 6790‘ 1428 F820 669 LDI X’20’ 6800‘ 142A 8F 670 PHI RF 6810‘ 1428 D48198 671 SEP CALL30A(TYPE) 6820‘ F 142E 000000 672 L8R SE01 6830‘ 1431 F030 673 NZER13ADI 48 6840‘ 1433 HP 674 PHI RF 6850‘ 1434 D48198 675 SEP CALL10A(TYPE) 6860‘ 1437 4A 676 SEC13LDA SNR 6870‘ 1438 F030 677 ADI 48 6880‘ 143A 8F 678 PHI RF 6890‘ 1438 D48198 679 SEP CALL10A(TYPE) 6900‘ .143E 0A 680 LDN SNR 6910‘ 143F F030 681 ADI 48 6920‘ 1441 8F 682 PHI RF ' 6930‘ 1442 D48198 683 SEP CALL30A(TYPE) 6940‘ 1445 2A2A 684 DEC SNR‘DEC SNR ' 6950‘ 1447 685 00 6960‘ 1447 686 00 6970‘ 1447 D483F0 687 SEP CALL30A(OSTRG) 6980‘ 144A 1859282F00 688 DC X’1859282F00’00PUTS CURSOR IN PLACE 6990‘ F 144F F800AA 689 LDI A00(80XNUH)3PLO SNR. 7000‘ F 1452 F8008A 690 LDI A01(80XNUH)3PHI SNR- 7010‘ 1455 0A 691 LDN SNR ' 7020‘ 1456 F030 692 ADI 48 ' 7030‘ 1458 8F 693 PHI RF 7040‘ 1459 D48198 694 SEP CALL‘9A(TYPE) 7050‘ 1450 F83F 695 LDI X’3F’ 7060‘ 145E 8F - 696 PHI RF 7070‘ F 145F F800AA 697 LDI A00(SN)1PLO SNR 7080‘ F 1462 F8008A 698 LDI A01(SN)‘PHI SNR 7090‘ 1465 D5 699 SEP R5 7100‘ 1466 700 0000888888888888888883888888 7110‘ 1466 D483F0 701 SCRN35EP CALL$IA(OSTRG): 7120‘ 1469 18185027 702 DC X’18185027’ 7130‘ 146D 3030303030463046 703 D0 T’OOOOOFOF" 7140‘ 1475 3043304330433043 704 DC T’OCOCOCOC’ 7150‘ 147D 18185022 705 DC X’18185022’ 7160‘ 1481 3030303033433343 706 DC T’00003C3C’ 7170‘ 1489 3043304330433043 707 DC T’OCOCOCOC’ 7180‘ 1491 18185023 708 DC X'18185023’ 7190‘ 1495 3043304330433043 709 DC T’00000000’ 7200‘ 1498 3043304330433043 710 DC T’OCOCOCOC’ 7210‘ 14A5 18185024 711 DC X’18185024’ 7220‘ 14A9 3043304330433043 712 DC T’OCOCOCOC’ 7230‘ 1481 3343334330303030 713 DC T’3C300000" 7240‘ 1489 1818502A 714 DC X’1818502A’ 7250‘ 148D 3043304330433043 715 DC T’OCOCOCOC' 7260‘ 1405 3046304630303030 716 DC T’0FOF0000’ I 7270‘ 1400 18185026 717 DC X’18185026’ 7280‘ 1401 3030303033463346 718 D0 T’00003F3F’ 7290‘ 14D9 3030303030303030 719 D0 T’00000000’ 7300‘ 14E1 18185030 720 DC X’18185030' 7310‘ ---14E5 3030303030303030‘H721“80“T‘00000000"“ Figure 0.]. (continued) 7320‘ 7330‘ 7340‘ 7350‘ 7360‘ 7370‘ 7380‘ 7390‘ 7400‘ 7410‘ 7420‘ 7430‘ 7440‘ 7450‘ 7460‘ 7470‘ 7480‘ 7490‘ 7500‘ 7510‘ 7520‘ 7530‘ 7540‘ 7550‘ 7560‘ 7570‘ 7580‘ 7590‘ 7600‘ 7610‘ 7620‘ 7630‘ 7640‘ 7650‘ 7660‘ 7670‘ 7680‘ 7690‘ 7700‘ 7710‘ 7720‘ 7730‘ 7740‘ 7750‘ 7760‘ 7770‘ 7780‘ 7790‘ 7800‘ 7810‘ 7820‘ 7830‘ 7840‘ 7850‘ 7860‘ 7870‘ 7880‘ 7890‘ 7900‘ 7910‘ 7920‘ 7930‘ 7940‘ 7950‘ 7960‘ 7970‘ 3346334630303030 18185028 3030303033463346 3043304330433043 18185029 3043304330433043 3346334630303030 1818433018185230 OCOAOA 2726262626262626 2826262626262626 26262622 2320534F4E415220 2320205748454540 2020202323205641 405545202320504F 534954494F4E2023 2309092309092020 2023 2309092309092020 2023 2A303C3C3C303C30 2930303030303030 30303024 202020 534F4E4152 20 554E49543A 00 85 04 0001030207060405 0F0£00080809080A 1F1E10181819181A 1011131217161415 3F3E303D3839383A 3031333237363435 2021232227262425 2F2E20282829282A 2000 FF oooooooooo 000000 254 722 723 724 725 . 