1 ,. m5 gummfim. . .5
I .. n a ,I
WhifiSliEUIig ‘6'
.1. 323......Rnfis .

$3.»... ......
1.11.2... u .. 2a.} .43,

Heir

.1

ii .o 5‘l.w\
. .51}:
I i .

Vt . s
35...“: 1...;

fit.’5.:n P75 (9...
.Iqoi§§: '
Sins.
‘..v’1....im3uflm2 i
.gl‘xiii. {‘2‘
ifiwvitr i... 9’
. .. my. . .
v“ J1! ”IT
flu . in”? I”... 0w}...
.113

0‘
.8: 2, .3;
3

£32.11
9| . .

a: .

‘1 .-
nwvsfli :5... IR II...
HEMP}... Naipaul!"
.. 1%.”); ‘

2
.iI. .
a... 3

 

 

34.0
iii... k. .
. I at
3?. 3.. i53§3it
3'3. .vsi¢.\\a 2".
2 la.‘iu..(‘ 4.
II... n} 5: .
Int E “31.11 ‘0.
. it!
.36

 

.1:

a. .
3.1.5“ .
35“.“. t. it
zufiiiul u. hm; .-

A . . . .flufik 5.5.3“, UM. .i‘.

. IV. $.ul.|i$..fiҤ. A1.-
.l...N.D.:.L.IZb.‘1§I\'I‘
$1. .5311..in "M...

.
ltkflhns
V I..33n..¢..sz..: mama.
fin-lg "31‘3
4.3.. .35... t .3... .
tJ I3..ii“h¥l 3 ‘0‘ t. IO: Vlr“
.5... . .1. .u trad
z. “3......
u ha...

 

 

 

 

in:
#53:? 3“.

i
.13. 3.79
). .gt :32:
7 ‘v‘- I‘

 

7.;

g. ..\.~vk .1.
~ .

”33......
J.

S. n . g. A . .
, . . V 5:.

 

 

3, UBPARY “
0‘3 Michigan State
University

 

 

This is to certify that the
dissertation entitled

Z-SOURCE INVERTER BASED POWER CONDITIONING
SYSTEMS FOR PV POWER GENERATION

presented by

Yi Huang

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Electrical Engineerigg

 

 

 

V Major Professor’s Signature
“.242. 4- .200 ?
I

Date

 

MSU is an Affirmative Action/Equal Opportunity Employer

 

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE
FEB 1 2 2011

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K:IProj/Aoc&Pres/CIRCIDateDm.indd

 

Z-SOURCE INVERTER BASED POWERCONDITIONING
SYSTEMS FOR PV POWER GENERATION

By

Yi Huang

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Electrical Engineering

2009

ABSTRACT

Z-SOURCE INVERTER BASED POWER CONDITIONING
SYSTEMS FOR PV POWER GENERATION

By

Yi Huang

The world’s energy demand is increasing quickly. As photovoltaic system is one of
the most promising alternative energy sources in DG. The power electronics converters

based PCS in the PV system now becomes the key point in cost reduction.

The main tasks for inverter are draw the power out and injected a sinusoidal current
to the grid. To competitive with other power sources, the most effective way is to
develop inexpensive and reliable inverters. The aim of this thesis is therefore to develop

new topologies and control strategies for PV system.

This thesis presented a new power conditioning system based on Z-Source inverter
for renewable energy sources. The PV system performance depends not only on
temperatures and irradiations, but also on maximum power tracking function of PV
inverter. So it is important to verify the inverter system too. To help verifying the
inverter system, a dc-dc converter based PV simulator was proposed and implemented.
Also, both the stand-alone split-phase PV system and grid-connected three-phase PV
system were proposed and analyzed. The simulation and experiment results were

shown to verify the proposed system.

With unique X shape network, the Z-Source inverter can realize buck and boost
function without additional dc-dc converter. For grid-connect action, the proposed
harmonic injected unity power factor current control which also employed modified P&O
MPPT method are suitable for PV power applications. By utilizing the Z-Source
inverter, the volume, the cost, and the switching device count are minimized. Because
of the single stage operation, the efficiency of the system can be greatly improved. The
reliability can be enhanced greatly due to the shoot through states. With all these
advanced features, the Z-Source inverter based PCS is very promising for renewable

energy applications.

To my mother, father, and husband.

iv

ACKNOWLEDGEMENTS

Foremost, I am heartily thankful to my advisor, Dr. Fang. Z. Peng, who has supported
me throughout my thesis with his patience and knowledge. I attribute the level of my
Ph.D degree to his encouragement and effort and without him this thesis would not have
been completed. One simply could not wish for a better advisor. This dissertation
comes from numerous discussions with him, under his keen insight and guidance. I
would also like to thank all my committee members, Dr. Strangas, Dr. Schlueter, and Dr.
Promislow. Their insightful comments and suggestions have enhanced the technical
soundness of this dissertation. I am grateful to my friends and colleagues from the
Power Electronics and Motor Drive Laboratory. I would like to express my
appreciation to Yuan, Dr. Lee, Miaosen, Richard, Lihua, Wei, Joe, Dong, Honnyong, Dr.

Y00, and all the lab mates in the Power Electronics Lab.

My thanks go to my family, especially my husband and my parents. Without their

years of encouragement and continuous support, I would not have reached this point.

TABLE OF CONTENTS

 

 

 

 

LIST OF TABLES ----.-ix
LIST OF FIGURES ................................................................................. x
CHAPTER 1. INTRODUCTION -- - -_l
1. 1. BACKGROUND ......................................................................................................... 1
1. 2. DISTRIBUTED GENERATION AND PHOTOVOLTAIC GENERATIONS .......... 2
1. 3. POWER CONDITIONING SYSTEM ........................................................................ 3
l. 4. OUTLINE OF THE THESIS ...................................................................................... 4
CHAPTER 2.- PHOTOVOLTAIC SYSTEMS - ........ -- - - 6
2. 1. INTRODUCTION ....................................................................................................... 6
2.1.1 History of Photovoltaics and new trends ........................................................ 6

2.1.2 Basics of Photovoltaics ................................................................................... 8

2. 2. TYPES OF PHOTOVOLTAIC SYSTEMS [15-18] ................................................. 11
2. 3. INVERTERS IN PHOTOVOLTAIC SYSTEM ....................................................... l4
2. 3. 1 Introduction ................................................................................................ 14

2. 3. 2 Types of Inverters ....................................................................................... 16
CHAPTER 3. SUMMARY OF PREVIOUS TOPOLOGIES - ..... - 18
3. 1. INTRODUCTION ..................................................................................................... 18
3. 2. PV INVERTER STATUS [15-19] ............................................................................ 19
3. 3. PV INVERTER BY POWER STAGE [20-3 8] ......................................................... 20
3. 3 .1 Introduction ................................................................................................ 20

3. 3 .2 Single-stage Inverters ................................................................................. 22
3.3.2.1. Four switch topologies .................................................................................. 22
3.3.2.2. Six switch topologies .................................................................................... 25

3. 3 .3 Multi-stage Inverters ................................................................................... 26

3. 4. INDUSTRIAL PV INVERTER [37-53] ................................................................... 33
3. 4. 1 Introduction ................................................................................................ 33

3. 4. 2 Manufacture PV Inverter ............................................................................ 33

CHAPTER 4. Z—SOURCE INVERTER BASED POWER CONDITIONING

 

SYSTEM FOR STAND-ALON E SYSTEM 47
4. 1. INTRODUCTION ..................................................................................................... 47
4. 2. THE BASICS OF PCS FOR RESIDENTIAL USE .................................................. 48
4. 3. THE PROPOSED Z-SOURCE INVERTER SYSTEM ............................................ 52

4. 4. CONTROL AND OPERATION PRINCIPLE .......................................................... 54
4. 5. DESIGN GUIDELINE .............................................................................................. 57
4. 6. SIMULATION AND EXPERIMENT RESULTS .................................................... 59
4.7. SUMMARY ................................................................................................................ 78
CHAPTER 5. PV SIMULATOR SYSTEM .................................................................. 79
5.1. INTRODUCTION ...................................................................................................... 79
5.2. PREVIOUS PV SIMULATOR .................................................................................. 80
5.3. PROPOSED PV SIMULATOR ................................................................................. 84
5. 3 .1 Mathematical modeling of PV cell ............................................................. 84
5. 3 .2 Control strategies ........................................................................................ 87
5. 3 .3 Hardware designs ....................................................................................... 93
5. 3 .4 Experiment results ...................................................................................... 94
5.4. SUMMARY ................................................................................................................ 99

CHAPTER 6. GRID-CONNECTED Z-SOURCE INVERTER WITH PV

 

SIMULATOR SYSTEM ......... - - - -- ............ - ....... ..-101
6.1 . INTRODUCTION .................................................................................................... 101
6. 1 .1 Background ............................................................................................... 101

6.2. PROPOSED GRID-CONNECTED Z-SOURCE INVERTER SYSTEM ............... 102
6.3. CONTROL ISSUES ................................................................................................. 105
6. 3 .1 General issues ........................................................................................... 105

6. 3 .2 Voltage control and current control .......................................................... 106

6. 3 .3 Modulation methods ................................................................................. 106

6. 3 .4 MPPT methods ......................................................................................... 109

6. 3 .5 Synchronization in grid-connected applications ....................................... 112
6.3.5.1. Synchronization methods ........................................................................... 1 12

6.4. PROPOSED PLL SYNCHRONIZATION .............................................................. 114
6. 4 .1 Proposed PLL synchranization ................................................................. 114

6. 4 .2 Proposed harmonic injected feed-forward control ................................... 115
6.4.2.1. Third harmonic injection ............................................................................ 116
6.4.2.2. Current control loop .................................................................................... 126

6. 4 .3 Modified P&O MPPT control .................................................................. 132

6.5. HARDWARE AND SOFTWARE ........................................................................... 136
6. 5 .1 Hardware of system setup ......................................................................... 136

6. 5 .2 Hardware of control unit ........................................................................... 138

6. 5 .3 Software implementation .......................................................................... 140

6.6. SIMULATION AND EXPERIMENT RESULTS ................................................... 146
6.7. PCS REQUIREMENTS AND STANDARDS ......................................................... 156
6.8. CONCLUSIONS ...................................................................................................... 158
CHAPTER 7. CONCLUSIONS _ _ - ..... ...... - - ...-159

 

vii

7.1. SUMMARY OF CIRCUIT TOPOLOGIES AND THEIR SUITED APPLICATIONS

 

......................................................................................................................................... 159
7.2. CONCLUSIONS ...................................................................................................... 162
7.3. CONTRIBUTIONS .................................................................................................. 163
7.4. RECOMMECDED FUTURE WORKS ................................................................... 164
APPENDICES“--. ....... - - --.....165
A.l. HARMONIC DISTORTION FACTOR OF THE CURRENT RIPPLE ................. 165
A.2. SOME EXPERIMENT HARDWARES .................................................................. 168
A3. FUNCTIONAL BLOCK DIAGRAM OF 2407A DSP CONTROLLER ............... 170
REFERENCES - _ _- - - - - - - - 172

 

viii

LIST OF TABLES

Table 3.1: Manufacture PV inverters (Xantrex, Ballard, Beacon,Phoenixtex power ........ 44

Table 3.2: Manufacture PV inverters (Fronius, studer, sma) ............................................. 45

Table 3.3: Manufacture PV inverters (Magnetek, PV Powered, Solectria, Xantrex). . .....46

Table 5.1: The PV array parameters (SM-60 Module)86

Table 6.1: shoot-through interval, boost factor and voltage gain of three boost control

methods ............................................................................................... 117
Table 7.1: Summary of central, string, and module integrated inverter .................... 160
Table 7.2: Summary ofthe inverter topologies. . 161

ix

Figure 1.1:

Figure 2.1:

Figure 2.2:

Figure 2.3:

Figure 2.4:

Figure 2.5:

Figure 2.6:

Figure 2.7:

Figure 2.8:

Figure 2.9:

Figure 3.1:

Figure 3.2:

Figure 3.3:

Figure 3.4:

Figure 3.5:

Figure 3.6:

LIST OF FIGURES

Major types of power electronics devices ........................................ 2

Evolution of the solar electrical capacities till 2030 [14] ..................... 7
I-V characteristics with constant temperature ............................................. 9
P-V characteristics with constant temperatur .................................... 9
I-V characteristics with constant irradiation .................................... 10
P—V characteristics with constant irradiation ................................... 10
Maximum power point tracking ................................................. l 1
An Off-grid PV system ............................................................ 12
A grid-connected PV system ....................... . ............................ 13
A hybrid PV system ............................................................... 14
Line-frequency transformer ...................................................... 20
High-frequency transformer 1 ................................................... 20
High-frequency transformer 2 ................................................... 21
Centralized and decentralized inverter topologies [21] ...................... 21
A single stage inverter ............................................................ 22
F our-switch boost inverter by Céceres and Barbi [22] ....................... 23

Figure 3.7: Four switch buck boost inverter by Vazquez et al [23] ....................... 24

Figure 3.8: Four switch buck boost inverter by Kasa et al [24] ........................... 24

Figure 3.9: Four switch isolated bidirectional buck boost inverter by Kjasr and

Blaabjerg [25] .................................................................................... 24
Figure 3.10: Four switch resonant buck boost inverter by Wang [26] .................. 25
Figure 3.11: Six switch buck boost inverter by Kusakawa et al [28] .................... 25
Figure 3.12: Dual power processing inverter ................................................ 26
Figure 3.13: Two stage boost inverter [29] ................................................... 27
Figure 3.14: Two stage nonisolated buck boost inverter by Saba and Sundarsingh

[30] ............................................................................................................................... 28
Figure 3.15: Two stage isolated buck boost inverter by Saba and Sundarsigh [30]. . .28
Figure 3.16: Buck-boost inverter by Funabiki et a1 [31] ................................... 28
Figure 3.17: A flyback inverter with enhanced power decoupling by Shimizu [32]...29
Figure 3.18: Multi-stage boost inverter with pseudo-dc-link .............................. 30
Figure 3.19: Multi-stage boost inverter by GEC [33] ....................................... 30
Figure 3.20: Bidirectional dc-ac-ac converter by Beristain et al [40] .................... 31
Figure 3.21: Cascaded inverter system [54] .................................................. 32
Figure 3.22: Xantrex PV series topology [42] ............................................... 34

xi

Figure 3.23: Ecostar Power Converte [44] ................................................... 35
Figure 3.24: Smart power M5 power system [45] .......................................... 36
Figure 3.25: Sunville inverter topology [46] ................................................ 37
Figure 3.26: FRONIUS IG 2000-LV [47] ................................................... 38

Figure 3.27: A] series inverter [48] .......................................................... 39
Figure 3.28: Sunny Boy 1100U inverter [49] ............................................. 40
Figure 3.29: Sunny Central [49] .............................................................. 40
Figure 3.30: AURORA isolated outdoor model : PVI 2000-I-OUTD-US [51].... . ....41

Figure 3.31: Starinverter PCS [52] ........................................................... 42
Figure 3.32: PVIl3KW inverter system [53] ................................................ 43
Figure 4.1: The traditional PV systems ....................................................... 49

Figure 4.2: Direct PV inverter systems for i120 V split phase residential power.... . .50 I

Figure 4.3: The six-switch inverter control scheme ........................................ 51
Figure 4.4: The proposed PCS for stand-alone PV system ................................ 53
Figure 4.5: Sketch map of the simple boost control ........................................ 55
Figure 4.6: Typical V-I and P-V characteristics ............................................. 60
Figure 4.7: V-I characteristics under different conditions ................................. 61

xii

Figure 4.8: A 10 kW Z-Source inverter prototype .......................................... 62

Figure 4.9: Simulation results of case 1 under the conditions of 1000 W / m2 , 60

°C, and va=230 V64

Figure 4.10: Experimental results of case 1 under the conditions of 1000 W / m2 , 60
°C , and va =230 V ............................................................................ 67

Figure 4.11: Simulation results of case 2 under the conditions of 250 W/ m2 , 0 °C ,
and va=450vn

Figure 4.12: Experimental results of case 2 under the conditions of 250 W / m2 , 0

°C , and va =450 V ........................................................................... 74
Figure 5.1: Typical PV module curves ....................................................... 80
Figure 5.2: PV simulator concept [5] .......................................................... 81
Figure 5.3: Block diagram of the PV generator simulator circuit [63] ................... 82
Figure 5.4: The circuit of the simulator with a light emission unit [65] ................. 83
Figure 5.5: An equivalent circuit of a PV cell ............................................... 84
Figure 5.6: The output characteristic of the SM-60 module ............................... 86
Figure 5.7: The PV simulator output characteristics ........................................ 87
Figure 5.8: Proposed PV simulator and its control strategy ............................... 89
Figure 5.9: Analysis of voltage control ....................................................... 91

xiii

Figure 5.10: Combined voltage and current control divisions ............................. 92

Figure 5.11: PV simulator for resistor load conditions ..................................... 94
Figure 5.12: Load current change ............................................................. 95
Figure 5.13: Solar irradiation change around load R= 18.5 ohm ......................... 97
Figure 5.14: Solar irradiation change around load R= 18.5 ohm ......................... 98

Figure 6.1: Traditional medium power stage grid-connected PV inverter system. . .. 103

Figure 6.2: Grid—connected Z—Source PV inverter system configuration. . . . . . . 104

Figure 6.3: SPWM method ................................................................... 108
Figure 6.4: Hysteresis current control ....................................................... 109
Figure 6.5: P&O MPPT method .............................................................. 110
Figure 6.6: INC MPPT method ............................................................... 111
Figure 6.7: PLL control loop ................................................................. 113
Figure 6.8: Zero crossing PLL ............................................................... 114
Figure 6.9: Phase detection methods ........................................................ 115
Figure 6.10: One-sixth of the fundamental injected ........................................ 116
Figure 6.11: Third harmonic injected maximum constant boost ......................... 118

Figure 6.12: The relation between the voltage gain and modulation index. . . . . 120

xiv

Figure 6.13: Z-Source inverter operation modes .......................................... 121
Figure 6.14: Relationship of the harmonic factor and modulation index with different
PWM ............................................................................................ 123
Figure 6.15: Relationship of the harmonic factor with modulation index and dc bus
voltage with and without harmonic injection ............................................... 124
Figure 6.16: Grid-connected three phase Z-Source inverter system .................... 127
Figure 6.17: Block diagram ofcontrol loop............... . .. 128
Figure 6.18: Open loop bode plot ............................................................ 130
Figure 6.19: PI correction bode plot ....... , .................................................. 131
Figure 6.20: Close loop system bode plot ................................................... 131
Figure 6.21: PV I-V and constant power curve ............................................ 133
Figure 6.22: PV I-V and constant power curve analysis .................................. 134
Figure 6.23: The block diagram of the MPPT control .................................... 135
Figure 6.24: The modified P&O MPPT method ........................................... 136
Figure 6.25: Universal DSP 2407A board ................................................. 139
Figure 6.26: Control Unit d1agramsl40
Figure 6.27: Interrupt handler flow chart ................................................... 142
Figure 6.28: Flowchart of DSP code ........................................................ 143

XV

Figure 6.29: CPLD control block ............................................................ 145
Figure 6.30: PV output characteristic during irradiation change ........................ 147
Figure 6.31: The grid connected z-source inverter output current ...................... 148
Figure 6.32: The grid line to line voltage and z-source inverter line to line voltage. 149

Figure 6.33: The experiment results of power curve changed with voltage ........... 149

Figure 6.34: Simulation results of va is around 400 V ................................ 151

Figure 6.35: Experiment results of va is around 400 V152

Figure6.36:Simulationresultsof va isaround 330 V153

Figure 6.37: Experiment results of va is around 330 V ............................... 154

Figure 6.38: Simulation results of va is around 230 V155

Figure 6.39: Experiment results of va is around 230 V156

Figure A. 1: Ripple current for two phase legs of three-phase inverter.... . ... ...........165

Figure A.2: PV simulator ........................................................................................ 168
Figure A.3: Z-Source inverter. . . .. ............................................................................. 168
Figure A.4: Part of grid-connected Z-Source inverter with PV simulator system ..... 169
Figure A.5: Functional block diagram of 2407A DSP board ..................................... 170

xvi

CHAPTER 1. INTRODUCTION

1. 1. Background

When the energy crisis in the seventies was going on, photovoltaic became more and
more popular as a substitute for fossil energy source. The worldwide improving
environment awareness and the needs of the growing population in developing countries
have increased the interest in photovoltaics as a long term, inexhaustible,
environmentally friendly and reliable energy technology. Photovoltaic sources are well
established in the alternative energy market. Also it is growing an average rate of 26%

per annum [98].