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 no T’3F3F0000’ no X’18105028’ no T’00003F3F’ no T'oooooooo' no X’10105029’ no T'oooooooo' no T’3F3F0000’ no X’1810433010105230’ no X’OCOAOA" no T'szzzxzz' no T'cxxxsxzx' no T’ttt" no X’2320534F4E415220’_ no X’2320205748454540’ no X’20202023232OS641’ no x'485545202320504r'j_ no X’534954494F4E2023’ no x'23o9o92309092020' - no x'2023' no x'23o9092309092020' no x'2023' ' DC T’I<<<<<<<’ no T’)<<<<<<<’ no T’<<<$’ no x'20202o', no T’SONAR’ SEP R5 0000118818188388131181311 000DATA 0 0 0 0.8888888: ' ORG X’1700’ TOPSTK8NOP ' 0000UHEELL POSITION LOOK-UP TABLE 000000CONUERT GREY CODE- 000000TO BINARY ' ORG X’1600’ BIN3DC X’0001030207060405’ DC X’0FOE00080809080A’ DC X’1F1E1C1D1819181A’ DC X’1011131217161415" DC X’3F3E303D3839383A’ DC X’3031333237363435’ DC X’2021232227262425’ D0 X'2F2E20202829282A’ ORG X’1650’ DSUP:DC X’2000’ USON3DC X’FF" NSON3D0 X'0000000000’ DC 0 NUHBSDC SN:DC 0 FLAG:DC 0 D00 D00 800 000 800 SONNUHX DCO BOXNUH3D01 ORG X’1750’ DST: LDR HSBE“ X’000000’ Figure D.l. (continued) 7980‘ 7990‘ 8000‘ 8010‘ 8020‘ 8030‘ 8040‘ 8050‘ 8060‘ OK- 1753 1756 1759 1758 1750 1760 1762 1763 1764 FBEFAC F88080 468F 3200 040404 3059 85 D5 255 788 789 790 791 792 793 794 795 796 ‘Nsse: LnI X’EF’i PLO Ro LnI x'80's PHI Ro ' NSOE1: LDA LINK; PHI RP 82 EXITH. NOP6NOP6NOP 8R NSOEI EXITH: SEP R5 TYP: SEP R5 END APPENDIX E DATA FROM PERFORMANCE TESTS OF THE HARVESTER'S STEERING CONTROL SYSTEM APPENDIX E DATA FROM PERFORMANCE TESTS OF THE HARVESTER'S STEERING CONTROL SYSTEM Appendix E contains the data that were collected during the performance tests of the harvester's steering control system as described in Sections 9.1 to 9.2. Tables E.1 to E.4 show the data that represents the harvester's position as the harvester moved over a simulated tree row during the performance tests. Performance tests were done with straight row, curved row, and a row with an 8 cm step-change. The data in these Tables E.1 to E.4 are the X-Y position coordinates of selected points on the lines drawn by two pens which were attached to the harvester's frame. One pen was attached at the front wheel and the other at the rear wheel (Figure 9.3). Note, that for each X-value in the tables for the front of the harvester, there is a corresponding set of rear X-Y coordinates. These coordinates for the front and the rear of the harvester completely define the position of the harvester at one particular instant of time during a performace test. Note that the X coordinate for the rear of the harvester was computed by triangu- lation. The steering control ysstem was required to maintain the harvester's centerline within about 20 cm fro each tree's centerline as ‘the harvester traveled over each tree. Thus, to determine the position of the harvester's centerline, the coordinate position of points A and B on the harveter were computed and are shown in Tables E.5 to E.8. 256 257 Points A and B were on the harvester's centerline (Figure 7.5). Point A was between the front wheels and point B was between the rear wheels. The data in Tables E.5 to E.8 was used to plot the path of the harvester observed during the performace tests and these plots are shown in Sections 9.4 to 9.5.1. 2258 Table E.1. Harvester's Position Data Collected for Three Performance Tests with a Straight Row. STRAIGHT RUN ( R ‘ 121.9 N ) JULY 112 v 198? SURFACE ‘ CONCRETE DRIVEUAY I RUN 1 I RUN 2 I RUN 3 I FRONT I FRONT I REAR I REAR I FRONT I REAR I REAR I FRONT I REAR I REAR I X I Y I X I Y I Y ‘1 X I Y I Y I X I Y I -------- I---——--—I--- I I------—-I--------I----—--I----~---1-——-—--—I-—----—1 0 I 59.5 -415.5 60.5 I 6205 -415.5 61.0 I 6005 "415.5 60.0 I 150 I 5805 ‘26505 6005 I 6105 "26505 6I00 I 6100 “26505 5900 I 300 I 5805 -11505 6000 I 6000 -11504 5100 I 5900 -115.5 59.5 I 450 I 5505 3405 59.5 I 5900 34.5 60.0 I 58.5 34.5 6000 I 600 I 5605 18405 58.5 I 58.5 18405 59.0 I 57.5 18405 59.5 I 750 I 5500 33405 5805 I 5900 33405 6000 I 58.0 -33405 5900 I 900 I 5600 48405 5800 I 57.5 48405 59.0 I 5705 484.5 58.5 I 1050 I 5605 63405 5705 I 5705 63405 5800 I 57.5 63405 5800 I 1200 I 5605 78405 5800 I 57.5 78405 57.0 I 5705 784.5 58.5 I 1350 I 5605 934.5 5800 I 5800 93405 5605 I 59.0 -93405 58.0 I 1500 I 5605 108405 5800 I 5800 108405 5700 I 5900 108405 58.5 I 1650 I 5505 123405 5800 I 5900 123405 57.0 I 5900 123405 5805 I 1800 I 5505 138405 5900 I 6000 138405 58.0 I 6000 138405 59.5 I 1950 I 5605 153405 5805 I 6105 153405 5900 I 6005 153405 59.5 I 2100 I 5700 168405 5800 I 5905 168405 5805 I 5900 1684.5 5900 I 2250 I 57.5 183405 5800 I 6000 183405 5800 I 5905 1834.5 6000 I 2400 I 5800 1984.5 5805 I 6100 198405 5805 I 57.0 1984.5 59.5 I 2550 I 5800 213405 6000 I 62.0 213405 5900 I 5700 2134.5 58.5 I 2700 I 5705 228405 6000 I 6005 228405 5905 I 57.0 228405 5705 I 2850 I 5605 243405 5905 I 6005 243405 6000 I 57.5 2434.5 57.0 I Table E.2. Harvester's Position Data Collected for Three Performance Tests with a Row Containing a 8 cm Step—change. STEPPED RUN ( STEP SIZE ‘ 8 0H ) JULY 12 9 1982 SURFACE ‘ CONCRETE DRIUEHAY STEP BEGINS UITH TREE 4 AT ( 1219029 8.0 0H) I RUN 1 I RUN 2 1 RUN 3 I FRONT I PRONT I REAR I REAR I FRONT 1 REAR I REAR I FRONT I REAR I REAR I x I Y I x 1 Y I Y I x I Y I Y I x I Y I -------- I--—-----1----——--I----——-I--------I--------I--—-—--I-—----—-1------—~I-------1 0 1 56.5 -415.5 58.5 I 59.5 —415.5 58.0 I 62.5 -415.5 57.5 I 150 I 54.0 -265.5 58.0 I 59.5 -265.5 58.0 I 64.0 —265.5 58.5 I 200 I 54.5 -215.5 58.0 I 55.5 -215.5 58.0 1 62.5 —215.5 58.5 I 250 1 54.5 ~165.5 57.0 I. 50.5 -165.4 57.0 I 61.5 -165.5 58.0 I 300 I 50.0 -115.4 57.0 I 47.0 -115.4 57.0 I 61.5 —115.5 59.0 I 400 I 47.0 -15.4 55.0 I 45.5 -15.4 54.5 I 62.0 -15.5 58.5 I 460 I 44.5 44.6 54.5 I 44.5 44.6 53.0 I 62.5 44.5 58.5 I 600 I 52.0 184.5 53.0 I 52.0 184.5 51.0 I 62.0 184.5 59.0 I 750 I 52.0 334.5 53.5 I 54.0 334.5 51.5 I 63.5 334.5 58.5 1 800 I 53.5 384.5 53.5 I 55.5 384.5 52.0 I 63.0 384.5 59.0 I 900 I 56.5 484.5 53.5 I 57.5 484.5 52.0 I 62.0 484.5 59.0 I 1050 I 56.0 634.5 55.0 I 60.0 634.6 53.5 I 63.5 634.5 59.0 I 1100 I 57.5 684.5 55.0 I 61.0 684.6 54.5 1 64.0 684.5 59.0 I 1200 I 61.5 784.5 56.5 I 64.5 784.6 56.0 I 66.0 784.6 59.5 I 1400 I 63.5 984.5 59.0 I 68.5 984.6 59.5 I 68.0 984.6 60.5 I 1600 I 66.0 1184.5 62.0 I 71.5 1184.6 63.0 I 70.5 1184.6 62.5 1 1800 I 66.0 1384.5 64.0 I 70.0 1384.5 66.0 I 70.5 1384.5 64.5 I 2000 I 65.5 1584.5 66.0 I 71.0 1584.5 68.0 I 72.0 1584.5 66.5 I 2200 I 65.0 1784.5 65.5 I 70.5 1784.5 67.5 1 70.0 1784.5 65.5 I 2350 I 64.5 1934.5 65.5 I 70.5 1934.5 68.0 I 68.0 1934.5 65.5 I 2400 I 64.0 1984.5 65.5 I 68.5 1984.5 68.0 I 68.5 1984.5 66.0 I 2500 I 63.5 2084.5 65.0 I 67.0 2084.5 67.0 I 70.0 2084.5 65.0 I 2700 I 65.5 2284.5 65.0 I 66.5 2284.5 66.5 I 70.5 2284.5 65.5 I 2750 I 67.0 2334.5 65.0 I 67.0 2334.5 66.5 I 71.0 2334.5 66.0 I 2750 I 67.0 2334.5 65.0 I 67.0 2334.5 66.5 I 71.0 2334.5 66.0 1 Table E .3. 