In the photovoltaic systems, power electronics converters are the key enable parts.
Energy is transferred from the PV, through power electronics converter, then finally to
the load or grid. The power electronics converter will regulate the power and voltage,
also realize maximum utilization of the energy fiom the PV. With the rapid improvement
in material science, the cost of photovoltaic cell has dropped tremendously in last two
decades, while as the price for the converter used in PV system remains almost the same.
Thus, to lower the cost and at the same time achieve the best performance of the power
electronics converter in the PV system now becomes the key issue for applications of PV

system.

Generally, by substitute old power processors with Power Electronics converters will
help a lot in system efficiency and dynamic response. The earliest power electronics can
be dated back to 1948, from which Bardeen, Bratain, and Schockley invented the silicon
transistor at Bell Telephone Laboratories. Since then, many new power semiconductor
devices have been invented and evolved for control of power and energy. Figure 1.1
shows the rating and application range of major devices used in power conversions. For
PV system, IGBTs and GTOs are most commonly used devices because of the voltage

and current ratings.

 

 

 

 

 

 

 

 

Current
1k - I
‘GTO
500 ' IGBT SCR
200 "'
MOSFET Voltage
. >
200 2k 5k

Figure 1.1: Major types of power electronics devices [19].

1. 2. Distributed Generation And Photovoltaic Generations

Based on above introduction, the extension of distributed generation (DG) is widely
used to avoid the energy exhaustible from only the fossil fuel supply. Distributed

generation, also called on-site generation, generates electricity from many small energy

sources typically in the range of 3 kW to 10,000 kW. With the widely used distributed
generation system, consumers and power utilities can get much more benefits.
Distributed generation can give more varied energy options, increase generation and
transmission efficiency, and improve power quality and system stability, etc. The types
of DG include Gas Turbines, Reciprocating Engines, Microturbines, and as well as the

"green powers", such as fuel cell, wind turbine and photovoltaic (PV).

Among those green powers, photovoltaic has its mrique advantages than other sources.
Since the generation of electricity is directly fi'om the sun, no fuel is needed. The
production of electricity by the photovoltaic process is clean and produces no carbon
dioxide or other toxic fumes. That’s non-polluting energy. The PV system has a high
reliability with at least 20 years of service time. The operating costs is low, because there
is no moving parts, the PV cells need little upkeep. Photovoltaic system also has a good
modularity due to their portability and sizebility. Usually photovoltaic system is near the
point where the electricity is used, thus the wires connection can be reduced and the same
to the construction time, the total construction costs can be reduced. Based from [101],
the PV panel production cost is $0.99 to 2.00/W (2007) plus installation and supporting
equipment unless the installation is Do it yourself (DIY) bring the cost to $6.50 to 7.50
(2007). As the price of PV cell itself is still decreasing, the PV system is very promising

for DG.

1. 3. Power Conditioning System

In nowadays distribution systems, many different types of DG fiom under 10 kW to

tens of megawatts generation are located around communities and industrial facilities.

When the power source of the DG does not meet the grid or load requirements, power
converter is required to bridge between the source and load to act as the Power
Conditioners.

Generally, a Power Conditioner for PV system will need to meet the following
requirements:

1) Power conditioner system can process power conversion from dc voltage into ac
voltage with the required frequency. The dc voltage can be higher or lower than the ac
voltage.

2) The power quality must satisfy the low total harmonic distortion (THD), voltage
and frequency deviation requirement.

3) Power conditioner system has protection for electric power systems, and some

also can have anti-islanding protection and electrical isolation if necessary.

1. 4. Outline Of The Thesis

The goal of this thesis is mainly to research power conditioning system for PV.
The previous topologies for PV inverter system were summarized and compared. In
addition, a novel Z-Source power conditioning system is proposed for split phase stand
alone PV system; also, the grid-connected three-phase Z-Source inverter system is
proposed and analyzed. For test purpose, the dc-dc converter based PV simulator system
is proposed and built. The outline of each chapter is listed as the following:

Chapter 2: Introduces the Photovoltaic system as the background for the whole
thesis.

Chapter 3: Reviews the inverter topologies for photovoltaic power conditioning

system. Compares of those topologies are presented. The advantages and disadvantages
for different topologies are also summarized.

Chapter 4: Proposes a novel Z-Source power conditioning system for split-phase
stand alone residential photovoltaic system. The design and control is analyzed. The
simulation and experiment results are presented to verify that.

Chapter 5: Proposed a dc—dc converter based PV simulator for Z-Source inverter
system for test purpose. A new combined voltage and current control method is proposed
and tested. The experiment results are shown for prove.

Chapter 6: Proposed a grid-connected three-phase Z-Source inverter with PV
simulator system. The design and control is pr0posed and discussed in detail. The
simulation and experiment results are shown to verify the proposed system and control
strategies.

Chapter 7: Summarizes the inverter topologies, and makes a conclusion for the
whole thesis. The contribution of the whole thesis is listed, and the future work is also

recommended.

CHAPTER 2. PHOTOVOLTAIC SYSTEMS

2. 1. Introduction

2.1.1 History Of Photovoltaics And New Trends

Renewable energy sources derive their energy from existing flows of energy, from
on-going natural processes, such as sunshine, wind, flowing water, and geothermal heat
flows. Now the most feasible alternative energy sources include solar power, firel cell,

and wind.

The energy available at the surface of the sun is 60,000 kW/ m2 , where the sun’s

radiation at the top of the earth’s atmosphere is only about 1.4 kW/ m2

. After it passes
through the atmosphere, about 80 trillion kW of solar radiation energy is available
globally. This is about 13,000 times the present world energy use [100].

The history of PV’s dates back to the early 19th. In 1839, Edmund Becquerel, a
French experimental physicist, discovered the photovoltaic effect while experimenting
with an electrolytic cell made up of two metal electrodes [12]. In the early 1950’s, the
Czochralski meter was developed for producing highly pure crystalline silicon. In 1954
Bell Telephone Laboratories produced a silicon PV cell with a 4% efficiency and later

achieved 11% efficiency. Then with the development of the semiconductor technology

and PV module manufactory, now photovoltaic has a wide range of applications.

From the US PV roadmap [13], the goal of the industry is to meet 10% of US peak
electricity generation capacity by 2030. Figure 2.1 shows the different projections of the
Japanese, US and EPIA roadmaps [14]. PV generation shows promising in the near
future.

Now there are several major market players in US. For BP-solar, it was the number
two to sell 58 MW in 2001. Then there is shell solar, shell solar sold 44.4 MW in 2001.
And shell solar will invest 0.5 to 1 billion dollar in solar photovoltaics and wind energy
from 2001 to 2006. Astropower sold 26 MW in 2001. And there are also ASE-Americas,

United Solar Systems, Evergreen Solar, etc [14].

\,~

 

 

 

M“
140000 / III USA
, ‘ Europe
120000 , Japan
7 ~ 4 Wold
100000 , r __

80000

60000 '

 

40000

 

Install Capacities (MW)

20000

0

 

Figure 2.1: Evolution of the solar electrical capacities till 2030 [14].

(Sources: Japanese, US and EPIA Roadmap)

2.1.2 Basics Of Photovoltaics

The basic element of PV system is PV cell. PV cell can produce electricity due to
quantum-mechanical process. Usually a PV cell consists of a p-n junction formed in a
semiconductor material similar to a diode. If light is incident on a PV cell, current will
flow when an electrical load is connected.

The characteristics of a PV cell are nonlinear. They depend on solar irradiation and
cell temperature. For most crystalline silicon solar cells the reductions in voltage with
increasing temperature is around 0.50 %/ °C , though the rate for the highest-efficiency
crystalline silicon cells is around 0.35 %/ °C . Averagely, the rate for amorphous silicon
solar cells is 0.20-0.30 %/ °C , depending on how the cell is made [99]. Figure 2.2
shows I-V characteristics with constant temperature condition, and Figure 2.3 shows P-V
characteristics with constant temperature too. Similarly, Figure 2.4 shows I-V
characteristics with constant irradiation, while Figure 2.5 shows P-V characteristics with
constant irradiation. While for each curve there is only one operation point for maximum
utilization of power as shown in Figure 2.6. Thus maximum power point tracking
(MPPT) is used to control output power. There are several different methods to realize

MPPT.

 

G=1000W/m2

 

 

   

G=800W/m2

   

600w/m2

G:

 

 

 

 

 

4 3 2

8 5:5

Voltage (V)

Figure 2.2: I-V characteristics with constant temperature.

 

 

 

 

H_ u u - I
. q _ . _ . _ .
_ _ _ . _ . _ .
_ . _ . _
_ _ _
11.11 1111 1 11 1 1 111
_ _ . _ . _ _ _
. _ _ _
_ _ _ . . .
. . — ._ . . _
_ . _ _ . _ _
. ._ _ _ _ . .
IITI TII. ILII+IIT ITILIII
_ . _ _ _ .
_ . . _ . _
_ . . . _ _ _
_ _ _ _ _ . .
_ _ . _ . . _
_ . _ . _ . _
.llqulqltJll II I l.ll.111
_ _ _ _ _ . _
_ . _ . .
_ _ G _ .
_ _ g _ _ . .
_ . .m _ . _ .
31..--.. a 1.1L--_. 1. 1-
_ . e _. _ _
_ . a . _ _ _
_ _ n _ _ _ _
_ . I _ . . _
_ _ _ _ _ .
_ _ . . . . _ .
b i—r L p _ - — _
m m m m m .o. .0. m m

36 coach

Voltage (V)

Figure 2.3: P-V characteristics with constant temperature.

 

 

 

 

 

 

 

 

 

_
.
151111
.
Yfi
\nfll\\+\|l|l|l
” n
_ .
1r1 h111-
_ _
_ _
_ _
_ _
_
.
m m
xwmw “
w u
. _
_ _
411141-11fi1114111-
_ _
. .
. .
_ .
_ _
2 1
A<V 30:30

15 20

1 0
Voltage (V)

Figure 2.4: I-V characteristics with constant irradiation.

 

 

q——-'

_.._.__..__r___._..._'__

.._____L_.....
t
I
L.

- — - Increasing Tc

 

 

90

80-------

50-------
40
30-————--

05v coaom

Voltage (V)

Figure 2.5: P-V characteristics with constant irradiation.

10

 

 

\

 

 

a
\.

E
5'

 

a
\
f
/

 

 

Solar Current (A)
-h Ur
O C
”an
/

 

 

 

 

 

 

 

 

 

 

\ .
:2 / 1
,0 / l
O l
0 5 10 15 20

Voltage (V)

Figure 2.6: Maximum power point tracking.

2. 2. Types Of Photovoltaic Systems [15-18]

According to the different operation requirements, the component configurations, and
ways to connect the power source and loads, photovoltaic systems are generally classified
to three types: off-grid system (or stand-alone system), grid-connected system (or utility-

interactive system), and hybrid system.

Off-grid systems can also be classified to off-grid domestic system and off-grid non-
domestic system. Off-grid domestic systems provide the electricity for low power loads
(such as lighting), which used by households and villages that are not connected to the
utility grid. Systems of this type have been widely used to satisfy the energy demands of

off-grid communities. The system size is usually around 1 kW. Off-grid non-domestic

11

systems provide power for a wide range of applications, such as water pumping,
telecommunication, etc. Figure 2.7 shows the general off grid structure of PV system. In

this system, a energy storage device, such as a battery pack, is usually required to supply

DC
Load

PV Charge — ' AC
_’ .
Generator Controller {\’ Load

Inverter/
Power
Conditioner

transient power at the peak load.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I [(191123 1L

 

 

Figure 2.7: An Off-grid PV system.

Similarly, grid-connected systems can be classified to grid-connected distributed
system and grid-connected centralized system. Grid-connected distributed systems
provide electricity to a building or other loads that are connected to the utility grid.
These systems are widely used and account for most of the installed PV capacity.
Generally system capacity varies fiom 1 kW to 100 kW for different requirements.
When on-site generation exceeds the need of loads, the excess energy can be fed back to
the utility grid. With grid connected PV system, the distribution loss can be reduced

because the systems are installed at the point of user. Since no extra land is required for

12

PV systems, costs for mounting systems can be reduced. One distinguish feature of PV

system is that PV array can be used as roofing material as BIPV (building integrated PV).

Compared with off-grid installation, grid connected PV systems cost less because the

energy storage is generally not required. Figure 2.8 shows the grid connected structure of

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PV system.
DC Bus AC Bus
, DC
PV __, Load
Generator
~——/—e
: Grid
Battery ‘-—-F‘ [\1 AC
Inverter/ Load
Power
Conditioner

Figure 2.8: A grid-connected PV system.

Besides the aforementioned two types of PV systems, there is an alternative solution
which is called as Hybrid PV system as shown in Figure 2.9. The hybrid system can
either works as a stand alone system or a grid connected system. It provides the end user
more options as the penalty of higher cost. Hybrid systems can be classified to three
types: series hybrid system, switched hybrid energy system, and parallel hybrid energy

system. Usually hybrid PV systems contain at least one other generation source such as

13

wind turbine, which is not a utility. Those systems can provide continuous power

whether the PV can provide enough power or not.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DC
Load
PV Charge Battery
Generator Controller
3 AC
[\1 Load
Engine
Generator, ’\/ Inverter/
wind turbine, — Power
grid backup Conditioner
Rectifier

 

 

 

 

Figure 2.9: A hybrid PV system.

2. 3. Inverters In Photovoltaic System

2.3.1 Introduction

Inverter is an electronic device that converts dc electricity to ac electricity. Because
solar panels and batteries can only generate and store dc electricity, while the appliances
people used from the utility grid need ac electricity, the inverter is the key part in the

power processing unit in PV systems.

Power Electronics inverters convert dc electricity to ac electricity by switching
mechanisms to break the continuous current into pulses. Nowadays, inverters can be

made small enough to along with laptop for working on the road, and large enough to

send megawatts of renewable energy into the utility grid. Different inverter has different
configuration, power output quality and expected service life.

The PV inverter is a costly and complex component of PV systems that convert dc
power to ac power from the PV modules. Inverter mean time to first failure (MTFF) is
around five years [19]. To minimize inverter cost and at the same time improve inverter
performance and reliability is the primary goal of PV manufactures. Since the inverter is
not only used in the PV system, but also used for other Distributed Energy Resources
(DER), such as wind, fuel cells, etc. The inverter requirements are similar that universal
designs with replaceable modules are the best solutions for a ‘next generation’ inverter
[19]. For all the residential grid-connected DG types, the inverter should be able to
convert between 2 and 10 kW of dc power, prevent islanding, reduce radio frequency
interference, provide the low harmonic distortion, and optionally provide backup power.
Therefore, the residential grid-connected inverters for solar, wind, and fuel cells can be
nearly identical.

Inverter technology is the key technology to have reliable and safety operation of PV
system. It is also required to generate high quality power to ac utility system with
reasonable cost. By means of high frequency switching of semiconductor devices with
Pulse Width Modulation-PWM methods, high efficiency conversion with high power
factor and low harmonic distortion power can be achieved. The remaining problem is the
cost.

In a PV system, the inverter has to realize the following firnctions:

1. convert dc power to ac power usable by most business and household

devices

15

2. modify the voltage and current output from the PV array

3. converter the ac power to the specific voltage and frequency range of utility
network

4. safeguard the utility network system and its personnel fiom possible harm
during repairs

5. prevent damage to the PV array and other component during unusual
operating conditions

Major design considerations for inverters include their capacity, voltage rating and

battery capability. In addition, the high operating voltage minimizes transmission losses.
2.3.2 Types Of Inverters

As described before, there are various types of inverter configuration. Self
commutated voltage type inverter is employed in all inverters with a capacity of 1 kW or

under, and up to 100 kW.

The self commutated inverter has more advantages than the line commutated inverter.
The self commutated inverter can freely control the voltage and current waveform at ac
side, and adjust power factor and suppress the harmonic, and highly resistant to the
system disturbance. Therefore, most inverters for distributed power sources such as

photovoltaic generation now employ a self commutated inverter.

The self commutated inverter can be classified to voltage and current types. The
voltage type is a system in which the dc side is a voltage source and the voltage
waveform of the constant amplitude and variable width can be obtained at the ac side.

The current type is a system in which the dc side is the current source and the current

16

waveform of the constant amplitude and variable width can be obtained at the ac side. In
the case of photovoltaic power generation, the PV output is dc voltage, so the voltage

type inverter is employed.

Both the voltage control and the current control can be performed to the voltage type
inverter. The current control scheme is used more popular because a high power factor
can be obtained with simple control, and transient current suppression is possible during
disturbances, such as voltage changes in the utility power system. For the output ac

voltage control, general PWM control can be utilized.

17

CHAPTER 3. SUMMARY OF PREVIOUS
WORK

3. 1. Introduction

For growing PV market, there is a lot of occupation for grid-connected applications.
And around 95% are used for AstroPower domestic residential business [20]. Now the
residential system power rating ranges from 1.6-9.6 kW. And customer can also choose
2.4, 3.6, 4.8, 7.2 kW sizes due to their requirements. For commercial system, the power

rating can be 10-300 kW or more.

Grid-connected PV system has the advantage of more effective utilization of
generated power. According to [20], PV grid-connected inverter systems have good
performance, including high conversion efficiency and power factor exceeding 90% for
wide operating range, while at the same time maintaining current harmonics THD less

than 5%.

With the development of power electronics devices and the corresponding
technologies, the cost and size of PV inverter reduced a lot recently. The progress of
circuit design of inverter and control methods also helps a lot. The control circuit also

provides protection functions like the maximum power point tracking.

18

3. 2. PV Inverter Status [15-19]

In a PV system, the PV array output voltage cannot be applied to the load directly in
common. Thus the inverter is needed to process that. It is very important to make the

power meet the specific demand of the equipment.

There are various inverter employed in the PV system. Now for the PV inverter, the
input voltage is higher than ever, some up to 600 V. The single unit costs around $0.6-
$1.0/watt. The inverter efficiency is around 85% in average. The system becomes easier

to install with higher dc voltage input.

Inverter can be operated for normal fluctuations of voltage and frequency at the utility
grid side. The standard voltage and frequency for a single phase circuit is 120/240 V and
60 Hz in US, for a three phase circuit is 480 V and 60 Hz. For the standard values, the
inverter can operate substantially without any problems within the tolerance of 10% and

—15% for the voltage, and 21:04 to 1% for the frequency.

The operable dc voltage range is determined by the rated power of the inverter, rated
voltage of the ac utility grid system. For a capacity of 1 kW or below, the operable dc
voltage ranges from 14-25 V, 27-50 V, 45-100 V, 48-120 V, and 55-110 V. In addition,
for a capacity of 1 kW to 10 kW, the operable dc voltage ranges from 40-95 V, 72-145 V,

75-225 V, 100-350 V, 125-375 V, 139-400V, 230-450 V, 250-600 V, and 450-800 V.

The harmonic current from the inverter is very small, because the PWM control
scheme is used as the output waveform control of the inverter. Thus fewer problems

arise.

19

The developments of PV system need more requirements for the inverter. For the
future application, inverters maybe need utility interactive and stand alone capability,

more reliability, longer life cycle, and lower failure probability.

3. 3. PV Inverter By Power Stage [20-3 8]

3.3.1 Introduction

Some power conditioning systems employ transformers in the topologies. Figure 3.1
shows that a line-frequency transformer (LFT) is located between the grid and inverter.
This transformer can avoid the injection of dc current into the grid. Figure 3.2 shows that
a high-frequency transformer is located in an HF-link grid-connected ac-ac converter.

Figure 3.3 shows that a high-frequency transformer is located in a dc-link PV-module-

t e

-— LF

connected dc-dc converter.

 

 

 

DC
PV

 

 

 

 

 

 

Figure 3.1: Line-frequency transformer.

 

 

Grid

 

 

 

 

 

 

 

 

 

Figure 3.2: Hi gh-frequency transformer 1.