259 Harvester's Position Data Collected Tests with a Curved Row on a Campus CURVED RUN ( R ‘ 1085.0 1135.0 1235.0 125500 1385.4 141505 1535.4 1555.4 1635.9 1661.1 1686.2 1716.2 1786.4 1836.2 1856.2 1868.3 1936.9 1987.2 2017.5 2057.5 2137.6 2167.2 2217.8 2273.5 2288.9 2329.5 121.9 N 1 JULY 9 SURFACE ‘ LAUN 50.0 ‘8005 ‘4605 “5305 -55.5 '61 05 I RUN 2 I FRONT HtflhiHlflb4~IHhIHIHFOHINHIHINHIiiIHI‘F‘flIIHDdFIHI‘FIHOflFIHtflhifltiht M (I 0 UI I -118.0 1084.9 1135.0 1234.9 1254.9 1385.4 1415.5 1535.3 1555.3 1635.9 1661.1 1686.2 1716.2 178602 183601 1856.1 186802 1936.8 1987.2 201704 2057.4 2137.3 2167.1 2217.7 227303 2288.3 2328.4 1982 -2600 for Three Performance Lawn. ‘41505 -265.5 ’11505 34.6 184.5 284.5 334.5 1384.5 484.5 634.5 644.5 784.6 934.6 964.6 1034.9 1084.9 1135.0 1234.9 1254.9 1385.4 1515.4 1535.3 1555.3 1635.7 1660.9 1686.0 ”7716.0 1786.1 1836.0 1855.9 1868.0 1936.7 1987.2 2017.4 2957.4 2137.5 2167.2 2217.7 2273.2 2288.3 2328.3 “3‘05 -53.0 “610° HMO-0“!HNH"NHHHHIQHHHHHHHHHHNHHHHHNNHHHHHHHHHNHHH Table E. 4. 260 Tests with a Curved Row on a Concrete Driveway. CURUED RUN ( R ‘ 12109 N ) SURFACE ‘ CONCRETE DRIUEUAY -24.5 ”2805 '3‘05 ‘8000 ’4100 -4205 ~55.0 ”6305 -6800 -73.0 “810° -84.5 -9505 I “10805 I -111.5 I '11805 REAR I REAR X I Y -1 -415.5 58.0 -265.5 6100 -11505 61.5 34.5 60.5 184.5 60.5 284.5 6000 334.5 60.0 384.5 60.0 484.5 5905 634.5 59.0 784.5 5705 934.5 54.5 964.5 54.0 103407 52.0 1084.8 5000 1134.8 47.5 1234.8 42.5 125408 41.5 1385.3 3400 1415.3 32.0 1535.3 24.0 1555.3 22.5 1635.8 16.5 1661.0 14.5 1686.1 12.0 1716.2 9.5 1786.2 2.5 183602 ’205 1856.1 -4.5 1868.2 -5.5 193606 ’1300 1986.9 -19.0 2016.9 -2300 205700 ’2705 2136.9 -3605 2166.9 -4000 221703 ”4705 2273.1 -54.0 228802 -56.5 2328.4 -62.0 I RUN 2 I FRONT I 0~ O O 0-01-4nuunuuunnunnuuuuuuunuHHH—nnnuu-n-au-cnv-u-n-n-u-n M L! O O I ‘11000 I ”11205 I ’11905 JULY 7 9 1982 REAR I REAR X I Y - I -415.5 60.0 -26505 62.0 -11505 62.5 34.5 61.0 18405 6100 28405 60.5 334.5 60.5 38405 60.5 484.5 60.5 63405 59.5 784.6 5805 93406 55.0 96406 55.5 103408 5205 1084.9 50.5 1134.9 48.0 1234.8 4300 1254.8 4200 1385.4 35.0 141505 3300 1535.6 25.0 1555.6 2300 163602 17.5 1661.3 15.0 1686.4 1300 1716.4 905 1786.4 2.5 1836.5 -2.5 1856.5 ~4.5 186806 -5.5 193700 -1300 1987.3 -19.0 201703 -23.0 2057.3 -27.5 2137.0 -37.0 2167.3 540.5 2217.5 -48.0 227302 -54.5 2288.3 ~56.5 2328.4 ~62.5 HHNHHHHHHHHMHHHuHHHMH—H—nufi‘uuhHflFOHI‘HHHHV I -11100 I -118.0 REAR X I I I I I I I I H I I I I I I I I ‘41505 1254. 1385.7 1415.7 1535.7 1555.7 1636.3 1661.4 1686.6 1716.7 1786.6 1836.7 1856.6 1868.8 1937.2 1987.6 2017.6 2057.7 2137.9 2167.6 2218.0 2273.6 2288.8 2328.9 -——---— -3200 ‘3505 ‘4205 -4905 ‘5105 -5705 Harvester's Position Data Collected for Three Performance III—4HHHHHNHHHHHHHI‘HH~HHHN~HHH~HHHHHHMHHHF‘HHHNH 261 —~H~~~_~~~~~HNHHHHH~H~~H 0.n1 0.—1 n.«1 n.—1 0.N1 n.~1 0.N1 D.~1 0.—1 IIIIIIIIII ~I.I.II.1. I.II~IIII,I.III~IIIIIIII > 2ch m ucwom-gmom new < pcwom-ucogm mo mmumcmugoou :owuwmoa umusasou 0.00Nn D.~1 5.0—«n 0.n1 0.000N 0.n1 v.0—DN 0.n1 n.000n n.1 0.0«hw 0.—1 0.00nn n. 0.0~NN 0. 0.000N 0.~1 0.0«0— 0.~1 0.00- 0.u1 0.0—on n.N1 n.000— n.N1 0.0—nu m.N1 0.00u— 0.N1 0.0—On D.N1 0.000 n.—1 5.0—k 0.~1 «.00m 0.— 0.0.? m. a x n > u u «cum u hzoxu a 0. ..vnvfl .D. 0.0'ON n 3.1 0.nnVN D.n1 «.0'0N “.1 u.'cwm .n. 0.000N u 0. 00nmmfl n..1 «.000N 0.~1 c.0n—N 0.“ 0.0'mN u 0. h.nn—N 0.m1 N.0QDN m.—1 '.¢O0u 0." 0.00MN — n.u1 0.nm0~ 0.31 c.00nN 0.N1 n.0no— 0. 0.0'NN - 0.N1 0.nno~ n.N1 Q.0¢NN n.u1 u.¢&o~ 0.1 0.000N ~ 0.m1 0.n00— 0.n1 H.000N 0.—1 v.vnn— m.— 0.0.0. u fi.