20

 

 

AC

 

 

 

 

 

 

 

 

 

 

 

Figure 3.3: High-frequency transfonner 2.

For PV inverters, there also can be classified to the centralized inverter and
decentralized inverter. Figure 3.4 shows the centralized and decentralized inverter
topologies in [21]. Centralized inverter only needs only one inverter for the power
conditioner, but decentralized inverter needs more than one inverter to connect PV
modules to a common bus (dc, low voltage ac, or the grid). Compared 'those two
topologies, the one central inverter has the advantages of higher efficiency, lower cost,
and without the considerations of shading effects. The low voltage ac bus is best

selection as the common bus employed in the decentralized inverter.

Decentralised

Grid C51? Grid
—1:—@

 

 

 

 

 

 

 

Central R
Inverter Common
' Bus
:-
| \.
nverters

Figure 3.4: Centralized and decentralized inverter topologies [21].

21

For different power stages, inverter systems can also be classified as single-stage
inverter and multi-stage inverter. Both of these two will be described in detail in the

following.
3.3.2 Single-stage Inverters

Single stage inverters only need one power stage to process power conversion from
dc to ac and boost the dc input voltage to required output voltage level as shown in
Figure 3.5. Single-stage inverters can be classified to four switch topologies and six
switch topologies. Single-stage inverters usually have a relative simple topology, and
employ the fewer components, thus cause a higher efficiency. The past literatures have

listed a lot of inverters for this type.

 

 

DC .
PV Grid

AC

 

 

 

 

 

 

Figure 3.5: A single stage inverter.

3.3.2.1. Four switch topologies

Figure 3.6 shows a nonisolated boost inverter proposed by Caceres and Barbi [22].
The circuit employs two dc-dc converters, and the load is between the two outputs. Each
converter provides a unipolar dc-biased sinusoidal output with a 180 degree phase

difference. Therefore, the output waveform is pure sinusoidal. Figure 3.7 shows abuck-

22

boost inverter proposed by Vazquez et al [23]. In this circuit, similarly to inverter in Fig.
3.6, two buck-boost dc-dc converters instead of boost converters are used in the system.
This system can generate output voltage lower or higher than the input voltage due to
buck-boost converter. Figure 3.8 shows a four switch buck boost inverter by Kasa et a1
[24]. The input PV voltage can have a wide operation range, but requires split dc voltage
source. A Four switch isolated bidirectional buck boost inverter is shown in Figure 3.9
by Kjar and Blaabjerg [25]. The advantage of this system is the galvanic isolation
provided by the two high frequency transformers due to the transformers. But the
transformer also will increase the system cost and lower the efficiency. Figure 3.10
shows a zero-current switching buck-boost inverter proposed by Wang [26]. The
advantage of this circuit is that the switching power loss can be reduced.

11031

.1 {A J {A
31 52

\I
\
II

S3 S4

 

 

 

 

Figure 3.6: Four-switch boost inverter by Caceres and Barbi [22].

23

 

 

J Clx V:—
31 I
Ll
C2»: C39]

 

 

 

 

S3 ] Load |-4

Figure 3.7: Four switch buck boost inverter by Vazquez et al [23].

 

 

 

 

 

 

 

I— —1 D1
V T1 T2 i: Grid
V1_-=r -- C1 C3
1?
VZE 2:
T C2 T3 T4
_1 I— D2
7 51 - 7L“—

 

Figure 3.8: Four switch buck boost inverter by Kasa et al [24].

 

 

 

 

 

 

 

 

 

 

 

] PWMl PWM2[

Figure 3.9: Four switch isolated bidirectional buck boost inverter by Kjaer and Blaabjerg
[25].

24

 

\I
II

< -11:
Q

 

 

 

 

Figure 3.10: Four switch resonant buck boost inverter by Wang [26].

3.3.2.2. Six switch topologies
Figure 3.11 shows a six switch isolated buck-boost inverter proposed by Nagao and
In this system, the energy storage inductor is charged from different

Harada [28].
directions in each half cycle, thus provide an alternative output [28].

 

 

 

 

 

 

 

 

 

 

 

31
SA ‘ “ 55
—1 L2
S2 m
' s5
J_ "'I
a 7‘ L1 C2__ C3__ -
V C1 % I‘" I‘\ (I? Gnd
532— 5.6
1—
s4 EL
54 ' N '

 

 

Figure 3.11: Six switch buck boost inverter by Kusakawa et a1 [28].

Single—stage inverters usually have higher efficiency and lower cost, but the power

capacity is limited, and dc source range is also limited.

25

3.3.3 Multi-stage Inverters

Multi-stage inverters have more than one stage to process power conversion as shown
in Figure 3.12. Dc-dc converters will perform usually voltage buck or boost or electrical
isolation, and then dc-ac can be performed to get the required output amplitudes and
waveforms. Multi-stage inverters can solve the problems when single-stage inverters

meet with high power, high performance requirement.

 

 

PV Grid

 

 

 

 

 

 

 

 

Figure 3.12: Dual power processing inverter.

A two-stage boost inverter is shown in Figure 3.13. In the circuit, a dc-dc boost
converter is ahead of the dc-ac inverter. With the first stage, input voltage is boost to
suited dc voltage, and output ac voltage can be obtained by a high frequency PWM
inverter through the second stage. The whole process flows as the dc-dc-ac process.
Figure 3.14 shows a two-stage nonisolated buck boost inverter proposed by Saba and
Sundarsingh [30]. This inverter was designed to operate a low dc input voltage as 100 V
due to safety reason, thus make the PV Operate voltage range very small. The isolated
buck-boost inverter with a high frequency transformer shown in Figure 3.15 can draw the
power fiom the PV source even when the distributed generated dc voltage is low. The
inverters operate as current source inverter in line frequency in Figure 3.14 and Figure

3.15. The inverter in Figure 3.16 does not have the high-frequency transformer, but

26

adding a capacitor to the charge loop of energy storage device. This inverter can have a
wide range of PV operate voltage. While, Figure 3.17 shows a flyback inverter with
enhanced power decoupling by Shimizu [32]. In this kind of inverter, the first stage has
the buck boost converter, the second stage has the flyback converter, and the last stage
has a transformer. The voltage across intermediate capacitor can have both the dc
component and ac component alternating at twice the frequency of the load. This system
can have a longer service life of inverter because the large electrolytic capacitor can be
replaced by the small intermediate capacitor. In paper [41], similarly topology has been
used. The first stage is dc-dc converter to boost voltage, and then the second stage
processes the dc-ac conversion. There are only four switches in the inverter operation in
line frequency, thus the dead time effects can be disregarded, and the line noise of
voltage and current caused by high frequency switching can be reduced. The above

several inverters all process as dc-dc-ac topologies.

Grid

<1|1k

 

 

 

 

 

 

 

Figure 3.13: Two stage boost inverter [29].

27

 

gllt

85 D2
:: C2 ;: 25 D3
C1 L1
D1 J
S6

 

 

 

 

 

 

 

 

 

 

 

L1
J 31
J SZJ
1:1
. 240v
;: C2:
D1 533 54-1

 

 

 

 

 

Figure 3.15: Two stage isolated buck boost inverter by Saba and Sundarsigh [30]

 

we

D1 A D2
SIJ SZJ

 

"(22¢

 

 

 

 

 

53:] S4]
D3 D4

28

Figure 3.16: Buck-boost inverter by Funabiki et a1 [31].

Sflybackl

   

 

1r

 
   

\/

Sac2
L.

J

    

Sbuck-boost

Sflyback2

Figure 3.17: A flyback inverter with enhanced power decoupling by Shimizu [32].

Figure 3.18 shows a multi-stage boost inverter with pseudo-dc-link. The circuit has a
PWM dc pulse train. Compared to traditional dc-link, the dc filter in the pseudo-dc-link
can be eliminated. The inverter operates as line-frequency switching inverter to get the
required output voltage from the high-frequency dc pulse train. Therefore, the losses and
cost can be reduced because not all stages work at high switching frequency. Figure 3.19
shows a multi-stage boost inverter proposed by GEC [33]. During the last stage, the line-
frequency inverter processes the half sine wave to full sine waves and produces ac current
to the grid in phase without output filter. This topology has been applied to General
Electric Company (GEC) for 10 kW grid-connected PV systems [39]. Those above two

are both dc-ac-dc-ac topologies.

29

 

HIIF

\I
II
O
)—A

 

 

 

 

 

 

S7 S8 L2

 

 

 

 

Figure 3.18: Multi-stage boost inverter with pseudo-dc-link.

 

illl

 

 

 

 

 

 

 

 

Figure 3.19: Multi-stage boost inverter by GEC [33].

Figure 3.20 shows a bidirectional dc-ac-ac converter proposed by Beristain et al [40].

In the topology, the second stage is a bidirectional ac-ac converter, and a high-frequency

transformer is also employed for voltage variation and electrical isolation.

30

 

 

III]

 

 

 

 

 

Figure 3.20: Bidirectional dc-ac-ac converter by Beristain et a1 [40].

Paper [42] also discussed the cascaded dc-dc converter for PV power system. Buck,
boost, buck-boost and (er converters are checked as possible cascaded converters. The
results show that the boost converter is the best if a specific boost is needed. While
flexible in voltage ranges, buck-boost and Ct'rk Converters usually have the less efficient
and the higher cost. Figure 3.21 shows a cascaded inverter PV system [54]. More than
one single-phase H-bridge inverters are in series. The cascaded inverter used ready-made
ICs for class D audio amplifier was realized. The ready-made integrated circuits (ICs)
have the advantages to reduce the parts of the inverter and mass production [54]. Usually,
the class D audio power amplifier uses the ready-made ICs, but ICs could not provide
utility grid level. Similarly, The PV inverter for grid connected by the ICs without LFT
meets the same problem. The cascaded inverter can solve the problem and reach high

voltage efficiently.

31

INVl

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1
DC/ 7 . 11 11
3‘
va1 DC
INV2
DC/ /\l
3:
DC
Vout=n(V1 1 - V12)
INVnV
DC/
3: DC

 

 

 

 

 

 

 

 

 

 

 

 

1c_sl

Figure 3.21: Cascaded inverter system [54].

For multi-stage inverter systems, the high frequency transformers are generally used
to provide electrical isolation and voltage boost in the front stage, while the line

frequency transformers are usually used in the last stage to limit the switching losses.

32

3. 4. Industrial PV Inverter [37-53]

3.4.1 Introduction

PV system has the numerous markets in the world. More than 1900 MW of PV were
installed worldwide. Japan has the highest installed capacity, followed by Germany and

the US. Those three countries represent about two thirds of global PV capacity [43].

3.4.2 Manufacture PV Inverter
° Xantrex

Xantrex’s inverter has a single stage employed with high efficiency transformer.
Figure 3.22 shows Xantrex PV series topology. A three phase bridge is used between the
PV array and a 60 Hz isolation transformer. The transformer steps the utility line-tie
voltage for more efficient power transmission. The circuit has the advantages that the
single stage makes the system more efficient, and much more utility grid-tie voltage
options can be selected due to the isolation transformer. The circuit has the disadvantage
that the overall system efficiency is lowed by using the transformer. The Xantrex
Technologies PV 10208 system uses this kind of topology inverter, and was tested at

Sandia National Laboratories. The efficiency is around 95%.

33

 

 

va [ DC contact

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

‘4" _ _ AC 1
' Va
i Ipv 4a *0 4a contact :1
i . IMM— ‘—-1 l—le- _.
E GUM)— .——-n..-.11-—.-1 h,
l I ‘----—-
1 t ”K? 14:1 {I i
l . . l
: N I ground 1G‘l‘attetd‘trrre srgnals E
= PV Inverter 1 vab’bc’ca
- ‘ " I
1) PV voltage regulatron a—-—tlt pv
: 2)AC current ‘——T ram d
I ‘ I
] 3)Islanding,etc. 1 groun
Figure 3.22: Xantrex PV series topology [42].
0 Ballard

Figure 3.23 shows the Ecostar Power Converter for PV applications. The topology is
similar to Figure 3.22. This inverter is a grid tie utility interactive inverter. The rated
voltage is 208 V or 480 V. The output frequency is 60 Hz. The PV array dc voltage
range is from 290-595 V. The current harmonic is less than 4%. PV array MPPT ranges
from 240 V to 600 Vdc. The peak inverter efficiency is 97% at 75% output. The peak
efficiency with transformer is 94.6% at 75% output [44]. This kind of inverter has been
installed in Chicago Center for Green Technology, Illinois. The number of kW hours

produced from a 76 kW array is around 13,274 kW hours.

34

To AC Disconnect

= 5 l = 30kVA
External Output
Transfomer

   
 

 

 

To DC
Disconnect

To Combiner{
Boxes

 

Site
Ground

Figure 3.23: Ecostar Power Converte [44].

° Beacon Power
The Smart power M5 inverter is used for grid-connected PV system with battery backup.
A 5 kW grid tie PV power system is shown in Figure 3.24. The system employs the
dc/dc converter before dc/ac inverter. The PV operating ranges from 48 to 110 Vdc, DC
input battery voltage is 44-60 Vdc. AC output voltage is between 106 to 132 Vac (120
Vac nominal). Output frequency is 60 Hz nominal. And the total harmonic distortion is

less than 3%. The peak efficiency is 93%.

35

  
  

100A DC/DC SkVA Inverter Inverter

 

 

 

 

   
   

 

 

 

 

 

 

 

output
— breaker
___.... IV
i AC
Battery bus transfer
Contactor Current contactor
/ shunt

 

 

 

 

 

 

I
I
-t —
+ __
e e

PV/DC svvitchgear AC switchgear

? : I Communications

Battery
temperature

sensor

 

 

 

 

 
  

 

 

 

 

 

 

 

 

Battery (48V)

Figure 3.24: Smart power M5 power system [45].

° Phoenixtec Power

Figure 3.25 shows Sunville grid tie inverter. For Sunville 1500, the output power is
1500 W, the nominal dc voltage is 360 Vdc, PV array MPPT operates 150-450 Vdc,
operational voltage is 196-253 Vac, the operational frequency is 50/60 Hz, THD is less

than 3%, the maximum conversion efficiency is 95%.

36

DC/DC DC/AC
boost inverter

T—m *”

 

 

\J

 

EM] __
PV FIL Filter " FIL - Grid

 

 

 

 

-p----_-___-_-

 

 

 

 

 

Figure 3.25: Sunville inverter topology [46].

' Fronius

FRONIUS IG 2000-LV has the topology shown in Figure 3.26. The nominal output
power is 1.8 kW, the nominal output voltage is 240 V, the THD is less than 5%, the

maximum efficiency is 94.4%.

37

 

I F RONIUS IG mm
. .
I I DC disconnect

+— —1-
'5. /—'_I -— 1—1—/
_/

I1; [I] 411*"?

I Grounding Terminal I I
l

  
  
  
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PV frame ‘ Energy __E ,
ground meter ——3 /

‘ "*1: t“
IG 2000/3000/ Lockable AC
2500-LV: 20A _ disconnect
IG 4000/5100/ Main SWItch
4500—LV: 30A _ grounding

system

 

 

 

 

 

 

 

Figure 3.26: FRONIUS IG 2000-LV [47].

° Studer

Figure 3.27 shows the A] series stand by inverters. For AJ 500, the battery voltage is

12 V, the input voltage is 10.5-16 V, the maximum efficiency is 93%, the output voltage

is 230V(115 V).

38

 

"" “"" .
1 R1 1
Metalic housing _

.—-y--‘.. DC/AC . L--O Brown

r
C UM! "
: Fuse I :
O-+ = t

 

 

 

 

 

II

t

 

 

’--0 Blue

 

 

 

"' * t
I L: e T—I—r-o dear
; 1uF/63 R2 20nF I Yellow

 

, wivw‘ v. rw'wifi—wmwvvv— ‘V vv-v— — w-rvtrwrv v vw‘rrvvv‘v—vwvvwrrva—w'

R1=R2=2,7M ohm(not existing in A] 275-AJ 3 50-AJ400)

Figure 3.27: A] series inverter [48].

° SMA

The SMA Sunny Boy inverter is not only the most popular grid tie PV inverter in
Europe, but also in North American. Over 250,000 Sunny Boy inverters have been
installed in worldwide. Figure 3.28 shows the Sunny Boy 1100U. The ac input voltage
is 213-262 Vac (nominal 240 Vac), ac input fi'equency is 59.3-60.5 Hz (nominal 60 Hz),
dc input voltage is 129-400 Vdc, the maximum ac power output is 1100 W, the power
factor is unity, the peak inverter efficiency is 93%, PV start voltage is 180 Vdc. Figure
3.29 shows Sunny Central SC125U. It is a 125 kW grid-tied string inverter for PV
system. For SC125U-240, dc input voltage is 275-600 Vdc, the peak power tracking
voltage is 275-550 Vdc, ac output voltage is 212-264 Vac (240 Vac nominal), current

THD is less than 3%, power factor is unity.

39

 

   
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

      
 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sunny Boy 1100U
Inverter 1 Utility
1 l
E
”i: IBreakeEI
DC disconnect AC discoEnect
PV switch swrtc AC _:
Array distribution Earth
panel Ground
Figure 3.28: Sunny Boy 1100U inverter [49].
PV Array . Disconnect Tran. Line AC
- switch Inverter 1251(VA contaCtor disconnect
“.3. ' I T — To Grid
' 480/240/208
:2: .,. WYE
6on
’ External @
Optional . Options
Combiner Box /4 Temperature Control Power
Irradiance Transformer
Ground Other sensors
Fault
Detector

 

 

 

 

 

 

Figure 3.29: Sunny Central [49].

40

° Magnetek

Figure 3.30 shows AURORA isolated outdoor model: PVI 2000-I-OUTD-US. The
maxmiurn power rating is 2000 W, the absolute maximum voltage ranges from 0-600
Vdc(360 Vdc nominal), power tracking is from 90-580 Vdc( 360 Vdc nominal), nominal
ac voltage ranges from single phase 211-264 Vac, the nominal AC frequency is 59.3 to
60.5 Hz, the line power factor is 1, ac current distortion is less than 2.5% THD at rated

power with sinewave voltage, the maximum efficiency is 94%.

 

 

 

 

 

 

Booster] 2?: Isolation T PLM

7:1“- “I: a it"

4%

 

+

 

 

 

 

 
 
   

 

 

II

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E ‘._-_‘___‘ _ __ Grid
‘ _-J:I__ ——-——J Vnom=240Vac
PV Array ~——-—-~ ~
Voc=600Vmax 303? C my C
Vnom=360Vnom o—I System controller WMW
ern=150Vm1n i Ground ,fiI-T. I
Fault Fuse PE

 

 

 

 

 

1

Figure 3.30: AURORA isolated outdoor model: PVI 2000-I-OUTD-US [51].

' PVPowered
Figure 3.31 shows PV Powered Starlnverter topology. For PVP2000-120, dc input
range is 135-450 Vdc, dc operating voltage ranges from 135-320 Vdc, operating ac

voltage ranges from 106-132 Vac, operating frequency ranges from 59.3-60.5 Hz,

nominal ac output voltage is 120 Vac.

41

PVP3200/PVP2900
PVP2000

Control PCB Power Distribution PCB

PV Vdc - '- - gr: .. 240 VAC
240:170- . -_ 60112 ,

450VDC

DC 1
Power . 1

System Block Diagram

 

Figure 3.31: Starinverter PCS [52].

° SOLECTRIA

Figure 3.32 shows PVIl3KW inverter topology. This inverter is used in 5-13 kW, 60
Hz, 208 Vac, 480 Vac, 3 phase, grid tied commercial PV systems. Multiple PVIl3KW
inverters can be used together for 20-26, 38, 52 kW (ac) or larger PV systems. Fully
integrated design includes transformer, filters, ac & dc disconnects, dc combiner-fuses.
The maximum continuous power is 13.2 kW, power factor is unity, ac output voltage is
208/480 Vac, THD is less than 5%, ac frequency is 602t1 Hz, the inverter peak efficiency

is 94%, the MPPT voltage ranges from 225-385 Vdc.

42

 

I 13 kW 3-PHASE COMMERCIAL PV INVERTER BLOCK DIAGRAM I

 

  
     
 
 
   
 

 

 

 

 

 

 

 

 

 

 

 

 

PVI l3kW IIII Utility
AC
DC AC I Distribution
lI Fuses Fan Disconnect , Panel
[L‘fl Swrtch L,

 

 

 

 

 

Loads

 

 

DC EM]
Disconnect Inverter I Delta/W ye F' l t
Switch Transformer 1 er

 

 

 

Figure 3.32: PVIl3KW inverter system [53].

Also, micro inverter is a type of grid-tie inverter. These inverters are small, usually
less than 500 W, and are mounted directly on a solar module or in close to it. This type
of inverter reduces the amount of dc wiring and protective equipment used in a PV
installation. Xpower Micro inverters-800 has the continuous ac output power 640 W, ac
output voltage is 120 Vac nominal, ac output frequency is 60i4 Hz, dc input voltage
range is 10.5-15.5 Vdc. This inverter has a wide application for cell phone, camcorder,
stereo, etc.

Inverter manufactures are developing new products to compete in the rapidly growing
North American grid tie PV markets. There are a lot of PV inverters in the market as

summarized in the following table.

43

(Xantrex, Ballard, Beacon, Phoenixtex Power [44-46, 53])

PV commercial

Table 3.1: Manufacture PV inverters.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

inverter
éfigeggx) 32101;: Beacon M5 Sunville 1500
PCS topology dc/ac with dc/ac with dc/dc with dc/dc with
transformer transformer dc/ac dc/ac
A“ “’th vouage 106-132 Vac 196-253 Vac
range
Ac out”? ”Rage 208 Vac 208 Vac 120 Vac
nominal
Dc input voltage 290-595 Vdc 48-110 Vdc 360 Vdc
nomlnal
Ac frequency 60+0.5/-.7Hz 60 Hz 60 Hz nominal 50/60 Hz
power factor >0.99 0.99 >0.99
THD <5% <4% <3% <3%
MPPT range 330-600 Vdc 240-600 Vdc 50-85 Vdc 150—450 Vdc
Efficiency >95% (peak) 94.6% (peak) 93% (peak) 93%-95%
Battery 48 V
Con.power rating 10 kW 30 kW 5 kW 1500 W
Weight 1151b/52kg 4301b,3751b 1201b/54.5kg 8.5 kg
Dimensions(cm) 66*41*30 104*107*41 107*41*26 31.5*26.9*12

 

44

 

 

PV commercial

Table 3.2: Manufacture PV inverters.

(FRONIUS, Studer, SMA [47—49])

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

inverter
FRONIUS 1G Sunny Boy SunnyCentral
2000 Stud“ A15 00 1100U SC125U-240
dc/ac with dc/ac with
PCS topology dc/ac trans former dc/ ac transformer
Ac ““1"“ 212-264 Vac 219'242’109' 213-262 Vac 212-264 Vac
voltage range 121 Vac
Ac outpui 240 Vac 230/115 Vac 240 Vac 240 Vac
voltage nomrnal
Dc input voltage 27°ch 10.5-16 Vdc 129—400 Vdc 275-600 Vdc
nomrnal
Ac frequency 59.3-60.5 Hz 50/60Hz:1:0.05% 59.3-60.5 Hz 59.3-60.5 Hz
power factor 1 1 1
THD <5% <4% <3%
MPPT range 150-400 Vdc 145-400 Vdc 275-550 Vdc
Efficiency 94.40% 93% 93% 95.70%
Battery 12 V battery
Con.power rating 1800 W 400 W 1100 W 125 kW
Weight 261b/l 1.8kg 4.5kg 56.26lb/21kg 33071b/1500 kg
Dimensions(cm) 47*41.8*22.3 14.2*8.4*24 32.2*32*18 235*150*60

 

45

 

Table 3.3: Manufacture PV inverters.

(Magnetek, PV Powered, Solectria, Xantrex [SO-53])

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PV commercial
inverter
Aurora PVI- PV Powered Solectria Xpower Micro
2000-I-OUTD PVP2000-120 PVIl 3KW Inverter800
dc/dc wrth dc/ac with dc/ac with dc/ac with
PCS topology dc/ac wrth
transformer transformer flyback
transformer
Ac output single phase _
voltage range 211-264 Vac 106 132 Vac
A“ “up“? 120 Vac 208/480 Vac 120 Vac
voltage nomrnal
. 0-600Vdc
Dc 1nput voltage (3 6 Onominal) 135-500 Vdc 10.5-15.5Vdc
Ac frequency 59.34505 Hz 59.3-60.5 Hz 60sz:1Hz 60Hz:t4Hz
power factor 1 l
THD <2.5% <5%
MPPT range 90-580 Vdc 135-360 Vdc 225-385 Vdc
Efficiency 94.00% 93% 94% 90.00%
Battery Yes
confowe’ 1500 w 2000 w 1320 w 640 w
rating
Weight 25kg 76lb 3801b/173kg 11b40z/852g
Dimensions(cm) 42*32.6*23.2 38.l*19.1*55.3 87.6*66*34.5 19.1 *1 1.4*6.1

 

 

 

 

 

46

CHAPTER 4. Z-SOURCE INVERTER BASED
POWER CONDITIONING SYSTEM

4. 1. Introduction

With the development of solar cell technology, the price of solar module has dropped
dramatically. Recent worldwide survey shows that in last three years, the retail price of
solar module has dropped 16.95%. However, at the same time, the prices for the PCSs
almost remain the same. Further more, compared with converters used in drive systems,
the prices for the converters used in PV systems are still up to 50% higher. To lower the
cost of the PC85 has become a very urgent issue of grid connected PV system [1].

PCS is required to convert the dc output from PV to grid synchronized 50 or 60 Hz ac.
Here a Z-Source inverter based PCS, which connects the PV arrays for residential
systems that are 60 Hz, 120/240 V split phase power in the US was proposed. By
utilizing the Z-Source inverter, the number of switching components and the total volume
of the system can be minimized. Moreover, the Z-Source inverter makes it possible to

use PV that has a wide range voltage change. Thus the cost of the PCS is minimized.

47

4. 2. The Basics Of PCS For Residential Use

In order to feed energy from a PV array into the utility grid, PCS converter system

has to fulfill the following three requirements:

1) to convert the dc voltage into ac voltage;

2) to boost the voltage, if the PV array voltage is lower than the grid voltage; and
3) to insure the maximum power utilization of the PV modular.

Figure 4.1 shows the two most commonly used converter system configurations in
practice. In the system shown in Figure 4.1(a), a transformer at line frequency is utilized
to boost the voltage after the dc-ac inverter. Usually, a line frequency transformer is
associated with huge size, loud acoustic noise, and high cost. In addition, the inverter has
to be over-sized to cope with the wide PV array voltage change. The KVA rating of the
inverter is doubled if the PV voltage varies at a 1:2 range. So to eliminate the
transformer and to minimize the required KVA rating of the inverter, in many
applications, a high frequency dc/dc converter is used to boost the voltage to a constant
value as shown in Figure 4.1(b). Unfortunately, the switch in the dc-dc converter
becomes the cost and efficiency killer of the system [2-3].

Another option is to use a single-stage inverter for direct dc-ac conversion as shown
in Figure 4.2. For the split-phase system used in US residential power, two 120 V ac
outputs with the same ground and 180 degree phase difference are required. For this
purpose, there are two circuit choices for the dc/ac inverters in the PCS: four-switch
inverter and six-switch inverter. The circuit structures of these two choices are shown in

Figure 4.2.

48

  
 

 

 

PV
array

  
 
 

DC/AC

Inverter To Gnd

   

(a). Dc/ac with step-up transformer.

 

 

 

 

\I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l—____—__———-____—___-—____—'
I
_ _I_.
03,053,, g .2342; Filter I To one
_ I_.
.'
____________ 5&5““"""‘

(b). Dc/ac with dc-dc boost.

Figure 4.1: The traditional PV systems.

49

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Utility

 

 

 

 

 

 

 

 

Vdc ;: J .1
2
"6‘0"
PV "
Array __
_Vdc F: J J
2 Load
(a). Four switch inverter.
‘ ‘ Utility
J J J
”1111‘
Array dc" i __
? o—rfl‘
J J J
I Load I
(b). Six switch inverter.

Figure 4.2: Direct PV inverter systems for i120 V split phase residential power.

For the four-switch inverter as shown in Figure 4.2 (a), the neutral point is tapped
from the center of the two dc capacitors, whereas in six-switch inverter shown in Figure
4.2 (b), the neutral point is connected to the third phase leg.

50

The two phase legs in the four-switch inverter are controlled by SPWM. Two
sinusoidal control references with a 180 degree phase difference and the same amplitude
are utilized to compare with a triangular carrier.

The basic control scheme of the six-switch inverter is shown in Figure 4.3. Two of
the phase legs, “b” and “c”, have the exact same SPWM control as in four—switch inverter.
The third phase leg, “a”, is usually controlled to produce a square waveform with 50%
duty ratio at the carrier frequency to serve as the neutral phase and at the same time
achieve the maximum utilization of the dc bus voltage. The switching frequency of the
third leg can be different fiom the other two phase legs. By proper coordinating the
control of the neutral phase leg and the other two phase legs, the equivalent switching
frequency can be doubled, thus the output filter can be optimized. It is generally believed
that the six-switch inverter has better performance than the 4-switch inverter for the split-

phase application [4].

Sap
pr
Scp
San
Sbn

Scn

 

Figure 4.3: The six-switch inverter control scheme.

51

4. 3. The Proposed Z-Source Inverter System

In the proposed PCS, a Z-Source inverter [5] is utilized to realize inversion and boost
function in one single stage. Figure 4.4 shows the proposed system. Unlike the

traditional voltage source or current source inverters, the Z-Source inverter employs a
unique impedance network with split inductor L1 , L2 and capacitor C1, C2 connected

in X shape. With the impedance network, the Z-Source inverter can advantageously use
the shoot through states to boost voltage. Further more, with the ability to handle shoot
through state, the inverter system becomes more reliable [5-6]. The inductors and
capacitors in the Z-Source are both energy storage devices, so their value can be
optimally designed to ensure small size and low cost.

Compared with the systems in Figure 4.1, in the proposed system, there is neither
bulky transformer nor a dc-dc converter to boost the voltage in the circuit. The size and
cost are minimized. Because no dead time is needed, the control accuracy and harmonics
can also be improved. Further more, the split-phase Z-Source inverter naturally inherits
all the advantages of the split-phase six-switch inverter. Thus, the 120 V ac output filter
can be optimized.

Compared with the direct inverter systems in Figure 4.2, the Z-Source inverter has
minimum KVA requirement. For the inverter system in Figure 4.2 with a PV voltage
change of 1:2 ranges, the PV dc voltage needs to be 340-680 V minimum to produce a
i120 V split-phase power. Therefore, 1200 V IGBTs are needed in the system. Given a
10 kW PV system, a 20 kW inverter is needed to cope with the voltage change. Using

the proposed system of Figure 4.4, the PV voltage can be designed to be 225-450 V,

52

which can be inverted to $120 V split-phase power by the Z-Source inverter using 600 V
IGBTs. In addition, the required KVA rating of the Z-Source inverter remains the same

10 kW for a 10 kW PV system.

 

 

120/240
V3 60Hz
Split Phase

 

 

 

 

 

 

 

 

--_..J

    

Z-Source 3-phase Inverter

Figure 4.4: The proposed PCS for stand-alone PV system.

Therefore, by utilizing the Z-Source inverter, the volume, the cost as well as the
number of active switching devices are minimized. Because of the single stage
operation, the efficiency of the system can be greatly improved. The proposed system:

1) has only one stage to realize inversion, boost, and maximum power tracking;

2) has the minimized number of switching devices;

3) needs no dead time;

4) can have shoot through state in the inverter; and

5) inherits all the advantages of the six switch inverter system.

53

4. 4. Control And Operation Principle

As shown in Figure 4.3, when the triangular waveform is greater than the maximum
value or lower than the minimum value of the three reference waveforms, all upper three
switches or all bottom three switches are turned on respectively, as indicated in the
shadowed area. During these periods of time, the output voltage of the inverter is zero,
i.e., zero states. For the Z-source inverter, the basic idea of control is to turn zero states
into shoot through states and keep the active switching states unchanged, thus we can
maintain the sinusoidal output and at the same time achieve voltage boost from the shoot
through of the dc link [5-6].

Figure 4.5 shows the boost control method of the split-phase Z-Source inverter.

Phase legs “b” and “c” are controlled by SPWM to synthesize ac output, and the shoot-

a:
through command, Vsikp and Vsn is used to boost voltage as desired. The control of

these two phase legs is similar to the simple boost control proposed in [5] and [6]. Two

*
straight lines, Vs?) and Vsn are used to control the shoot through duty ratio. When the

carrier is greater than the upper straight line, phase leg “b” goes to shoot through state,
whereas phase leg “c” goes to shoot through state when the lower straight line is greater
than the carrier. Phase “a” in the inverter is switching at 50 percent duty cycle without
shoot through. The switching frequency of phase “a” is the frequency of the carrier. By
doing this, each phase leg only shoots through once during one carrier cycle, the

equivalent switching frequency can be doubled for the output filter.

54

 

 

 

 

/
/
+an
"it:

 

VMI IV]: EN:
11

g_

align
fifi
. g:\\
3 ; : :: : :Kégg
9. F144. a- a...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E
L
t.
if
1:.
h 4;:
E1 4;

 

Sci 1111 ii 1? WW

I I

     

Figure 4.5: Sketch map of the simple boost control.

As described in [5] and [6], the inductors L1 and L2 , which can be wounded around
the same core, have the same inductance, and the capacitors C1 and C2 have the same

capacitance, the relationship between the output ac voltage and input dc voltage is found

as
where 170 is the output peak voltage, va is the PV output voltage, M is the

modulation index, which is defined by

M = Vpeak /Vm-, (4.2)

55

and B is the boost factor, which is determined by

B: 1 ,
1—2-T0/T

 

(4.3)

where T 0 is the total shoot-through period per carrier cycle, and T is carrier cycle. In

the simple control method [5,6], the amplitude of the two straight lines is the peak of

modulation waveform, therefore the relationship between modulation index, M and the

shoot through ratio, T 0 / T can be found as
T0/T =1—M. (4.4)

Substituting (4.4) into (4.3), the boost factor becomes:

 

l
B = 2M _1 . (45)
From (4.1), and (4.5), the peak amplitude of the output can be expressed as
130 = 22—5—33 va. (4.6)
While the capacitor voltage is:
1 - 2%)— B + 1
VCl = Vc2 = :z—Tj va = —2— va. (4.7)
T

When the PV output voltage value is high enough to produce the required ac voltage,
the shoot through state is no longer needed, i.e., T0 = 0 and B =1 . Under this

condition, the relationship between the inverter peak output voltage and the PV output

voltage can be calculated by

56

%=M%/2 mm

It also should be noted that the shoot-through states can be created by shorting both
legs “b” and “c”, or all the three legs simultaneously during any given shoot through
states according to the two straight lines. For all these shoot-tlrrough cases, the resulted
boost effect and output voltage waveforms remain the same.

In the proposed control scheme, only one phase leg is used to create shoot through at
any time, thus minimizing the switching frequency. However, at the same time, the
current stress on each switch during shoot through is doubled when compared with
shooting through two phase legs simultaneously at any time. A trade off in the control
must be made, one can

1) either reduce the switching frequency by shorting one or two phase legs; or

2) reduce the current stress on each device by shorting all phase legs during shoot-

through periods.

To make a decision in real applications, the switching and conduction losses at

different conditions need to be calculated and investigated for different cases.

4. 5. Design Guideline

In the Z—Source based PCS, the maximum voltage over phase legs Vpn is controlled

to maintain the split phase output at different input voltages. For 120 V split phase
output, PV cell with maximum output voltage of 450 V can be used. Assume the

minimum output voltage is half of the maximum voltage, to achieve the same output ac

57

voltage of 120 V, the device voltage stress can be calculated by manipulating (4.5) and

(4.6), which results

V

p" = 3va = 455. (4.9)

Thus, 600 V device can be used.

The maximum current stress on the device can be simply calculated based on the

following equation

__2_ “V8
1m ————+21L_ - +2—. (4.10)
V va

Where I L is the Z-source inductor current, Pmax is the maximum transient output
power, and Pavg is the average output power. The current stress can be reduced by
turning on two or all phase legs during shoot through to distribute the 21L into different
phase legs.

To determine the inductance and capacitance of the Z-Source network, the input
power is assumed to be constant dc. The ripple power is absorbed by the capacitors of

the Z-Source network. The power ripple absorbed by the capacitors is calculated

1
480 1 2 1 2
AP=P 1 cos(120*2*7r)tdr=§C(V+AI/) -§CV , (4.11)
1
480

58

where V is the average voltage across the capacitor, C is the total capacitance of two
capacitors in the Z-source. To limit the voltage ripple to be less than x%, the capacitor

value can be calculated.

Based on the voltage ripple across the capacitors, the ripple voltage across the
inductor is the voltage difference between the capacitor and the voltage across the PN.
The capacitor voltage is already known with x% ripple. The PN voltage is zero during

shoot through and ZVC — va for others. Assume the input voltage is a constant dc

value, the PN voltage ripple is 2VcD0 , therefore, the voltage ripple across the inductor
is

2T0

Vripple =(1— T)*x%*V. (4.12)

To limit the ripple to a special percent, the inductance can be determined.

4. 6. Simulation And Experiment Results

Photovoltaic solar cells have nonlinear V-I characteristics. Its output voltage and
power change according to temperature and irradiation. Figure 4.6 shows the typical V-I
characteristics for a PV module. For a specified temperature and irradiation, the
intersection of the load line with the photovoltaic voltage-current characteristic, is the
operation point. In real practice, PV modules are first cascaded then paralleled to form

PV array, thus to meet the voltage and power requirement.

59

 

 

 

 

 

 

 

 

 

 

 

15
A 12.5 g
f} 10 "E,
s: "U
a 7.5 g
=3 (D
$3 5.0 I.
a. r:
’5 2.5
o
0

 

 

 

Output Voltage (V)

Figure 4.6: Typical V-I and P-V characteristics.

Figure 4.7 shows the V-I curves of PV array at different temperatures and irradiations.
Usually the photovoltaic output voltage changes mainly with the temperature, while the
photovoltaic output current changes mainly with the irradiation. With constant
irradiation specified, the PV output power decreases when the temperature rises. With
constant temperature specified, the PV output power increases when the irradiation
increases. One of the functions of the PCS is to extract the maximum power out of the
photovoltaic at any given temperature and irradiation. Based on the curves shown in
Figure 4.7, simulations and experiments are performed to prove the concept proposed in

this thesis at the following two conditions:

2

l) at 1000 W/ m , and 60 °C , va =230 V, the maximum PV output power is 7200

W;

60

2

2) at 250 W / m and 0 °C , va =450 V, the maximum PV output power is 3360 W.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

A 40 2
3:, 35 _ 1000 W/m ,60°C 2
’g t_____ 1(0) W/m ,25°C
5 30 -
9 ,/
=3 25 "
e- 2 o
5 20 _ 8mW/m ,45 C

15 i f“- -

10 _ 500W/m2,25°c

______7_- _________
5 ' 250W/m2,0°C
0 g . .

 

 

 

0 80 160 240
Output Voltage (V)

Figure 4.7: V-I characteristics under different conditions.

For both cases, the simulation and experimental systems are setup with the following
parameters: L1 = L2 =1 mH at the line frequency and C1 = C2 =13,300 ,uF. The
switching frequency is 10 kHz. Z-source inverter produces PWM voltage waveforms just
like the traditional inverter. Thus, a LC filter is added after the inverter to achieve
sinusoidal waveform. The filter parameters are L =1 mH and C = 120 ,uF . We
assume that output voltage range of PV arrays is 1:2. So 600 V IGBTs, which has
maximum dc voltage as 450 V, was used to operate in 225-450 V in experiments.

Resistive load is used for the simulations and experiments. For case one, 4 Q is used for

61

each phase, whereas 8.57 (2 is used for case 2. A 10 kW Z-Source inverter prototype

used in the experiment is shown in Figure 4.8.

f!
Diode Brldge .‘r' 7

h

DSP Control

Figure 4.8: A 10 kW Z-Source inverter prototype.