~1 n.nnnu n.n1 N.0v0— 0.N1 n.tan— 0. 0.00n— — 0.—1 h.n3n— n.v1 0.000— 0.nl n.0nw— 0.«1 0.0'0— n 0.N1 0.nnflu fl.vl «.0'ou 0..1 —.?00— 0.N1 0.00?“ a 0.N1 Q.n%0u n.n1 n.00v— n.n1 N.vn0 0.N1 k.0¢n— h 0.51 w.n£0 “.31 n.0vn— 0.n1 «.Ccn n.N1 0.00". h 0.61 O.nmh $.81 ”.00"— 0.N1 0.nm0 n.NI (.0'0— n c.ml O.nn0 n.n1 n.010— 0.—1 D.nzt n.N1 n.00m u 0.N1 h.no' 0.?1 N.00O 0. O.nnn 0.u1 ”.000 n t.u1 n.nnn 0.01 0.0Ch 0.—1 0.n3« n.—1 v.00“ — fi.u1 h.n¢— m.n1 «.000 0. O.nn 0..1 n.0vv u 0.1 '.nfl $.01 0.0?0 0.01 o.¢—~1 0. 0.00” n 0. N.oun1 n.ul n.00fi 0.« 0.00N1 n.. c.0vu — n. n.00w1 n.u1 N.0v« 0.“ 0.0—'1 m.N ”.1 n n. n.0u'1 n.1 3.1 11111111 ~11111111~11111111~1111 u u n ~11111111 > a x > u x n > n x u > u x «(mu m ¢¢mz pzoak a black n ecu“ x ¢¢w¢ u bzomk u bzozu . ~23.— .22. ><3w>~¢a ubmzuzcu a mucuXDm N00— . may >420 2:2 hzo~¢zhm .3om “cmwwgum saw: mama» mucmELomgmm muggp Loy m m mpnwh 0.000N v.00on n.0vnn «.00nN v.0QNN n.000n 0.0v0n 0.005" 0.0ccn 0.00v~ 0.0an n.00—n v.0qo— H.000 n.0th H.00n n.0vt c.00m 0.0?“ v.1 x bleak m 222 h~~~H~H~~HH~~H~~HHHHHNNH HHHHHHNNHHhHH~~~—H~~NN~— 262 III—"HFOHHNHHHHi-OHHHHHMHMHHHHHHHHHHHHHHHHHHHH .. 0.0..1 0.00.0 0.05.1 0..n5N . 0.0N.1 5.0.0m 0.55.1 n..n5N . 5.0N.1 5.0.0N 5.00.1 0.005N . 0.0..1 0.00.n 0.05.1 0..00N . 0.0..1 5.0505 0.05.1 0..00N . 0.0..1 5.05N5 0.05.1 0.000N . 0....1 N.N00n 0.50.1 0.050N . 0.N..1 5.00NN 0.00.1 0.0505 . 0.5..1 5.0005 0.05.1 N.050N . 0.00.1 ..0000 0.50.1 ..NNON . 0.00.1 0.00NN 0.50.1 ..NNON . 0.00.1 5.0055 0.00.1 0.NNON . ..001 n.000n 0.00.1 5.N50~ . ..00.1 0.00.5 ..00.1 0.N50N . 0.00.1 0.00.5 0.50.1 0.050N . 0.001 0.0005 0.00.1 0.5005 . 0.001 0.0N.N 0.00.1 0.NO0N . 0.001 0.00.0 0.00.1 N.NO0N . 0.001 0.050N 0.00.1 0.NO0N . 0.001 0.0000 0.00.1 0.NO0N . 0.001 0.0000 0.00.1 n.~00N . ..001 0.0005 0.05.1 0.5NON . 0..01 0.0000 0.00.1 0.N50N . 0..0J 0.000N 0..0.1 n.NNON . 0.051 0.0005 ..0N.1 5.5005 n 0.551 0.050. 0.0N.1 0.N0na . 0.051 0.050. 0.00.1 0.500N . 5.051 0.0050 ..0..1 0.0005 . 0..51 0.000. 0.0..1 ..0005 . ..N51 0.000. 0.0..1 ..0000 . 0.001 0.0005 5.00.1 0.0555 . 5.001 0.500. 5.00.1 ..05NN . 5.001 H.000. 0.00.1 0.0505 . 0.001 0.NO0N 5.001 0.0050 . 5.001 0.000. N..0.1 0.00NN . 5.001 H.000. 5.N0.1 0.00NN . 0.001 N.NO0N 5.501 0.00am . 5.N01 H.000. N.001 N.00NN . 5.501 N.000. 5.00.1 0.00NN . 0.001 0...0N 5.n01 N.00.N . 0.501 ".005. «.001 0.00.5 . «.501 N.005. 5.001 0.00.N . 0.001 ..NO0N 5.001 0.05—N . 5.001 0.0.5. 5.501 ..0N.N . N.001 N.0.5. 0.001 0.0N.N . 0.001 0.0.05 «.501 0.000N . 5.501 0.000. «.001 ..000N . N.501 0.000. 5.001 ..000N . 0.001 0.5005 0.051 0.000N . 5.001 0.000. 5.001 , n.000m . 5.001 0.000. N..01 N.000~ . 0.n01 0.NO0N 0.051 0.0005 . 5.501 0.000. N.551 0.0000 . 0.001 0.000. 0.551 0.0005 . 0.501 5.0005 0.001 5.000. . 0.001 0..00. 0.001 5.000. n 0.001 0.000. 0.001 0.000. . 0.001 0.0005 0.N01 0.000. . 0.001 0..n0. 0..01 5.000. . 0.001 0.000. 0.001 0.000. . 0.551 0.00NN n.001 0.050. . 0.0N1 0.0.0. 0.001 0.030. . 0.0N1 0.0.0. 0.001 0.050. . 0.0m1 0.0.55 0.501 0.005. . 0.0N1 0.00m. 0.N01 0.005. n 0.051 0.000. 0.001 0.005. . 0.0.1 0.0005 0.001 0.000. . 0.5.1 0..0N. 0.001 0.500. . 0.5.1 0..0N. 0.501 5.000. . N.5.1 0.000N 0.001 5.000. . 0.0.1 0..nN. 0.001 0.000. . 0.0.1 0..MN. 0.001 0.000. . 5.5.1 0.000. 0..01 5.000. . 0...1 0..n.. 0.001 5.000. . 0...1 0..n.. 0.an 0.000. . 5.01 0.0.0. 0.051 5.000. . 0.01 5..00. 0.5N1 0.000. n 0.01 0..00. 0.051 5.000. . 0.51 ..000. 