 

2

Case 1.-1000 W/m and 60 °C, va =230V

For the first case, the input voltage, va, is 230 V, to achieve 120 V split phase
output, the modulation index can be calculated by the following equation, which is the
inverse of (4.6):

2170 _ 2 x J2 x 120

T 400 —va _ 4x./2x120—230

 

=0.755 (4.13)

The boost factor is

62

B: 1 =1.961. (4.14)
2M—1

 

Thus the shoot through duty cycle T 0 / T can be calculated from (4.3), which will result

a T 0 / T equals to 0.245. The Z-Source capacitor voltage would be

 

l—TO/T V 1—0.245

__ = 230:340 V. 4.15
1—2T0/T 1’" l—2x0.245 ( )

VCl = VC2 =

Figure 4.9 and Figure 4.10 show the simulation and experimental results of case 1.
Figure 4.9(a) and Figure 4.10(a) show the simulation and experimental waveform of the
input voltage and Z-Source capacitor voltage waveforms, respectively. In both
simulation and experimental results, the capacitor voltages are close to 340 V.
Meanwhile as shown in Figure 4.9(b) and Figure 4.10(b), the simulation and
experimental results of the output voltages are both close to the desired split phase 120 V
rms. The output power can be calculated based on the output voltage and the load current
waveforms, which are also shown in Figure 4.9(b) and Figure 4.10(b). As the load
current is 30 A, the total output power of the inverter is around 7200 W. Thus the
maximum power output from the PV at this condition is realized. Figure 4.9(c) and
Figure 4.10(c) show the output filter inductor and Z-Source inductor current. The Z-
source inductor average current is around 31 A, which on the other hand, proofs again
that the PV outputs its maximum power. Both the simulation and experimental results

are consistent with the theoretical calculations.

63

Figure 4.9: Simulation results of case 1 under the conditions of

2

1000 W/m ,60 °C,and va=23o V.

 

 

GraphO

 

400.0 ' V
3000.: . . ,,,,,,

 

100.0“ ’

°-°‘ (V)=t(5)
500.0' ' ,

400.0 .. .......................................................... .
0.0 J'1—————l>-———'—'—''—""'—"'_"'—"""—'l=-—’—-' l
2-2 2.3 2.4 2.5 2.6

t(s)

 

 

 

 

 

 

fir

 

 

 

 

 

 

 

(a). Input Voltage and Z-Source capacitor voltage.

64

 

 

 

(A)

(A)

 

 

 

Figure 4.9 Continues:

 

(A)

 

 

    

GraphO
(V) : 1(8)
100.0 200.0 '
1’2
50.0 7 100.0
0 0 - E o o (A) 11(5)
-5o.o ' -1 00.0 ’2

 

—100.o - -200.o (V) :t(s)

 

 

 

100.0 - 200.0
50.0 ' 100.0
0.0 7 E o 0
-50.0 - -100.o

-1oo.0 ' -200.o -,

 

2.525 2.55 2.575 2.6
t(s)

 

(b). Load voltages and currents.

65

 

Figure 4.9 Continues:

 

 

Grapho

(A) 11(5)

 

 

 

 

 

 

 

 

 

 

 

 

2.5 2.52 2.54 2.56 2.58
t(S)

 

(c). Filter inductor current and Z-Source inductor current.

66

 

 

Figure 4.10: Experimental results of case 1 under the conditions of

1000 W/m2, 60 °C,and va =230 V.

 

 

Main: 10k

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

va (100V/div)
I: ........................... .---fi-- -fi- ..........
:_r; ....................
Chllzl Ch21:l
0.100V/div 0.100V/div
DCFull DCFull

 

 

 

 

(a). Z-Source capacitor voltage and Input Voltage.

67

Figure 4.10 continues:

 

 

 

Main: 10k

12 (50A/div)
V

 

 

 

Chl 1 :1
lOmV/div
DC50 Full

 

 

Ch2 1:1
IOmV/div
DC50 Full

 

 

C113 1:1

Ch4l:l

0.200V/div 0.200V/div
DC Full DC Full

 

 

 

 

 

(b). Load voltages and currents.

68

 

Figure 4.10 continues:

 

Main: 10k

 

 

 

 

 

 

 

 

Ch3 1:1 Ch4 1:1
IOmV/div IOmV/div
DC50 Full DC50 Full

 

 

 

 

 

 

(c). Filter inductor current and Z-Source inductor current.

69

 

“H

 

Case

Fc

Strateg

Ill

Figure 4.10 continues:

 

Main: 10k

 

 

 

 

 

 

 

......... Vl(100V/div)

 

 

 

 

 

 

 

 

Ch2 1:1 Ch4 1:1
lOmV/div 0.200V/div
DC50 Full DC Full

 

 

 

((1). Input current Iin and load voltage.

Case 2: 250 W /m2 andO °C, va =450 v

For the second case, as the input voltage is much higher, va =450 V, the control

strategy is different with the first case. Shoot through is not used for this case.

The modulation index can be calculated as following

M=2i=m=0755 (4.16)
p, 450

70

 

 

 

A5 11
capacito‘
conditio

Figt
Similar

well.

 

As no boost is needed for this case, the boost factor B is 1. Thus, the Z-Source
capacitor voltage would be the same as the PV voltage, which is 450 V. Under this
condition, the theoretical maximum output power from PV is around 3360 W.

Figure 4.11 and Figure 4.12 show the simulation and experimental results of case 2.

Similar as in Case 1, the simulation and experimental results also verify the analysis as

 

 

 

 

 

 

 

 

 

 

 

 

well.
Figure 4.11: Simulation results of case 2 under the conditions of
250 W/m2,0 °C,and va =450V.
GraphO
(V) 11(5)
500.0‘
........... . .. .. .. .. VP"
400.0 ‘
200.0“
100.0. . ...
0-0‘ (V) :1(s)
600.0 '
VCl
A4000-
3
200.0" "'
°-° ‘I———r'——r——-r——l———“'_—’
3.3 3.32 3.34 3.36 3.38
t(s)

 

 

 

(3). Input Voltage and Z-Source capacitor voltage.

71

 

 

 

(A)

(A)

 

Figure 4.11 continues:

 

 

(A)

60.0
30.0 7

0.0 7
-30.0 7
-60.oi
50.0 7
30.0 7

0.0 7

-30.0 "

 

-60.0 ‘

(V)

200.0

100.0

0.0

-100.0

-200.0 '

200.0

100.0

 

 

 

 

 

3.3 3.32 3.34

t(s)

3.36

 

(V) 2 1(3)

V2

‘ (A) 11(3)

12
(V) :t(s)

V1

(A) :t(s)

’1

 

(b). Load voltages and currents.

72

 

 

 

 

(A)

(A)

 

Figure 4.11 continues:

 

 

40.0 ‘

20.0‘ "

0.0 ‘

(A)

-20.0' ‘ ’

 

-4o.0 ‘
20.0 7

5.0‘ "

 

0.0 ‘
3

GraphO
(A) 21(8)

 

’23

 

 

(A) 21(8)

 

iLl

 

 

 

I l I I I
.3

3.34 3.36 3.38

t(s)

3.32

 

(c). Filter inductor current and Z-Source inductor current.

73

 

Figure 4.12: Experimental results of case 2 under the conditions of

2

250 W/m ,0 °C,and va =450 v.

 

 

 

 

 

 

 

 

 

Mairi: 10k
Vcl (250V/div)
;

ooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo
I

 

 

 

 

 

 

 

 

 

 

 

 

 

Ir. 3
3E .........
Chl 1:1 Ch21:l
0.500V/div 0.500V/div
DCFull DCFull

 

(a). Z-Source capacitor voltage and Input Voltage.

74

Figure 4.12 continue:

 

 

v2 (100V/div) Maim {0k 1'2 (20A/div)

 

 

 

 

 

I V1 (100V

......

ydiv) 1.1 ......... 11(20A/div) I ........ .........

...... .........

 

 

    
     

   

     

9H l?

 

Ch11:1 Ch21:1 Ch31:1 Ch4121
0.200V/div 0.200V/div 10mV/div lOmV/div
DC Full DC Full DC50 Full DC50 Full

 

 

 

 

 

 

 

 

 

 

(b). Load voltages and currents.

75

 

Figure 4.12 continues:

 

Main: 10k I

......... I :13 (sz/div)- I

 

 

 

 

 

 

 

 

 

 

 

Ch3 1:1 Ch4 1:1
10mV/div lOmV/div
DC50 Full DC50 Full

 

 

 

 

 

(c). Filter inductor current and Z-Source inductor current.

76

Figure 4.12 continues:

 

 

 

 

 

 

 

 

 

 

 

 

 

—— IMairi:10kI
......... I Iin(2A/div) I
. i 6 2 2 2 - E i .
1:: .............................................................................................
3:: ............................................................................................
v1 (100V/div)
Ch21:1 Ch41:1
lOmV/div 0.200V/div
DC50 Full DC Full

 

 

 

 

 

 

((1). Input current Iin and load voltage.

From the above two cases, it can be concluded that the simulation and experimental
results show that at different input voltage, the proposed PV system’s output voltage
maintained at i 120 V rrns, whereas the output power tracked the maximum PV output
power at different temperatures and irradiations. The basic principle of the proposed

system was verified.

77

 

4.7. Su

A ne
propose
single;
of Z-S
togeth
the ad

for P‘

4.7. Summary

A new PV power conditioning system based on Z-Source inverter is proposed. The
proposed system realizes the boost and inversion with maximum power tracking in one
single power stage, thus minimizing the number of switching devices. All the advantages
of Z-Source inverter and six-switch split-phase inverter are inherited and integrated
together to create a highly reliable PCS system with minimized volume and cost. With
the advanced features summarized in section III, the pr0posed system is very promising

for PV power conditioning applications.

78

 

5.1.111

The
also on r

inverter

T0 e
range, a
appropr
recorde
UneXpet
Simular

Condnh

 

 

CHAPTER 5. PV SIMULATOR SYSTEM

5. 1. Introduction

The PV system performance depends not only on temperatures and irradiations, but
also on maximum power tracking function of PV inverter. So it is important to verify the

inverter system also.

To evaluate PV inverter system, ideally the inverter need be tested through the entire
range, as shown in the Figure 5.1. One method is that inverter is connected to the
appropriate size of PV panels in a sunny day, and PV'panels’ output characteristics are
recorded. This method has the disadvantages that the limitations of sized PV panels, the
unexpected weather conditions, and the long running time. Another way is to use a PV
simulator, which is able to produce repeatable conditions respect of environment

conditions.

79

 

 

 

- t (A)
Curren
Module

5.2.1

Tilt
5111 1101
i0“I ph

0016 it

0.1.

 

 

Char gig-1

1&th

 

 

I-V Curve
0.7 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Isc
-——--\ Knee point
0-6 \ / \ / \
\/ \><
05 /
3 / \ /‘
+- 0.4
1: /
5 \ >. i
3 > /
:3
'8 / Area=Pmax >< I
2 0.2 / \ \ / \
\ \,
0.1
\ \
o
0 5 10 15 20 25
Voltage (V) V013 Vmax

Figure 5.1: Typical PV module curves.

5. 2. Previous PV Simulator

There already exist some methods to build up a PV simulator model for simulation,
but not too many literatures for hardware implementation. Most of them are capable for
low power rating application; some are capable for medium and high power rating. The
core technology of the simulator circuits is a control circuit that simulates PV l-V
characteristics curves. People can approach it use microcomputers or some other analog

techniques.

80

Figure 5.2 shows a PV simulator concept in [62]. In this paper, a Si and
polycrystalline panels with a simulated light source were used. The output current and
voltage were recorded as a function time. With the use of class A regulator and switched
operation, the fast dynamic response can be achieved. The internal current control which

allows additional units to be connected in parallel in order to get higher output currents.

Power Supply Module

mega A
_J—/
I] i B

__.,
___l

/

 

 

 

 

 

 

/

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”J C Class A
.. Regulator
Vdes + Verr
T [1
Dynamic
Inverter
Load
Comparators CT

3 H Vdes ]
C

 

 

 

Figure 5.2: PV simulator concept [5].

81

Figure 5.3 shows a block diagram of the PV generator simulator circuit in [63]. In the
circuit, the output of a pn photo sensor, just like a small solar cell, was magnified through
a dc power amplifier. The maximum power is around 30 W. But this method is not
suitable for high power cases.

Small PV cell .
DC power amplifier

I A Vo Output

 

 

 

 

 

 

 

 

 

 

Operational point control

 

 

Figure 5.3: Block diagram of the PV generator simulator circuit [63].

Also, some test equipments used lamp to simulate the solar irradiation with its
consequent need of huge power for bigger power rating larger than 1 kW, thus to
eliminate atmospheric dependency [64]. Others tried to use a current source and a diode
chain, which can represent 3 PV panel through its electrical scheme. But it is still have
thermal stability problems and was limited by its diode chain’s fill factor. Due to the
reasons above, the PV simulator which are applicable for high power was designed in this

thesis.

82

DC amplifier

.—-—I>-—M—-—o v.

1

 

DC bias unit
+

R__

LED driver

 

 

 

 

  

 

 

 

Figure 5.4: The circuit of the simulator with a light emission unit [65].

There already exist some commercial PV simulators in the market, for example
Ainelec. This simulator is based on a simplified curve which has the short-circuit current,
the open-circuit voltage and the maximum power point. The curve is composed of two
parts. The first part is decided by the short-circuit point and the maximum power point,
while the maximum point and the open-circuit point works for the other one [92]. This

simulator makes the curve simple, but not accurate.

For simulator from Elgar solar, it is has a wider power range and better controllability.
The simulator can output more array strings through building blocks [93]. This one is

much more complex than the Ainelec.

83

5. 3. Proposed PV Simulator

5.3.1 Mathematical Modeling Of PV Cell

In this thesis, a dc-dc converter which can output I-V curves was designed and
implemented. To have the I-V curves from the manufacture PV panel, we first need to

get the mathematical model in order for it to be useful in a computer controller simulator.

Figure 5.5 shows an equivalent circuit of a PV cell. The short circuit current I sc is

ideally equal to the generated light current Iph .

I19h 1d

 

 

Figure 5.5: An equivalent circuit of a PV cell.

The open circuit voltage Voc is expressed as the following:

I
Voc =k—Tln(Lh+1) (5.1)
q 10

Where, I ph is generated light current, I 0 is saturation current of diode, k is

Boltzman constant, q is electric charge, and T is operating temperature.

84

The output current is dependent on temperature since the current increases slightly as

the temperature increases. However, as the temperature increases the open circuit voltage

tends to decreases. Relationship of the open circuit voltage Voc and short circuit current

I SC according to the operating temperature variation is as follows,

Isc = 0 -eq V06 / H (5.2)
The diode saturation current can be represented as,
[rs = Irr '(%)3 'CXP[%'(%—%;)l (5-3)
The generated light current can be represented as,
Iph =[Iscr+kt°(Tc"Tr)J°—£' (5.4)
100

The operating temperature for output characteristics of PV array can be represented:
T=(0.3xS)+(0.9xTc)+273 (5.5)

So the output current-voltage characteristics of PV array can be represented as

equation (5.6) and it generates nonlinear output.

 

q V
I=Np’ ph—Np'lrs'iexP(KTA'E)-II (5-6)

For a given module, SM 60 module shown in Fig. 5.6, the PV array parameters can be

listed as the following:

85

1

Table 5.1: The PV array parameters (SM-60 Module).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

A : P-N junction ideal
2.25 5' :Solar radiation 100
factor
k :Boltzrnan constant 1.3e-23 Np :Parallel CifCUit 1
, 1.60226- , . ,
q :Electrlc charge 19 N S :Serles c1rcu1t 36
E g :Energy band gap 1.11 T: Temperature 45
T : Cell reference I Short circuit
' 300 S” 3.75
temperature current
I rr : Reverse saturation
2.1e-5 kt :Constant of I scr 2,313.4
current
Solar Cell (SM60)
4.0 120
3.5
‘ 100
2 3.0 ‘
r: "U
5 2.5 80 c2
5 a
2.0 . 60 A
5 a
8* 1.5
8 40
1.0
0.5 20
0.0 . . . . . . . . . . . 0
0 2 4 6 8 10 12 l4 16 18 20 22
Output Voltage (V)

Figure 5.6: The output characteristic of the SM-60 module.

86

 

 

To parallel nine PV cells and series twenty one PV cells, we can get the desired
output PV simulator output characteristics shown in Figure 5.7. It is noticed that the
parameter values must be changed, depending on the manufacturing process and type of
PV cells. The parameter values given here are only for this particular PV array. To
change the temperature and irradiation parameters, we can get different curves for PV

array. Regardless of values chosen, the DSP controlled PV array will function properly.

 

 

 

Solar Cell (SM-60, 21x9)
40 20

 

 

 

35”
30’
25-

 

 

 

20'
15-

Power (kW)

Output Current (A)

 

 

 

O .‘ i i r 1 r 1 1 r J r 1 r a .1
0 42 84 126 168 210 252 294 336 378 420 462

 

 

 

 

Output Voltage (V)

Figure 5.7: The PV simulator output characteristics.

5.3.2 Control Strategies

There are voltage control and current control for simulator. As voltage control, the
output current is measured and used as index to look up the reference voltage for
simulator. While as current control, the output voltage is measured and used to look up

the reference current. However, PV characteristics are nonlinear, which is a current

87

source when resistive load is small and is a voltage source when resistive load is great
enough. At current source segment, the change rate of voltage is large while the change
rate of current is small, which as a result is infeasible for measured current to look up
voltage reference practically. Thus, only voltage control and current control can not

satisfy the requirement.

For PV array, the V-I curve changes due to the temperature and irradiation
nonlinearly. This nonlinear characteristic source can be simulated by a dc chopper.

Figure 5.8 shows the proposed PV simulator and its control strategy.

The method is to use a DC/DC converter to simulate the PV output characteristics, by
producing voltage and current defined by PV curves for specified temperatures and
isolations. For the simulator, as a current regulator, the voltage is measured and input to
the memory which gives a desired current value based on a look up table. As a voltage
regulator, the current is measured and input to the memory which gives a desired voltage
value based on a look up table. A PI controller is used to adjust the switch duty ratio to

achieve desired output voltage.

Because PV inverters always start from open circuit condition, and then approach to
their MPP and oscillate around it during the real cases. The dynamic response of the PV
array simulator is of particular in order to avoid any significant impact on the MPPT and

current control of the inverter’s input stage.

88

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 5.8: Proposed PV simulator and its control strategy.

To see the detail of voltage control as shown in Figure 5.9, we assume PV simulator
works at a point A originally. When the equivalent load of simulator changes from R1 to
R2 , the operation point will change to point B at first since there is a capacitor at output
side and the output voltage could not change immediately. According to the voltage
control rule, current I R2 + A11 is measured and the voltage reference VR2 + AV2
(point C) will be given after looking up I-V table, as a result, operation point will move to

point D.

The relationship between AV1 and AV2, A11 and A12 is:

89

:_ .
KR 1 V1 (5 7)
K A11
PV 1V2 (5 )
IAV2I =IAV1I-IKR /prI (5.9)

Where K R is the slope of the load curve, K 191/ is the slope of tangential line at
supposed operation point on PV curve. It is clear that IK R / K PVI plays a key role in
the adjustment procedure of operation: if IK R / KPVI<1, and AV2 < AV1, operation
point right on the current-voltage curve of PV simulator (point B) will be achieved after
several adjustments; if IK R / KPVI>1, and AV2 2 AV1, then the control procedure

would not arrive at stable point.

The equivalent load of PV simulator is resistive load, thus one can get K R as

 

d1

KR: RL = 1 (5.10)
dVRL RL

KpV = 5% = -—K1K2eK2V (5.11)

4

Where K1=Np1re, K2 =W.
S

Thus, one can get the critical point where IK R / K pVI equal to one.

90

1

——————=R
KZV C

2 (5.12)
K1K2e

RL

Where RC is critical load.

When R L > RC , voltage control will work effectively based on above analysis; on

the contrary, when R L < RC , current control will have a better performance.

 

 

 

 

 

 

I A PV Curve (va) . R2 (KR)
C B
E 1A1,
1R2 ———————————— —————————— _____ __
A12 D l / R1
AV2 AVl /,-”
BL ,, ’ A
’/’4’
’,---7’ l
,,,-” I
,-” 1 -
0 VR2 V

Figure 5.9: Analysis of voltage control.

This thesis applies a hybrid control strategy based on above analysis. The PV curves
were divided to three regions shown in Figure 5.10. When R L 2 RV , voltage control is
used. Voltage reference is given by measured current through I—V table. While

R1 < R L 5 RV , the measured equivalent resistance is used to look up reference voltage.

This is because at the maximum power point where IK R / K pVI equals to 1, using

91

either measured current or measured voltage to look up table would suffer an oscillation

(IAV1I =IAV2I), while the value of measured resistance is relatively stable no matter

how much measured voltage or current is. When R L S R I , current control method is

applied. In order to avoid frequently switching between two control method when R L is

around RV or R1, an overlap is set at each boundary (i.e. RV i AR, R] iAR) in

 

 

 

 

 

practice.
12 m . AR R1
................. _ . AR
‘ w
10 t , 1 RV
‘1' " _ AR
8 ' RL<RI RV<RL<RI AR
5?, ’ I’ \\
E 6 L I ll .:\
Q) .l [I I.‘
E ./ ,I :1
U 4 .l l’ 7" -
.’ I 2‘
z /’ R - 1
/ / L > RV '.\ ,
2 I’ . 1
'_l
-l
0 I l l I l l l l l
O 40 80 120 160 200
Voltage (V)

Current Control

Voltage Control
(Resistor)

Voltage Control
(Current)

Figure 5.10: Combined voltage and current control divisions

92

The simulator act electrically as real solar module for a consumer load, but take the
energy from the public grid instead from the sun. The advantage compared to a real solar
module is, that the PV simulator makes its power available independent from time and
weather situation and additionally the user can change the values of the simulators in
wide range. So certain conditions can be produced independently of the weather exactly

when they just are required.

5.3.3 Hardware Designs

To build up a PV simulator, the inductor and capacitor need to be designed. The
requirement of inductor design should follow the way that the converter need operate
under the continuous conduction mode. Also, satisfy the power rating and current ripple

requirement. The inductor should be calculated:

 

L 2 V(1-D) (5.13)
fAI pk
The output capacitor, to satisfy the voltage ripple requirement:
C 2 D1 (5.14)
fA Vpk

Where D is the duty cycle in which converter runs at maximum power,

f is the switching frequency,

AI pk is the peak to peak inductor current ripple,

93

V is the output voltage under maximum power,
I is the output current under the maximum power,

AVpk is the peak to peak capacitor voltage ripple.

5.3.4 Experiment Results

Experiment for the PV simulator is based on the Figure 5.11. The maximum output

voltage is happened on 400 V. And the maximum output current is 26 A.

 

 

 

 

 

 

 

 

 

 

 

 

30
100%

A 25 ~ 0 '
3 ”87/0 16.7ohm
E 20- 75%“ ..
E ‘ 18.50hm
U
.5 15
5‘
Q 10 -

5

0

2 59 116 173 230 287 344 401

 

 

 

Output Voltage (V)

Figure 5.11: PV simulator for resistor load conditions.

To test the PV simulator, we used the resistive load. Figure 5.12 shows the

experiment results for load change conditions. During the load change, the output

94

voltage and current still follow the PV characteristic outputs. Figure 5.13 shows the test
result during the solar irradiation changed from 100% to 87% and its reverse conditions.
Similarly, Figure 5.14 shows the test result during the solar irradiation changed from

100% to 75% and its reverse conditions.

Figure 5.12: Load current change.

 

 

 

I va (100V/div) I J

 

I . . u . . . : . . . .
1 II." . 1 . 1 : : :
. q I - 1. r u

++177r4 +7~1791i veer-'7‘: r l r j r r v v 1 , ,

 

 

I Ipv (10A/div) I

 

 

 

 

 

 

 

 

(a). From 20A to 25A.

95

Figure 5.12 continues:

 

 

 

t i 3 I va (100V/div) I

 

 

 

 

 

 

 

........

 

 

 

(b) From 25A to 20A.

96

 

 

 

 

Main: 10k

 

 

   
 

 

..................................................
-

“3‘s, iziéf'Iit I; I
mlllwllllllllfllllil . lg .l 5! ‘1 s
........ Iirl

...................................................................................................

 

va(SOV/div) ‘

 

 

 

 

 

 

 

 

 

(a). From 100% to 87%.

 

   

 

............................................

 

 

 

......

 

......... va(50V/div)

 

 

 

 

   

Ipv(5A/div) ' '

...................................................................................................

 

it

 

 

(b) From 87% to 100%.

Figure 5.13: Solar irradiation change around load R= 18.5 ohm.

97

F‘

Figure 5.14: Solar irradiation change around load R= 18.5 ohm.

 

 

Main: 10k

 

 

   
 

 

 

 

 

 

 

..............................................

 

 

 

 

............................................................................................

------

 

 

(a). From 100% to 75%.

98

 

Figure 5.14 continues:

 

 

 

 

Main: 10k

 

 

     

-----------------------------------------------

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

....... ' am...
L ............................................... .......................
......................................... . Ipv(SA/div) .....
(b) From 75% to 100%.
5. 4. Summary

The increasing number of PV inverters is coming to the market stresses the need to
carry out a dynamic characteristics under real conditions. In order to repeatable for the
laboratory, a PV simulator which is capable to reproduce the current-voltage output
characteristics of PV modules was developed. This circuit has the ability to simulate any
kind of PV arrays, especially suited for high power applications. The simulator is based

on 10 kW design, controlled by a DSP board. It allows testing PV inverters up to 10 kW.

The PV simulator suits research purposes as well as the test of device, that are fed by

PV cells, like battery charger controllers, inverters or dc-dc converters. The consumer

99

load can operate on each point of the I-V curve at the simulator and the characteristic can

be modified by the user.

100

CHAPTER 6. GRID-CONNECTED Z-SOURCE
INVERTER WITH PV SIMULATOR SYSTEM

6. 1. Introduction

6.1.1 Background

Because of the rapid development of rural area and increasing demand of clean
energy sources, photovoltaic (PV) based electricity production has one major grth
sector in distribution system. For most PV systems, grid connected operation are
preferred. Grid-connected system has the advantages over stand-alone system:

0 Reduced cost
0 Extended lifetime
' Higher efficiency

A grid connected PV system can either provide power to loads or feedback power to
utility line. But till now, even with great technology improvement of PV cell itself, the
cost of grid connected PV system is still high, which is now becoming the major obstacle
for the wide application of PV systems. Inverter constitutes almost 20% of the total cost
in a typical grid connected system [97]. The balance of system cost is becoming more
important since the PV array price drops. To deal with this problem, the most addressed
approach is to lower the cost of the inverter. One of the key issues now is to reduce the

cost of the power conditioning unit for the PV systems.

101

In order to connect the PV cells to utility, an inverter/converter based power
conditioning unit is required to transfer the dc energy to regulated ac. For the
grid-connected inverters, the general requirements are: low line current distortion, high
power factor, high efficiency, simple circuitry, high reliability, and most importantly low

cost [66, 67].

6. 2. Proposed Grid-connected Z-Source Inverter System

For PV system, the output power and voltage typically depends on a variety of
uncontrollable factors, such as irradiation intensity and temperature. Also, the output
voltage of PV cell changes with the loading current. The typical change rate can be as
high as 1:2. So to achieve the best performance of a PV system, besides converting DC
power to AC, the power conditioning unit will need to be accommodation a wide input
voltage range and realize maximum power utilization.

For high power PV system, traditionally, there are two basic circuits shown in Figure -
6.1. The first one is PV array plus inverter plus ac grid. The PV array is directly
connected to the inverter [68]. For this case, if to feed a 1:2 ratio PV array, generally a
higher number of PV cell in series is needed to accommodate the minimum required dc
voltage for a traditional inverter to output constant voltage. Since the number of PV cell
is increase, the maximum dc input is also increase, which results in higher maximum
voltage stress and high cost on the switching devices. So to reduce the overall system

cost, the system show in Figure 6.1 (b) is often used in real applications [69, 70, 71, 72].

102

DC/AC
Inverter

208V 3P

 

 

PV 7;
Array

I—_‘ I ——————————— I 7K: 69"

350-7oov L 1200v

 

 

 

 

 

 

 

 

 

 

 

 

(a). DC/AC inverter to grid.

DC/AC
Inverter

 

 

 

 

 

 

I
I
E 208V 3P
_ C i W! I fly
pV D IDC 1
l /\/

 

 

 

 

 

 

 

 

 

 

 

l
I
|
|
l
1
Array Converter I '
I «m \/\/
l
|
175-350v :
|

(b). DC/DC converter plus DC/AC inverter to grid.

Figure 6.1: Traditional medium power stage grid-connected PV inverter system.

In this case, a boost converter is added in front of the inverter. Thus, the minimum

dc voltage from PV cell can be lower. And the maximum voltage stress on the switches

103

can also be regulated to the minimum level. So the cost of the total system goes down.
But the additional dc/dc converter usually increases the cost and power loss of the power
conditioning unit, and at the same time lower the reliability of the system.

So to deal with this problem, this thesis proposes a Z-Source inverter based power

conditioning unit to utilize grid-connected PV system.

 

 

 

 

 

 

 

 

 

 

 

 

208V pv
3 .
phase ac Simulator I Z-Source Inverter Circuit
Breaker _
. ’an Vaberc __ 1'1
7 [m
- \%
swj— 208V
3 phase ac

Figure 6.2: Grid-connected Z-Source PV inverter system configuration.

For grid-connected inverter applications, the power quality is important. To avoid
the distortion of the utility grid, the injected currents to the grid should have low
harmonics and a high power factor [73]. Furthermore, when the output currents are in
phase with the grid voltage, the maximum active power is achieved by minimizing the
reactive component. The inverters are expected to have high power quality, high
efficiency, high reliability, low cost, and simple circuitry. Compared to single-phase
inverters, three phase grid-connected inverters have much more advantages.

The grid connected PV energy can be transferred to a grid through one or two power
conversions. In PV applications, the proposed one stage power conversion is more

efficient and reliable than multi power stage conversions.

104

6. 3. Control Issues

6.3.1 General Issues

To enhance development of PV grid-connected system, there are the control issues,
such as the input power control, improvement of the strategy of maximum power point
tracking, reduction of the voltage and current ripple, improving the total harmonic

distortion (THD), and increasing the system efficiency.

The input power of the PV inverter system can be controlled in different ways
depending on different control methods. The main task is to transform PV dc voltage to
the inverter output ac voltage with the desired amplitude and frequency, also has to

perform the maximum power tracking.

If the system is two power conversion stages, which means that a dc-dc boost
converter is utilized before dc-ac inverter, the dc-dc output voltage has to be controlled.
To achieve this goal, there are two methods usually used. The first one is the dc voltage

control, and the second one is dc current control.

For dc voltage control, the reference value is set to be output voltage. On the
contrary, for the dc current control, the reference value is set to be output current. Both

these two control methods will affect duty cycle of dc-dc converter, thus to get the PWM

accordingly.

105

6.3.2 Voltage Control And Current Control

To control the input power of the grid-connected PV system, there are generally two
basic control methods. The first is to get a fundamental 60 Hz voltage wave through the
control of switching instants. The power flow is realized through the control of
amplitude and phase of output PWM which are related to the line voltage, thus produce
the necessary voltage across the ripple inductance and get the desired current flow. The
second one is to directly control the current flow through instantaneous current feedback.

The ac voltages at the switch do not need to be directly controlled.

Voltage control is a simple method of power flow control. However, it has
disadvantages that current harmonics and over currents can not be directly controlled, and
the transient response is limited. While current control has better transient response,
also can reduce harmonics, inherent over current protection, and has sofi start capabilities

too.

In the grid-connected system, the load is utility grid which has infinite capacity, so
the grid connected current can be controlled. Then the outside of inverter can be
derived as a current source. So the current control has the advantages than the voltage

control.

6.3.3 Modulation Methods

In the power electronics converters and inverters, the pulse width modulation

methods are utilized to implement the control. The methods of pulse width modulation

106

method will affect the inverter energy efficiency, waveform quality, and voltage linearity

a lot. .

There are three main modulation methods: the carrier based PWM methods, the
hysteresis current control, and programmed pulse modulation methods. For the
programmed pulse modulation methods, the switching patterns are calculated for a
specified performance Optimization at beginning and stored in a memory which can be
accessed through a look-up table. This method has a disadvantage that the huge
computer memory resources are needed and the number of pulses per fundamental cycle

is limited to a small number.

Compared to the programmed pulse modulation methods, the carrier based PWM
methods can operate at the higher switching frequency, at the same time can achieve high
quality output waveform. The first carrier based PWM was implemented by Schonung
and Stemmler in 1964 [77]. As shown in Figure 6.3, the reference modulation wave is
compared to the triangle wave, and the switching times are controlled by the intersections.
Based on this fundamental, this thesis proposed a new control which is depicted in detail

in the next section.

107

_.4|liIiIIIi|iIrI ,,
‘Illllllll‘l’ “”

SLIM III I I ll HUI—Il—Ifll—I—Ifll—II ,

cot

O

 

Figure 6.3: SPWM method.

Also, the hysteresis control method is used in some applications. Compared to the
hysteresis current control, the proposed control method has some advantages.
Considered about hysteresis control shown in Figure 6.4, the error between the desired
current value and feedback current value is the input of hysteresis loop comparator. The
output of hysteresis loop comparator is the control signal to the power devices. The
harmonics components will be different according to the bands width [76]. If the
hysteresis bands width is too big, the harmonics components of grid current will be larger.
Otherwise, if the hysteresis bands width is too small, the harmonics components of grid
current will be smaller, the switching frequency will be higher. The output spectrum of

grid current is wide due to the changed switching frequency.

108

la,b,c

Av AW \W/A V/

Hysteresis bands width

 

 

Figure 6.4: Hysteresis current control.

6.3.4 MPPT Methods

Because of the PV’s nonlinear characteristic, the use of MPPT can make full use of
the system and thus reducing the cost of the system. The function of the MPPT is to
make the PV array work as close to the maximum power point under any circumstances.
There are many methods to realize the MPPT, such as perturbation and observation (P&O)
method, incremental conductance method, parasitic capacitance method, constant voltage,

and so on [14].

For the perturbation and observation method, the maximum power tracking operates
by changing the solar array voltage shown in Figure 6.5. If a perturbation voltage is
applied to the PV array, the corresponding perturbation (dP/dt) is caused at the same time.
Therefore, the maximum power tracking can always seek the maximum power conditions
by tracking the perturbation. The perturbation will be updated every single cycle.

When approaching the maximum power point, the output voltage of PV array will

109

oscillate around the operating voltage. Thus a power loss is generated. The
perturbation difference has the effect on the resulted power losses. In this method, the
worst case is that perturbation in one direction will lead to an operating voltage far away
from the actual maximum power point if the system has been oscillating around the

maximum power point.

 

 

Sense V(k),l(k)J
I

 

I P(k)=V(k)|(k) I

 

 

 

 

 

 

I D=D+AD I I D=D-AD I I D=D-AD I I D=D+AD I

C .i 3

Figure 6.5: P&O MPPT method.

 

 

 

 

 

 

For the incremental conductance method, the PV array voltage varies according to the
maximum power point voltage by changing the incremental conductance or instantaneous
conductance. This method has advantages over the perturbation and observation

method, it can track quickly changing conditions more effectively than the perturbation

110

and observation method. But this method will be more complex than the perturbation

and observation method. The INC flow chart is shown is Figure 6.6.

 

Sense Vk I k

 

 

 

 

N=Q-h4
AV = Vk —Vk_1

 

 

 

 

Yes
AV=O?

No

 

Yes £V+I=Oi7

A

Yes

 

 

N0

NO NO

 

 

 

Yes

 

 

 

 

Increase Iref

 

Increase Iref Decrease Iref

 

 

 

 

Figure 6.6: INC MPPT method.

The parasitic capacitance method is similar to the incremental conductance method.
This method considers the parasitic capacitance in the PV array. The switching ripple
of the maximum power point is used to apply a perturbation to the PV array. Then the

array conductance can be calculated according to the measured current ripple and voltage

111

values. Therefore, the incremental conductance method can be used to achieve the
MPPT. This method may not be effective when the parasitic capacitance is small.
Thus, this method is suited for large PV arrays that have relatively large parasitic
capacitance. In addition, the dc/dc inverter usually has an input capacitor to filter out
small ripple in the PV array. Thus that capacitor will complicate the parasitic

capacitance, thus limiting the use the method.

The constant voltage method assumes that the operating voltage at the maximum
power point varies little with irradiation levels. The operating voltage at the maximum
power point is usually selected to be 0.76 times of the open circuit voltage of the PV
array. In this method, the open circuit voltages of the PV array are updated and set to

76% of the measured open circuit voltage accordingly.

Among these methods, the incremental conductance method has a higher overall

MPPT efficiency than others. The constant voltage method has the least. MPPT

accuracy.
6.3.5 Synchronization In Grid-connected Applications

6.3.5.1 . Synchronization methods

Synchronization is a big issue in grid-connected applications. There are classical
two ways to solve the synchronization problem: from hardware or from software
approaches. Accurate synchronization is very important in distributed ac power systems

and is a key requirement in any real time measurement and control system [74, 75].

112

The most famous synchronization is the phase-locked loop (PLL) control. The PLL
was used in a lot of applications until the development of integrated circuits, thus to make

it easier to perform the PLL [76].

A PLL is a device which can track with another signal. With PLL, the output can be

synchronized with the desired reference in phase and amplitude.

Figure 6.7 shows one method of a three phase PLL control loop.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

at: —*k°/S
Vd l 6
—+>O——> kp “S *——’
Vd
Vq *—
Va ——>
Vb ___,.
VC ——-—>

 

 

 

Figure 6.7: PLL control 100p.

113

6. 4. Proposed Control Strategies

6.4.1 Proposed PLL Synchronization

The proposed system used zero crossing detection method for PLL control. For zero
crossing detection PLL, there are hardware loop approach and software loop approach.
In this thesis, digital zero crossing detection PLL is implemented to synchronize with the

grid. Figure 6.8 shows a zero crossing detection PLL diagram.

 

 

 

Yr
_, Phase Low pass
_, Detector ’ Filter ’ VCO ‘ ’

 

 

 

 

 

 

 

 

 

 

 

 

Figure 6.8: Zero crossing PLL.
For phase detector, there can be classified to sinusoidal wave and square wave phase

71'
detections shown in Figure 6.9. The former has the detection interval from — 3 to
72'

3. The latter include triangle phase detection which has the detection interval from

H 7f . . . .
— 3 to 3 , the sawtooth phase detection whrch has the detection 1nterval from - 7! to

fl [77, 78, 79]. The sinusoidal phase detection was developed in this thesis.

114

 

sine

 

------- triangle

 

 

Saw tooth

 

 

   

 

Figure 6.9: Phase detection methods.

The zero crossing detection PLL has the advantages that it has the immunity to
harmonics in the input signal. The dynamic response of the zero crossing detection PLL

can be improved by adjusting the low pass filter and system parameters.

With the proposed PLL, we can synchronize grid very well.

6.4.2 Proposed Harmonic Injected Feed-forward Control

Grid-connected PV system has two main control requirements: seeking the maximum
power tracking and obtain the sinusoidal output current to the utility grid. In the
proposed solution, these two goals are realized in a single power conversion stage with

two control loops.

The inner loop realizes inverter current output control, which consists of grid voltage

and current sampling, phase calculation, and PWM generation. Usually the inverter is

115

controlled so as to generate the output current in phase with the grid voltage to achieve

the maximum active output power by minimizing the reactive output power.

The outer loop is a power control loop, which tracks maximum power point of PV
modules by adjusting inverter output power instantaneously. Current MPPT methods
are often based on perturb and observe, incremental conductance, parasitic capacitance,
voltage based peak power tracking, and current based peak power tracking [80, 81].
The modified P&O MPPT method is implemented and would be shown in detail later.
The peak power tracker operates by periodically increasing or decreasing the PV array
voltage. If a given perturbation causes an increase in airay power, the subsequent
perturbation is made in the same direction. Thus, the peak power tracking can be

realized [82, 83].