0.051 ..500. . 0.01 0..00. 0.051 . 0.500. n 0.51 0..00. 0.051 0.500. . 0.01 ..505. 0.0.1 0.050. . 0.01 0.N00 0.0.1 0.550. n 0.01 0.N00 0.5.1 5.550. . 0.01 0.505. 0.n.1 0.000. . 0.01 0.Nn0 0.N.1 0.000. . 0.01 0.500 0.0.1 0.500. . 0.1 0.5.0. 0.0.1 5.00.. . 0. 5.005 0.0.1 0.00.. n 0. 0.505 0...1 0.50.. . 0.. 0.050. 0.. 0.000. . 0.N 0.000 0.. n.000. . 0.N 0.000 0.1 ..000. . 0.5 0.000. 0.. 0.000. . 0.0 0.000 0.. 0.000. . 0.0 0.000 0.1 0.000. . 0.. 5.0.0. 0.. 0.000 n 0.0 0.000 0.. 0.000 . 0.N 5.000 0. N.000 . 0.. 0.0.5. 0.0 0.000 . 0.0 0.000 0.0 0.005 . 0.5 0.000 0.0 5.005 . 0. 0.00.. 0.0 0.005 . 0.0 5.000 0.5 .a005 . 0.. 5.000 0.0 N.005 . ..1 0.0... 0.0 0.005 . 0.N 5.005 0.5 ..005 . 0.. 5.005 0.0 ..005 . 0.1 0.0.0. 0.0 5.000 . 0.5 0.00. 0.0 0.000 . .. 0.00. 0.0 ..000 . ..NI 5.000 0.0 0.000 . 0.. ..00 0.0 0.000 . 0.1 5.00 0.0 0.000 . ..01 0.0.5 0.5 0.000 . 0..1 0.0..1 0.0 0.000 . 0.N1 0.0..1 0.N ..oon . ..01 0.000 0.N 0.00. . 0.N1 0.00N1 0.5 0.00. . 0.N1 0.0051 0.. ..00. . 0.01 5.0.0 0..1 ..1 . 0.01 N.0.01 0.0 n. . 0.N1 0.0.01 0..1 0.1 . 11111111 . 11111111 . 111111111 . 11111111 .1 1111111 . 11111111 . 11111111 n 11111111 m 111111 11. 11111111 .11111111. 11111111 . > . x . > . x . > . x n > . x . r n x n > . x . 0000 . «00¢ . 52005 . 02000 . «cum . 0001 . 02005 . 02000 . «<08 m 0000 . bzomu . 52000 . 0 2:0 . N 230 . . 23¢ 2304 n 0000030 000. . 0 >420 . t 0..N. u 0 0 2:: 000030 .czm; maasmu :0 3oz cm>g=u cam: mummh mucmELomgmm muggh go0 m u:.om-gmmm ucm < ucwoa-pcogm 0o mmumcwugoou cowuwmoa nmuzqsou .o.m mpam. ~~~H 263 HM~~H~~~~~~H~~h~H~~~HH~H~~H~~~~~H~~HH~~H~~~H 0.0..1 0.0..1 0.00.1 0..0.1 0.001 0..01 0..01 0.551 0.051 0.501 5.001 0.001 5.001 0..01 5.001 0.001 0.001 0.501 0.001 0..01 0.001 0.001 0.0.1 5.0.1 5.51 0.01 0.01 0..1 '0' ..0 0.0 0.0 0.00.0 0.55.1 0.00.0 0.05.1 0..000 0.50.1 0.5000 0.00.1 0.0000 0.00.1 5.5000 0.00.1 ..0500 0.00.1 0.0000 0.00.1 0.0000 0.00.1 0.0050 0.0..1 0.0000 0.00.1 0.0000 0..0.1 0.0000 0.001 0.0.00 ..001 0.0000 ..001 5.0.00 ..001 0.0000 0.001 0..000 0.551 0.0000 0.001 0.0000 0.001 0.0000 0.001 ..0.00 0..01 0.5000 0.001 0.0000 0.001 0.000. 0.001 0.0.0. 0.001 0.000. 0.001 ..505. 0.0.1 ..505. 0...1 ..5.0. 0.01 ..000. 0.01 0.0.0. 0.1 0.0.0. 0.. 0.00.. 0.0 0.0... 0.. 0.0.0. 0.. 0.000 0.0 0.0.5 0.0 0.000 0.0 0.0.0 0.0 x > . 0000 .2000 . 0.0050 0.0000 0.0500 0..000 0.0500 0..000 ..0000 ..0000 0.0000 0.0000 0.0500 0.0000 0.0000 0.00.0 0.00.0 0.0000 0.0000 0.0000 0.000. 0.000. 0.000. 0.005. 0.000. 5.000. 0.000. 0.000. 0.500. 0.550. 0.000. 0.50.. 0.000. 0.000 ..005 0.005 ..000 0.000 0.000 0.000 0.00. 02010 0 2:0 ~~HHH~H~HHH~~HHH~~H~H~~~~~~H~H~~~~~~H~N~~~HH 000. . 0..N.1 0.0..1 0.0..1 0.50.1 0.00.1 0.001 0.501 0.001 0.051 0.051 ..001 0.001 0.001 0.501 0.001 5.001 5.001 0.001 0.001 0.001 0.001 0.001 0.5.1 0.0.1 0...1 0.01 0.51 0.01 0.01 0..1 0.1 no no 0..050 0..000 0.0500 0.0000 0.0500 0.0000 0.0000 0.0000 0.0000 0.0000 0.0500 . 0.0000 n.o.nm o.mn~1 n.o-n o..n.1 n.voma 0.90.1 m.oon~ n.un.1 n.0nuu o.mc_1 o.o~.m _.~c_1 o.owou m.nn.1 o.ooo~ o._n.1 o.oso_ .n.o~.1 . 9.9m». ..m..- ..ucm. .o.oo—1 ..onm. ~.no_1 ..ono. u.no.1 ~.oou_ .5.¢on . ..o.5_ u.oo-. . ..omo. n.0m1 ~.nmo. u.na1 n.ono. .5.on1 , 5.onn— n.501 n.onn. n.nc- s.o_v. . o.nn1 o.ocn. .o.~n1 o..n~. o.cm- o..nu. o.nn1 o..n~_ 0..“- o._moq o.o~1 o.nno. o.u~1 o.~oo o.m.1 m.~no n.n.1 m.nmn o.o.1 z.nmc n..- m.m:c o..- o.ccm n. ..vnn m.. c.0an n. m.na. o. o.mn n. n.o_.1 m. o.oo~- o.m n.n_c1 o.m 11111111 .11111111_1:-11111. x » . «qua .zoau — 0.0000 0.00.0 5.00.0 5.0000 0.0000 ..0000 ..000. ..000. 0.000. 0.005. 0.500. 0.500. 0.000. 0.500. 0.500. .0.050. 0.000. 0.00.. 0.000. 0.000 0.005 0.005 0.000 0.000 0.000 0.000 0.00. 0.1 .2000 0 2:0 »¢:u>_¢a w.wzuzou u mucuzzm a . zac amazzu 5 >400 . : 0..N. u .amzm>.go mumgucou co 3cm um>gzu gum: mpmm. mucwsgo0gmm mugs. Low m unwoa10mmm was < ucwoa10cogg 0o mmumcwugooo cowuwmom uwpaasou HHhMH—Iflht-(HHHH—hHHhNhHHHHHHHflN‘~h~~~~~~~~~~~~ 0..0.1 5.0.00 0.0..1 5.0500 0.0..1 5.0000 0.50.1 0.0000 0.001 0.00.0 ..001 0.00.0 ..501 0.0000 0.001 0.0000 0.051 0.000. 0.051 ..000. 0.001 0.000. 0.001 0.000. 0.001 0.000. 0.501 0.005. 0.001 0.0.5. 5.501 0.000. 09001 0.000. 0.001 0.000. 0.501 0..00. 0.001 0..00. 0.501 0...0. 0.001 ...00. 0.0.1 ..000. 0.5.1 0.000. 0.0.1 0.00.. 0.01 0.000. 0.01 0.000. 0.01 0.000 0.01 0.000 0.01 0.005 0..1 0.000 0.1 ..000 0. 0.000 0. 5.000 0.1 0.000 0. 0.00. 0. 5.00 0.. 0.0..1 0.. 0.0001 0.01 0.0.01 > x 0000 “(00 -11. . . 0.55.1 0..050 0.05.1 0..000 0.50.1 0.0500 0.00.1 0.0000 ..00.1 ..0500 0.00.1 ..0000 0.00.1 0.0000 0.50.1 0.0000 ..00.1 ..0000 0.0..1 0.0000 0.00.1 ..0500 5.00.1 0.0000 5.001 ..0000 0.001 ..00.0 0.001 0.00.0 0.001 0.0000 0.001 0.0000 0.051 5.0000 0.001 5.000. 0..01 5.000. 0.001 5.000. 0..01 0.005. 0.001 0.500. 0.001 0.500. 0.001 0.500. 0.001 0.500. 0.001 5.500. 0...1 5.050. 0.0.1 0.000. 0.01 5.00.. 0.. 0.000. 0. 0.000 0.0 0.005 0.0 ..005 0.0 0.005 0.0 5.000 0.0 ..000 0.0 5.000 0.0 ..00. 0.0 0. .11 > . x block . 02000 . 2:“ .5.0 m.nmh Hun-IHHMMHflHI-t"NH—"C‘MHNHHNHHHHN-MNH~~~H~~IHHHHH C'DIDIDOOIDO N~~~~~~H~~Hfl~~~H~~~HH~HHNNHH .0 0.00un 0.~— N.00hN u 0.0 n.¢nnN 0.h 0.0?hu ~.0.n n.vnnn 0.h 0.0?hw .0 0.0—«n 0.0" N.00hN ~ 0.0 0.?0NN . 0.0 0.000N n 0.D «.VONN 0.0 0.000N .m n.0«ow 0.0— N.00mN a 0.h 0.?00N 0.h 0.00CN u 0.n 0.n00N D.» n.00vN .0 «.0—0N 0.0 0.00MN u 0.0 —.'00— “.0 0.00nN n 0.0 0.n00. 0.? 0.000N .m «.00NN 0.0 0.0'nN u 0.0 v.vno« n.0« 0.0?0N n h.n 0.000« 0.? n.000w .n 0.0.0N 0.0a «.00NN u m.& v.'0hn 0.0“ 0.00aN n 0.0 0.005u 0.6 C.00~N .0 0.0—Cm 0.N~ n.000u u 0.0 '.v00— 0.". 0.000u n 0.0 0.000" 0.0 v.000u .v 0.0—NW 0.0“ c.000— a 0.0 0.?0n— , 0.0— «.000. u 0.? 0.70nu 0.0 0.00m— .N 0.0—0N n.0— 0.000— n 0.0 n.00uu D.un b.000n u 0.N 0.?0uu 0.0 ".000u 0.0—mu 0.0 0.00'n n 0.! 0.000 “.0 0.00?« n 0.x! 6.000 0.0 “.00?" 0.! 5.0—0— 0.0 c.00Nn ~ 0.?! ”.mcn .n.w h.00Nu n 0.0! h.t0h D.u N.00Nu 0.0! n.0—n— 0.? N.00—u n v.0! 0.000 0.« 0.00“" n 0.0! 0.000 0.“! 0.000" 0."! 0.00v— n.n «.000u n n.0l 0.0n0 ‘0. c.0nou u 0.0! u.'m0 0.?! 0.070" 0.“! n.0nn— 0.N 0.600 a 0.0! 0.00? m.N! 9.000 n 0.0! v.00? 0..! 0.000 0..! n.0—w— 0.0 —.000 u 0.0! n.v0n A 0.?! 0.000 n 0.0! 0..0n n.0l 0.005 m.—! 0.00ua n.n N.onh u 0.0! v.0nn 0.0! 0.0Vh n n.0l 0.0mm 0.0! n.0vn 0.—! n.0—0— 0.N 0.000 ~ 0.0! —.'0u 0.0! 0.000 n 0.x! 0.n0« 0.0! 0.000 0..! n.0h0 n.N «.00? u 0.n! 0.N' m.n~! n.0mv n 0.0! A.3? n.n«! 0.00' m.—! n.0—0 0.N 0.00? n 0.0! N.h—! . n.'~! N.0)n u 0.0! . n.hul 0.nul 0.000 0.": —.o~n 0.. 0.00“ n 0.0! 0.5"“! 0.n—! 0.00m n 0.0! 0.m——! 0.0a! 0.00N 0.N! H.000 n.— 0.00m — 0.0! 0.00—! 0.0! n.0vm — 0.0! v.00“! 6.0! «.0'N w.—! n.0«0 m.N «.00N u 0.N! c.0am! 0.?! u.oon n 0.N! 0.0—N! 0.0! 0.00" m.—! 0.00n 0.0 n.onu n 0.0! 0.00m! 0.! n.0v— n 0.N! 0.00N! 0.0! 0.0?“ m.N! 0.0—v n.N N. — 0.N! 0.0w'! 0.! 0.! u n.«! 0.0.?! n.m! 0.! !!!!!!!! ~!!!!!!!!