6.4.2.1. Third harmonic injection

Third harmonic injected control method was first utilized by Buja and Indri. In this
method, the inverter output voltage can be maximized by adding a triple harmonic which

has one-sixth of the fundamental component modulation wave [78] shown in Figure 6.10.

Vaa Vbb Vcc

 

1.00
0.50

-050 _ ‘5

 

 

 

-1 .00 j 2
0.00 1 0.00 20.00 30.00 40.00

Time (ms)

 

Figure 6.10: One-sixth of the fundamental injected.

116

If the modulating input signal exceeds the amplitude of the carrier waveform, the
output will saturate. For three phase application, the line to line components are
immunity to the added third harmonic injection, but the fundamental amplitude can be

increased. Also, the switching losses can be reduced.

To connect the PV array with 1:2 voltage change ratio to 208 ac line, when the PV
array voltage is low, maximum constant boost control [84] for the Z-source inverter is
used in the pr0posed system. There are three boost control methods listed in table 6.1.
The shoot through interval, boost factor and voltage gain are compared in the table. The
maximum constant boost strategy configuration is shown in Figure 6.11. With certain
amount of third harmonic to the phase voltage waveform, the line amplitude of the
firndamental wave will increase without any over modulation [85]. Also, the output line
to line voltage wave still remains sinusoidal and undistorted.

Table 6.1 Shoot-though interval, boost factor and voltage gain of three boost control

 

 

 

 

 

 

 

 

 

 

 

 

 

methods.
D B G

Simple boost l — M (Max value) 1 M

2M —1 2M -1
Maximum boost 27: _ 3J3M 7: M7:

271' 3‘\/§M — 71' 3‘\/_3_M - 72'

Maximum boost J3 M 1 M
(with third harmonic l - — '—
injection) 2 JEM —1 fiM —1

 

In these three boost control methods, the modulation index must satisfy the following
requirements.

For simple boost control, the modulation index must be less than l—D , and

Z-Source capacitor voltage must be larger than 29(16-

117

 

For max

27r
imum boost control, the modulation index must be less than fig — D),

3 3 .
and Z-Source capacitor voltage must be larger than ——vac .
71'

For maximum constant boost control, the modulation index must be less than

2 .
E (1 — D), and Z-Source capacitor voltage must be larger than fivac-

 

 

 

 

VP
V I I / \
I I
b
V. :\- I
0 I1 275
_3 / 7

 

 

 

 

Sap u

E

 

_fl

2

 

 

 

 

 

 

mi /
E

 

 

 

 

 

 

 

 

 

J W
W LWJTEHJ‘L
Scp UWJ‘Lfl_fl4‘ rum Ifll
Sanu ULJT_IT_I‘LF_I_ ‘I
Sbn II I LFLT‘ mflmmfr
Sen l_fl_lfl_IT—LITJ HILr‘I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 6.11: Third harmonic injected maximum constant boost.

118

There are two periods in a switching cycle:

7!
For the first period (0-3- ), the upper and bottom curves can be expressed by

following equations respectively.

Vp1=\/3M+sin(6—-2—3£)M 0<9<§ (6.1)
. 27: 7t
an = Sln(9 - ?)M 0 < 0 < E (6.2)
_ 7r 27r . .
For the second penod ( -3— — —3—) , the curves meet the followrng equations
respectively.
V — sin(6)M i < 6 < 21 (6 3)
P2 3 3 '
Vn2 = sin(I9)M — J3M §< 6 < 337: (6.4)

Obviously, the distance between these two curves are always constant, that is 45M .

The Boost factor B and the voltage gain can be calculated:

1 1

B: :
1419 JiM—r

 

(6.5)

119

Figure 6.12 is a plot of equation (6.5) with the modulation index changing from 0 to

1.154. It can be seen that the voltage gain can be close to infinity when modulation

index is close to J3 / 3 .

 

 

 

 

250 j I I I T I
200 — I I
l
150 r [I ‘
100 . ,‘I .
m
50 — it .
; L.
I fimm~ - a -
0 152a. Izzarrirlrrsarrnsraarrsmagé ’ “WIRIHHEWIJ *
“r
-50 _ II -
2i
_100 l I I l I I
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
MI

Figure 6.12: The relation between the voltage gain and modulation index.

In this kind of control, the shoot through duty ratio is always constant while the
maximum voltage gain can also be achieved. When the carrier triangle wave is greater
or lower than the shoot through lines, the inverter is gated to a zero state, working as
shoot through mode, just like Figure 6.13 (a). Otherwise, the inverter works as the
traditional PWM inverter, which seems to a non shoot through mode, just like Figure

6.13 (b). One advantage of this method is to eliminate the ripple at line frequency.

120

 

 

 

 

 

 

 

 

 

 

 

 

(b) Non shoot through mode.

Figure 6.13: Z-Source inverter operation modes.

121

For the traditional three phase PWM, the average squared value of the current ripple

can be represented as [A- l ]:
0032 (6 + 365) — 3J3M 3 cos3 (6 +165)

2 2 71/3
Aizb =IK4£I Lul- I I—JEM4cos(9+£)(cos3(e—2—”) >610 (6.6)
a 2L 48 fl' "275/3 6 3

—cos3 6

L J

 

 

Then the above equation can be simplified to the following equation:

2 2
V AT
113mm: dc 2 ——I3M2— 4——‘/§M3+ +9M4I (6.7)

192L 2 7r 3
For the third harmonic injected constant boost PWM,
Similarly, the equation can be derived:

2 2

V AT

12 d—C— 3M2— 4——‘/:M3+ 31M4 (6.8)
hrms: L2192 2 7, 32

Based on these two equations, the harmonics distortion factor of the two PWM

strategies can be defined as:

Y(1)=3M2— 4——‘/_M3+ 2M4 (6.9)
2 7r 8
Y(2)=-:-M2—4—:_M3+ :éM“ (6.10)

Where Y (I) is for the traditional PWM and Y (2) is for the THICB PWM.

122

Respect of the switching frequency, dc voltage, and inductance, the relation of
harmonic current distortion factor and modulation index can be drawn in Figure 6.14.
In the figure, the curve Y (2) represents of third harmonic injected PWM, and the curve
Y (1) represents of traditional three phase PWM. It is clearly that the THICB PWM
would start to have significant less harmonics distortion factor when modulation index is

bigger than 0.5.

For the specific case that the dc voltage changes from 230 V to 400V, equation 8 and

9 are plotted at the condition that: T =100 uS and La=l mH .

The plotting results in Figure 6.15 clearly show that the current harmonics content in

THICB method is much smaller than in the traditional method.

 

0.7

 

6.. _ g
' Ya) 6"

 

 

 

0.5

0.4

0.3

Y(M)

 

0.2

 

 

 

0.1

 

 

 

1.4

 

Figure 6.14: Relationship of the harmonic factor and modulation index with different

PWM.

123

 

Square Ihrrns

 

Square Ihrrns

 

b.: Harmonics current content for THICB PWM.

Figure 6.15: Relationship of the harmonic factor with modulation index and dc bus

voltage with and without harmonic injection.

124

For the Z-Source inverter, the current through the inverter switches is composed of
current to the load and the current through the switches during shoot through. The
current during shoot through in average is distributed in balanced three phase paths.
The current through the inverter during the shoot through is twice of the inductor current.

The average current in shoot through for switch is

2
Iavs :31]; (6.11)

The average current through the diode is equal to the sum of the average current
through inductor and capacitor. During the steady state, the average current through the
capacitor is zero, and the average current through the inductor is equal to the current

through the diode.
ID=Pm/V,- (6.12)
Where Pm is the PV array maximum power.

Without shoot through state, the average current is the same as traditional inverter,

«510m

 

 

 

I = 6.13
avsn 3V0 cos I117: ( )

Thus, the average current in total is

[M =-2—IL(T0/T)+ fip’" (l—TO/T) (6.14)

3 3V0 cos l/lfl'

Under constant maximum boost,

T0 = (1 —MJ§/2)T (6.15)

V.
V0 = M l (6.16)

«BM—12.5

125

The average switch device power is

 

T0 Pm
SDP- =61 V =41 V- F. +8 l—T /T

_ 2pm (2 —\/3M) + Nip",
J3M —1 7r

 

6.4.2.2. Current Control Loop

In grid-connected distributed generation systems, three phase pulse width modulation
voltage source inverters are usually employed to achieve power conversion, grid
interfacing and control optimization. To feed grids with high quality power, the current
control of the grid-connected VSI plays an important role since a DO system would not

regulate the voltage at the point of common coupling (PCC) [86].

Due to PCC voltage can not be controlled, the power quality is determined by the
current quality only. To achieve unit power factor, the output current of the inverter
needs to be in phase with the utility voltage. The grid-connected three phase dc/ac
Z-Source inverter can be shown in Figure 6.16. And Figure 6.17 shows the unity power
factor control of the inverter current to the utility grid. The grid line to line voltages are
measured, also the phase lock loop is utilized to ensure system synchronization [87, 88].
Two currents are measured from the outputs of the inverter then fed back into a close

loop system. The close loop system is realized with a PI controller.

The PI controller is based on a transform function C(s) as follows:

F(s) , k,-
= k + — 6.18
15(6) 1’ s ( )

 

G(s) =

126

Where F (s) is the output, E (s) is the input error, k p is the proportional gain,
and ki is the integral gain.
The above function is an analogue form and need to be transformed to the equivalent

digital one before being implemented by the DSP controller.

The final form of the digital PI controller can be represented as:

F(n) = Yp[n] + mm] (6.19)
Where
Yp[n] = k pE[n] (6.20)
YiIn]=%E[nl+Yi[n-1] (6.21)
Y

p and Y1: is the outputs of the proportional and integral loops respectively, and

f is the sampling frequency.

Grid
Va Vb vc

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 6.16: Grid-connected three phase Z-Source inverter system.

127

 

 

Load

 

 

 

 

 

 

 

PV W Z-Source

Filter CB Grid
SIM Inverter —

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Vab
Vbc

 

 

 

 

 

 

Ia
Ib

 

 

 

 

 

 

 

 

Figure 6.17: Block diagram of control loop.

In Figure 6.16, the output voltage of inverter is:
di
v-=v +L —“+i R (6.22)
l g a d t a
Afier the transform of Laplace, the output current is:
z' (s)— 1 (v-(s) v a» (623)
‘1 sL + R 1 g '

Where vi is PWM output waveform of the inverter. R is equivalent resistance

for the inductors and circuit.

128

Since unity power factor control, the command current is:

c* * o
la = I a srn 6a (6.24)

So the command voltage of inverter becomes:
* ki =1: _ .
v,- = (kp +—)(Ia srn6a —za)+vg (6.25)
S

If no voltage feed-forward, when the grid voltage suddenly increases, the grid
connected current will decreases, thus the deviation occurs. Compared to with/without
voltage feed-forward control methods, the grid voltage feed-forward compensation can
work more effectively during disturbance. So in the current error compensation part,
the feed-forward grid voltage is used to generate voltage references for the PWM, thus

the disturbance due to grid harmonics can be suppressed well.

The system open loop transfer function is:

 

G(S) = :1 (6.26)
2(RCLs + Ls + R)
The PI transfer function is:
ki
Gp1(s) = kp + — (6.27)
S

Based on the following parameters,

Input voltage is 230 V,

129

Inductor is 1 mF,
Capacitor is 50 uF,

Rresistor is 5 ohm,

kp is 0.04, k,- is 60.

With the open loop transfer function, the system bode plot is:

 

 

 

Bode Diagram

 

40 7"
30%
20
10

< I
“l——O-—'-fi-l——.‘-t-I
I I ,‘

-10
-20
~30

Magnitude (dB)
0

-45
-90
-135
-180

Phase (deg)

 

T! r r Tll'zfi

..
gilt; f E‘Iléii‘

y,l§- ’ I 4 I 'i"
.4_._1——a——.—.-—L_ ————— '--—J aaaaa -—‘

 

 

 

 

Frequency (rad/sec)

Figure 6.18: Open loop bode plot.

The PI correction bode plot is:

130

 

Bode Diagram

 

 

 

 

 

    

 

 

 

  

 

 

 

 
 

 

 

 

 

a 5 Tu - - - -
- .3, O .F
I, .. . 1 u .
HUHHHHHHH w- .............
.- - -_ ............. ..
_ _
v ..
_ . .2
.6 W 4.0 L m
rI I 4 l O .. H
a H. r 1m; w .
_..-L e WHHH ..... ..
. ... 0 .m ..,. ,. a.
. a w b m .
,1" J 6 m a.
. a .fi .1 .— ”J
.2- ”. 3O (r\ m D w:
wl I A l H e .. .w
m _. y d .. r a,
e ._ m w o . .- ..
.7 m. e H B W n ”m
T I" ....................... u .. .. . a
m . a. 9 .m M M H.
.2 - a m 1|” t w -_, 1....-.11 w. IIIIIIIIIIIIII 1
“ H F 6 1m .. .. .. ”- ,_ m. ,. . _.
..-,_ ,... M m We ® .-- - flux. .H
I. m .m - _._
-.__ m 2, ,m .
-L. W .- T: :-::...
O O m a .. if. .- I. P. . N. .3 - __
3 2 s 5
m- m m 6 m. m... mm m. u m.
as 632662 $63 as: e . .
n 83 sea“: 93 as:

104

131

Frequency (rad/sec)

10 3
Figure 6.20: Close loop system bode plot.

102

The inverter feeds the sinusoidal current into the utility grid which its frequency and
phase is the same as utility gn'd. Thus, the unity power factor is realized. By
regulating the current loop parameters, tracking speed and tracking error can both be

modified [89].

From above, the grid-connected inverter with high power factor is realized and the
configuration of the system is very simple. This control configuration has higher
efficiency, good performance, and reliability because of its single stage. . Also the power

can be controlled well for its voltage forward feedback.

For the output filter, there are L filters, LC filters, and LCL filters. Among these
three types, the LCL filters is a third order filter which is easily to cause the stability
problem. The filter inductance is designed by the consideration of the utility grid
voltage, the switching frequency, the amplitude of grid connected current. The output
grid connected current may be distorted for the small inductance. Otherwise, the power

loss on the filter and damping and time-delay will increase also.

6.4.3 Modified P&O MPPT Control

PV inverter operation status will be affected by operation range on the PV curves.
Figure 6.21 shows the constant power curves for different values of power. From this
figure, it can be seen that there exist two equilibriums for a given power point. When

the constant power curve is tangential to the PV l-V curve, the maximum power point is

achieved. If the given P is higher than Pmax, there is no real solution and the

132

operating point will approach to the short circuit. So there exists the possible two

solutions for a given P.

4.0 '-

 

3.5 -
3.0 h
2.5 —

2.0 '

0.5 -

 

 

0.0 l l l l 1 I J L l l l ;

O 2 4 6 8 10 12 141618 20 22 va

Figure 6.2]: PV I-V and constant power curve.

From the Figure 6.22, it can be verified that the solution to the left intersection of
constant power curve with I-V curve is unstable; on the contrary, the right side is stable.
The voltage would collapse if the operating point is pulled toward to the left side. So

the PV inverter should operate on the right side.

133

4.0 ~ V /

3.5 '

V

 

H

3.0 "

2.5 '-

 

2.0 "
1.5 "

1.0 L1:

0.5 -

V
A

 

 

 

0.0 l l J l L l l l J l k

0246810121416182022va

 

Figure 6.22: PV I-V and constant power curve analysis.

Generally, the PV inverter starts from open circuit condition, then draws the power
from the PV panel and at the same time approaches to the maximum power point, finally

oscillates around the maximum power point.

Since the proposed Z-Source based PCS is directly connected to the grid, the PCS
will be controlled to transfer maximum power from the PV array to the grid all the time.
Because of nonlinear characteristic of PV models, the maximum power can not be
achieved by directly connecting the PV models. Tracking of the MPP must be used to
effectively get the maximum output power. Thereby, many researches of the MPPT
have been done [7]-[10]. There have been perturbation and observation method, the
incremental conductance method, and the hill climbing method, etc. Here a simple

power feedback method can be used for Z-Source based PCS to achieve MPPT as shown

134

in Figure 6.23. The power can be measured and used as feedback. The PV modules’

voltage can be regulated to an optimal point, which presents the maximum power.

 

 

 

PV

Array PCS .

  

 

 

 

   
 
  

 

 

 

 

 

Switching

 

mqpaaa
1nd4no

 

iuauremseaw
ramod

MPPT

 

 

 

Figure 6.23: The block diagram of the MPPT control.

In this thesis, the modified P&O MPPT method was implemented. Figure 6.24

shows the modified P&O MPPT method.

135

 

l Sense V(k),1(k) I

l

I P(k) = f[V(k),1(k)] l

 

 

Yes No

 

 

l AI(k)=AI(k—1) l l AI(k)=—AI(k—l)

I i
I

 

 

 

l 1(k) = 1(k — 1) + AI(k) l

l

P(k) = P(k - 1)
AI(k) = AI(k — 1)
1(k) = 1(k — 1)

l
C Return 3

Figure 6.24: The modified P&O MPPT method.

 

 

 

 

 

6. 5. Hardware And Software Implementation

6.5.1 Hardware Implementation Of System Setup

Testing of the 10 kW Z-Source inverter began with low power operation while

verifying the stable, safe operation of the inverters. The input power was supplied from

136

a PV simulator. This allows for the complete control of the input parameters,

independent of weather conditions, which is necessary to demonstrate the inverter

performance over a wide range of normal and fault conditions.

The following equipments were utilized in the testing of the whole system:

Rex manufacture Transformer 1, 10 kVA, three-phase, 230 V Delta to
260/460 V Wye.

Sylvania Transformer 2, ll kVA, three-phase, 460 V Delta to 460 V
WYE.

Powerstat Variac I, lOkVA, three-phase, O-240V.

Powerstat Variac 2, lOkVA, three-phase, 0-280V.

Z-Source inverter, 10 kW.

DC-DC converter, lOkW.

Rectifier bridge, three-phase, 60A.

Three-phase grid, AC power source, 208 V.

Circuit breaker, three-phase, 30 A, 20 HP.

LEM current sensors, LA 2058.

LEM voltage sensors, LV 25.

Inductors, capacitors, and resistor bank.

Tek TDS 7054 Oscilloscope.

Tek AM 503B current proble amplifier.

Tek TDS 2014 Oscilloscope,

Fluke Voltmeters.

Partial of the equipments are listed in the appendix 2.

137

6.5.2 Hardware Of Control Unit

The control unit of the system consists of DSP boards, three current sensors, four
voltage sensors, gate drive boards, and power supplies.

The universal DSP control board shown in Figure6.25 was developed in MSU Power
Electronics and Motor Drives Laboratory. It is based on TMS320LF2407A DSP from
Texas Instrument (TI). The board was designed to fit controls for all kinds of inverter
and converters. To maximize the control possibility of the board, a Xilinx CPLD is
used in series with the DSP chip to expand the logical calculation capability and available
I/O numbers. To enable feedback control, the board also includes the analog signal
processing circuits for the analog input channels of ADC converter.

The main features of the board are summarized as following:

' LF24OF operating at 40 MIPS with 64K words of zero wait state memory for

debugging or data stored

° 16 channels Analog to Digital Conversion with signal processing circuits

0 Dual event managers multiple PWM and capture channels on chip

0 On chip UART with R8232 Drivers, 485 and CAN divers

° 32K words of on chip Flash ROM

' SPI interface data memory for parameter stored

° On board IEEE 1149.1 JTAG Connection for Optional Emulation

' 15V, -15V, and 5V power input, (onboard 3.3 volt regulators)

° Expansion Connectors (data, address, I/O, control and PWM out signals)

° Xilinx CPLD (XL95288-10TQ208) providing many control or out signals

138

The ADC module in 2407A chip provides a flexible interface with 16 channels to
event managers A and B. The ADC interface is built through a fast, 10 bit ADC module
with total conversion time of 500 ns. Since the ADC has a resolution of only 10 bits,
while the registers of the timer are 16 bits, scaling has to be done for result registers
ADCFIFOs before the values in them involve calculations. This is realized by shifting

the ADCF IF 0 register result to the right.

 

Figure 6.25: Universal DSP 2407A board.