~!!!!!!I!~!:!!!!!!~I!!!!II!~IIIIIIII—!!!I!!!!~!!!!!I!!~! ! u n allllllll > u x u > _ x u > u x ~ > n x n > u x u > u x «cum — aqua n black n hzomm n «<01 h «<00 u 5201; n .202; a x¢w¢ h ctwu a block H hzozu n 23¢ n N 2:“ n u 2:“ >¢aw>~ma whmzuzoo n wo430 . to 0 n uNum awhm 0 2:0 awaawpm .mmcmzuuamum Eu w mcwcwmucoo 3cm 50?: mummp mucmELo$gma mugsh Low m ucmoa-gmmm ucm < pcvoaupcogm mo mmumcmugoou cowpwmoa mesanu .w.m mpnmp HHHHHfl~HNHHHHHH~~HHHHHHHHHHH LIST OF REFERENCES 265 LIST OF REFERENCES Ambler, B., Harries, G. 0. 1980. Optical Ranging for Tractor Guidance. ASAE Paper 80-1558. ASAE, St. Joseph, MI. Bibbero, R. J., 1977. Microprocessors in Instruments and Control. John Wiley, New York. Busse, H., Coenenberg. H., Feldmann, F., and Crusinberry, T. F. 1977. The First Serial Produced Automatic Steering System for Corn Com- bines and Forage Harvesters. Proceedings of the International Grain and Forage Harvesting Conference. September 1977. pp.43-47. Ciarcia, S. C. 1980. Home in the Range -- an Ultrasonic Ranging System. Byte 5(11):32-58. November, 1980. Coad, C. A. Ruff, J. H., Coble, C. G. 1979. Microprocessor- Based Ultrasonic Height Controller for Sugarcane Harvesters. ASAE Paper 79-1571. ASAE, St. Joseph, MI. Collins, R.L., Hong, J.P. 1974. A Comparison of Tire Influences on Vehicle Handling. Proceeding of the third Conference on Vehicle System Dynamics, Held at Virginia Polytechnic Institute and State University. Blacksburg Virginia, August 12-15, 1974. Durstine, J.w. 1965. The Truck Steering System from Hand Wheel to Road Wheel. SAE paper No. 730039 SAE Transaction, Booklet No. SP-374. Ellis, J. R. 1969. Vehicle Dynamics. London Business Books Limited. Gross, T. A. 1978. Controlling with Ultrasonics. Machine Design 5(5). March 3, 1978, pp. 90-96. Grovum, M. A., Zoerb, G. C. 1970. An Automatic Guidance System for Farm Tractors. Transactions of ASAE 13(5):565-573, 576. Kirk, T. G., Krause, A. E. 1975. Swather Edge Guide Steering Control System. ASAE Paper No. 75-1029, ASAE, St. Joseph, MI. Shukla, L. N., Goering, E., and Day, C. 1970. Effects of Tractor Parameters of Automatic Steering. Transactions of ASAE l3(5):678-681. Smith, D. E. 1980. Electronic Distance Measurement for Industrial and Scientific Applications. Hewlett-Packard Journal 31(6):3-10, 19, June 1980 266 Smith, L. A., Schafer, R. L. and Bailey, A. C. 1979. Verification and Testing of Guidance Algorithms. ASAE Paper 79-1618. ASAE, St. Joseph, MI. Swisher, G. M., 1976. Introduction to Linear Systems Analysis. Matrix Publishers, Champaign, IL. Tennes, B. R., Burton, C. L., Levin, J. H. 1976. Concepts for Merchandizing High Density Orchard Fruit Culture. Transactions of the ASAE 19(1):35, 36, 40. Tennes, B. R., Burton, C. L. 1979. A Rapid Planting Method for Fruit Trees and Bushes. Transactions of the ASAE 22(4):699-701, ASAE, St. Joseph, MI. Tennes, B. R., Brown, G. K. 1981. Design, DevelOpment and Testing of a Sway-Bar-Shaker for Horticulture Crops--A Progress Report. ASAE Paper No. 81-1059, ASAE, St. Joseph, MI. Upchurch, B. L., Tennes, B. R., Surbrook. T. C. 1980. Development of a Microcomputer-Based Controller for an Over-the-Row Apple Harvester, ASE paper 80-1556, ASAE, St. Joseph, MI. Young. S. C. Schafer, R. L., Johnsn, C. E. 1980. A Microcomputer-Based Vehicle Guidance Controller. ASE Paper 80-1557. ASAE St. Joseph, eph.