The DSP chip communicated with the program in the computer named Code
Composer Studio. Code Composer Studio interfaced to a JTAG emulator pod, which
was connected to the DSP. The JTAG emulator allowed for incremental stepping of the
control code for eased debugging. The JTAG emulator was also the way to download

the program to the DSP chip. The structure of the control system is shown in Figure

139

6.26. The control system is isolated from the power stage by using plastic optical fiber

and isolated sensors.

 

 

 

 

 

 

 

Fiber Optical
PWM output
Volta e sensors
--> :5
DSP Board Drive
Boards
--> it
[PM Device

 

 

 

  

Fiber Optical
Fault output

pc DSP XDSSlO
Emulator

 

Figure 6.26: Control Unit diagrams.

With the advantage of DSP technologies for power electronics applications, analog
feedback and status sensors are fed directly into the DSP and multiple PWM outputs
directly provide the drive logic for the converter and inverter power switches.

The PWM controller maintained the output current waveforms within the

performance specification requirements for THD and unity power factor.

6.5.3 Software Implementation

As mentioned above, the software environment is code composer studio for DSP, and

Xilinx for CPLD. For code composer studio, it has the features [90, 91]:

140

0 Integrated editor, debugger, profiler and project manager

° Probe points connected to the file I/O, graphical display

0 Fully integrated Code Wright Editor

° Source code debugger common interface for both simulator and
emulator targets

° DSP/BIOS Host tooling support

In the code composer studio environment, all the DSP code for experiment were
developed and written in assembly language. In the XILINX environment, all the
CPLD code was developed and written in ABEL language.

The control of the DSP is composed of an initialization routine followed by periodic
interrupts triggered by an internal timer. These interrupts occur once per switching
cycle and set the reference. This reference is then sent to the PWM modulator on the
DSP so that it can be transmitted over the optical fiber. A flow chart describing the
operation of the interrupt is shown in Figure 6.27. The system control flowchart is

shown in Figure 6.28.

141

n—

 

Cnterrupt Handler StarD
/ Sensor Process /

Y

 

 

 

 

Control Process

 

 

Protection

 

Update PWM output

Interrupt Return

 

Figure 6.27: Interrupt handler flow chart.

142

Figure 6.28: Flowchart of DSP code.

 

| System configuration

1

     
   
 
     

 

Sampling period
flag set?

 

 

 

 

 

 

 

 

 

 

 

 

[ Configure and start ADCJ
1 Reset sampling
Set up GP Timers and period flag
Full Compare Units 1
I Read ADC and
. . . . restart ADC
Initialize variables

 

 

 

 

 

Reset flags

 

1

Clear INT flags
Enable interrupt

 

 

 

No
Filter ADC

Calculate Vab,Vbc

J.

Calculate
Vsa,Vsb,Vsc

    

Initialize ADC

 

 

 

 

 

 

 

 

 

 

143

Figure 6.28 continue:

      
   
     

Get Vra from
Vsa by BPF

——l

Calculate
theta_B,theta_3A

1

Calculate
sinA,sinB,sin3A

1

Calculate open
pwm reference

1

Add PI parts

 

 

 

 

 

 

No

 

 

 

 

 

    
  

Get final
pwm
reference

  
   
   

 

Send PWM out

 

 

Make
0<=tmp(theta_
A)<1 80

L

 

All the PWM generated from DSP need be controlled in logic before sending out to

the gate drive board. Figure 6.29 shows the CPLD control block diagram.

 

CPLD Control
Logic
\ . Latch= =

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EUN/s-iflor) __">I CH Lamb &
T’ ‘ 'm Current
Gateblock .___+ AND '—. Boost —> OUT-m
FAULT“ T+ GROUPl
+5 Combination
PDPINTA ‘
PDPINTB <

Figure 6.29: CPLD control block.

Also, all the fault protections are implemented and updated in the system.
Inside the IPM module itself, it has:
0 Over current protection
0 Over temperature protection
' And UVLO protections
Besides these protections, the following protections were implemented through DSP
and CPLD:
° Z-Source inverter dc over voltage protection

0 Z-Source inverter inductor over current protection

145

' Z-Source inverter capacitor over voltage protection
0 Inverter output current protection
The fault signals will be sent to XILINX CPLD chip, then the output will be shut
down, and the protection signal is fed back to the DSP board. At the same time, the

LED which indicates the fault condition will light on instantaneously.

6. 6. Simulation and Experiment Results

To prove the proposed system, the simulation and experiment are both performed
based on the PV curves shown before. The maximum power is achieved when PV array
voltage approaches to 330 V. The experiment hardware of the Z-Source inverter and
DSP control parts are shown in appendix 2. In the crossed LC network, the inductor is 1
mH , and the capacitor is 1.3 mF . The system switching frequency is 10 kHz. The
Z-Source inverter is connected to the grid with 60 Hz frequency, 208 V line to line
voltages.

The Z—Source PV inverter can also work as manual mode, which means the inverter
can be tested intended for maintenance and debug test function, to provide with manual
control of basic inverter function. In this mode, one adjustable meter on the DSP board
is available to adjust the inverter power level. The manual mode was utilized during the
testing of inverter in order to maintain operating conditions independent of the MPPT
circuit, and to create fault conditions. The output current was adjusted from 5% to
100% of rated output current power while the input varied from the specified ranges.

The simulation of the whole system is performed with PSIM and Matlab software.

Figure 6.30 shows the PV output current, voltage and power when irradiation changes

146

from 1000 W/m

2 to 800 W/m2 . Figure 6.31 and Figure 6.32 shows the grid line

to line voltage and Z-Source inverter line to line voltage.

IPV (A)

va (V)

18

l7
16

15
l4
13
12

ll
10

600' 1 - r r z 1.
500.- .............................................. ......... ..... .
......... ......... ..... .
- “new: 5 %
......... -------..-.- ......... ........... ..... ...
...............................................
............................................

l l l E l l I

Figure 6.30: PV output characteristic during irradiation change.

 

 

I T I I 1 I 1 r
h ooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo -
h ........ :c c u'c‘: ...... :5 "if f f .................................. q

 

 

 

but.oncofncvdoooon‘.cocoon-J......-c-‘Ioooooon; .................................. d
b oooooooooooooooooooooooooooooooooooo : ......... . oooooooooooooooooooooooooooooooooo 4
l l l l l I 1

 

 

 

 

 

 

 

 

(b). PV array output voltage.

147

va (kW)

Ia (A)

Ib (A)

 

Figure 6.30 continues:

 

, v-vv~—~"—w--uw—vtv-VTt-‘vr‘(Vi-"T—-VVj-—T—‘—w—va—-VWTvvvvvvv‘v-vvv
.
¢ . I .
I I

 

 

 

C
.
Q - .
o . o .
b ......................................... $ ..............................
I j o‘
. 4 .. u- -A— 2 . b
w
V . .
o
c ' s
' a
o
C .
b .................................. . ‘- --------------------------------- t ggggg ‘
o
o . I
c
. o
. ' ‘ g.
o b " A‘ A;
Q fl 7 —— h w
1 . - ' I
h ccccccccccccccc gun-oooooucnde-ug‘nunqa ................................... g
-
O . ‘
O . . .
a . . .
C
.
o
p ................................................... ‘ 0000000000000000000000 0"
o
O
C
o
O
.
O
C
D- ................................. . ......................................... q
.
a
o
.
ccccccccccccccccccccccc :Con-q.u.‘..c.uouo‘o.-u-vuuou....a..'ouoo-ooooonocq
.
o
O
C
o - . ’ O
1 . i J .- 1. .. 1 1

 

 

0.3 0.4 0.5 0.6 0.7 031(8)

(0). PV array output power.

Figure 6.31: The grid connected Z—Source inverter output current.

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

 

 

Ill! ' II
..E. . ................. 1
1 I I l i I r i
I l l l I r r g
,. i ............ E.
l i i
l 'I'I : 1 4
, -
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII ‘.
I l i 1 i l r i

 

148

 

(V)
s

 

 

 

 

 

 

32
“is:
(U
>
-500
A 500
3: ?,
51.2 0
o 1
O .
§ § _500 i i i i i i
0 0.1 0.2 0.3 0.4 0.5 0.6 1 (S)

 

Figure 6.32: The grid line to line voltage and Z-Source inverter line to line voltage.

For PV voltage changes from 230 V to 400 V, the Z-Source inverter can work
continuously and automatic change fiom boost mode to non-boost mode and vice reverse.
The experiment results were recorded every 10 volts difference as the following Figure

6.33.

 

Power Curve

12000
10000
8000
6000
4000
2000

Power

 

150 200 250 300 350 400 450
Voltage

 

 

 

Figure 6.33: The experiment results of power curve changed with voltage.

149

Figure 6.34 shows the results when PV voltage is around 400 V. Figure 6.34 (a)
shows the grid line to line voltage and Z-Source inverter line to line voltage after filter.
These two waveforms should be the same due to. PCC. Figure 6.34 (b) shows the
inverter output current and Z-source inverter PN voltage. Because the PV voltage is
high enough, the shoot through is not needed, Figure 6.35 shows the experiment result
for the same condition of above simulation conditions. From the figure, it can be seen
that the inverter line to line voltage after filter is exact as grid line to line voltage
sinusoidal. Also, the output current is as desired value and waveform, the unity power
factor can be achieved finally. The experiment result consists with the simulation
results well.

Figure 6.36 (a) shows the grid line to line voltage and Z-Source inverter line to line
voltage afier filter. Figure 6.36 (b) shows the inverter output current and Z-source
inverter PN voltage. Figure 6.37 shows the experiment result. At this stage, the
maximum power is around 10 kW; the current injected to the grid is around 40 A peak
value just as shown in Figure 6.36.

When the PV voltage is lower, the shoot through is utilized to ensure the exact output

values. Also, Figure 6.38 and Figure 6.39 both confirm the analysis well.

150

gridVab
gridVbc

Ianab
Ianbc

Ia,Ib,Ic (A)

Vpn (V)

400
200

-200 '

-400
400

200

-200 ‘

-400

 

 

 

 

 

 

 

30

(a). Grid line to line voltage and Z-Source inverter line to line voltage after filter.

A
O

40 50 60
Time (ms)

70 80 90

 

 

 

 

 

 

 

 

 

 

40 50 60
Time (ms)

(b). Z-source inverter output current and PN voltage.

Figure 6.34: Simulation results of

151

Vp

v is around 400 V.

 

‘ Vgrid(250V/div) Vinv(250V/div)

1a (20A/div)

n -

 

:I—e--—i~-i-— ' '

 

 

 

 

 

 

Chl 1:1 Ch2121 Ch31:1 Ch41:1
5 OOmV/div lOmV/div 5 OOmV/div SOOmV/div
DC Full DC50 Full DC Full DC Full

 

 

 

 

 

 

 

 

 

Figure 6.35: Experiment results of va is around 400 V.

152

 

(a). Grid line to line voltage and Z-Source inverter line to line voltage after filter.

Ia,Ib,lc (A)

Vpn (V)

 

 

 

 

 

 

 

 

30 4O 50 60 70 8O 90

Time (ms)

 

 

 

 

 

 

500
400
300
200
l 00

 

 

 

 

30 40 50 60 70 80 90
Time (ms)

(b). Z—source inverter output current and PN voltage.

Figure 6.36: Simulation results of va is around 330 V.

153

 

.....

   

; Vgrid (250V/div) ‘ '

Vinv (250V/div) ' ' I?

 

 

 

 

 

 

 

 

 

 

 

 

Ch3 1:1 Ch4 1:1
5 00mV/div 10mV/div 5 00mV/div 5 00mV/div
DC Full DC50 Full DC Full DC Full

 

 

 

 

Figure 6.37: Experiment results of va is around 330 V.

154

lanab
Ianbc

gridVab
gridVbc

400
200

-200
-400
400

200

-200
-400

 

30 40 50 60 70 80 90

Time (ms)

(a). Grid line to line voltage and Z-Source inverter line to line voltage afier filter.

Ia,Ib,lc (A)

Vpn (V)

 

 

 

Time (ms)

(b). Z-source inverter output current and PN voltage.

Figure 6.38: Simulation results of va is around 230 V.

155

 

  
  

3 Vinv (250V/div) f

   
  
 

 

Vgrid (250V/div)

\l

     
  

f OW? l2 : .-
K. 3 T

 

.1"

 

 

 

Chl 1:1 Ch2 1:1 Ch3 1:1 Ch41:1 1'
00mV/div 10mV/div SOOmV/div SOOmV/div
DC Full DC50 Full DC Full DC Full

 

 

 

 

 

 

 

 

 

 

Figure 6.39: Experiment results of va is around 230 V.

From the above results, it can be concluded that the simulation and experimental
results both show that at different input voltage, the proposed PV system’s output line to
line voltage maintained at 208 V RMS, the current injected to the grid is sinusoidal
without distortion, also unity power factor is realized. The basic principle of the

proposed system was verified very well.

6. 7. PCS Requirements And Standards

Grid-connected PV inverter systems should handle the need to perform output control

and safety disconnecting or stop of the inverter if any failure mode happens. The major

156

failure modes of a grid connected PV system include loss of mains protection, over
voltage protection, galvanic isolation, system grounding problem, and islanding.

Islanding of a grid connected distributed generation system, occurs when the DG
continues to energize a portion of the utility system afier the portion has been
disconnected from the main utility grid [95].

Islanding detection methods can be characterized into two groups: passive and active.
Passive methods include detecting the frequency variation, the voltage phase jump, and
the three-phase drop. The active methods are active frequency drift, impedance
measurement, and reactive power fluctuation. No matter which islanding method is
used, there is a dead range for the islanding operations. It is not difficult to detect
islanding, and it is possible to reduce the likelihood of unintentional islanding almost
completely by monitoring several grid parameters. However, the maximum allowable
time that an inverter may continue to work after the grid has been switched off, together
with suitable test methods; need to be established before common international guidelines
can be reached [96].

There are many standards for PV PCS systems to comply with, such as IEEE Std.
929-2000 (Recommended Practice for Utility Interface of Photovoltaic Systems),
UL1741 (Standard for Static Inverters and Charger Controllers for Use in Photovoltaic
Power Systems), IEEE Std. 519 (Recommended Practices and Requirements for
Harmonic Control in Electrical Power Systems), 1999 National Electric Code (NFP A
70), IEEE Std. 1374-1998 (Guide for Terrestrial Photovoltaic Power System Safety), and

related ANSI and FCC standards.

157

In addition, all PCS topology should include a blocking diode, thus to prevent reverse
currents. The diodes should have a voltage and current ratings at least twice the

open-circuit voltage and short-circuit ratings of the source circuits.

6. 8. Conclusions

Utility connected PV power systems on residential and commercial buildings are
likely to become more important in alternative energy generation for the near future.
This chapter presented a new grid-connected PV power conditioning system based on
Z—Source inverter. The proposed system realizes the boost and inversion with
maximum power tracking in one single power stage, thus minimizing the size and cost.

The main elements in the control structure are the synchronization algorithm based on
PLL, inverter current controller with the power feed-forward and PI controller. A PLL
structure is used to ensure the current to synchronize with the grid. The control
structure has the following advantages:

° High reliability, high immunity
' Lower harmonic distortion
0 High efficiency

With the proposed control method, the system has low output distortion and unity

power factor. The simplicity and good performance of the proposed system make it a

great candidate for grid-connected PV based generation.

158

CHAPTER 7. CONCLUSIONS

7. 1. Summary Of Circuit Topologies And Their Suited
Applications

In PV world market, grid connected PV system has a large portion in the all PV
systems. Grid connected system is less expensive if the grid is used as storage instead
of batteries. In off-grid PV system, batteries cost up to 20% of the whole system. If
backup power is not required, a grid connected system does not need batteries and can get
the excess energy from the PV array. For grid connected PV system, there can be
summarized to three types: central inverter, string inverter, and module integrated
inverter. Table 7.1 shows different specifications of them. Table 7.2 shows summary
of different inverter topology.

From previous works, the single stage large central inverter is not a good choice, the
input voltage must be high to provide to inverter. For ac modules and cells, the dual
stage inverter is much better. It is effective to use HFT in large systems to avoid
resonance. Grounding on input and output terminals also helps on the resonance. LFT
CSIs are applicable for low power ac module applications; HFT VSIs are applicable for

low power and high power systems, the ac module, the string, etc.

159

 

Table 7.1: Summary of central, string, and module integrated inverter.

 

 

 

 

 

 

 

 

 

. Module Integrated
T e entral Inverter Stnn Inverter
yp C g L Inverter
Circuit
Topology
DC bus high voltage, high hlgh voltage, low low voltage, low current
current current
higher power losses, higher installation cost,
advantages mismatch losses, separate MPPT for each no mismatch losses,
(1 nonflexible design, string, higher overall individual MPPT,
, an higher current efficiency than central flexible design for
disadvantages harmonics, lower power inverter extendency, higher
quality efficiency
large scale PV Medium scale PV Small scale PV
Applications system(>5kW), three system(2-3kW), single system(<2kW), single
phase phase phase

 

 

 

 

160

 

Table 7.2: Summary of the inverter topologies.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Inverter Type Topology Characteristic
line fi'equency
__ 1 'l a transformer, low
dc/ac with LFT __ 4 ii‘ switching frequency,
= large transformer
Inverter volume
Wlth dc lac with high frequency
Transform dC/dC(HFr) transformer, small
er transformer volume
no dc/dc converter,
flyback inverter less component, low
cost
no boost stage, need
single stage dc/ac 69 required PV voltage,
T .... _ has a higher efiiciency
first stage is dc/dc
dc/ac with dc/dc i -- II!‘ Q boost converter, the
‘1‘. second stage is fiill
, bridge dc/ac.
dc/ac with i I lower output current
bi-directional :: Q ripple, lower
Transform switches 4‘ L switching frequency
er-less =
Inverter can have shoot
through, higher
Z-Source inverter inverter reliability,
less component, small
size, low cost
single stage half higher devrce rating
b 'd 'th d / voltage, lower
n ge M c ac switching frequency
6 switches for split
- phase, suitable for
Sllilgle Single phase single phase three
P ase three wire dc/ac wires residential
power system
suited for three phase
thr .
Three three phase three W I tee wrreip 0:; ert
_ dc lac d r r I l J I sys em, app ma ‘1: 0
phase wrre T - é } g} fi_-- -‘ large scale grid
connected PV system

 

 

 

 

161

 

' 7. 2. Conclusions

Solar energy is available everywhere. Photovoltaic is a new source of energy
where all of the world can participate. So PV power will become an effective
contributor for distributed resource applications. Photovoltaic has a lot of benefits
for use as a distributed resource including peak demand shaving and improved asset
utilization.

As photovoltaic system is one of the most promising alternative energy sources in
DG. The power electronics converters based PCS in the PV system now becomes

the key point in cost reduction.

This thesis presented a new power conditioning system based on Z-Source
inverter for renewable energy sources. The PV system performance depends not
only on temperatures and irradiations, but also on maximum power tracking function
of PV inverter. So it is important to verify the inverter system also.

This thesis first introduced the basics of PV power system, then summarized all
the existing inverter topologies of PCS mainly for PV systems. Then, the Z-Source
inverter for split single phase with stand alone application was proposed and
implemented. Continually a dc-dc converter based PV simulator was proposed and
implemented for test purpose. Finally, a grid-connected Z-Source inverter system

with PV simulator system was proposed and implemented.

By utilizing the Z-Source inverter, the volume, the cost, and the switching device
count are minimized. Because of the single stage operation, the efficiency of the
system can be greatly improved. The reliability can be enhanced greatly due to the

shoot through states. With all these advanced features, the Z-Source inverter based

162

PCS is very promising for renewable energy applications.

7. 3. Contributions

This thesis has the following contributions which already illustrated in the thesis:

O Summarized previous PCS topologies for PV power system, helpful for other
educations for further research. Also summarized commercial PV inverter

features, helpful for applications of industrial engineers.

O A new split-phase Z-Source inverter based stand-alone residential PV power
system has been proposed and implemented. The proposed control strategy
suits application well. The simulation and experiment results both consistent

with analysis.

9 A new three—phase grid-connected Z-Source inverter system has been
proposed and implemented. The proposed harmonic injected, unity power
factor current control which also employed modified P&O MPPT method is
used for the control strategy. The simulation and experiment results verified
the proposed system. The proposed scheme gives a low cost and high

quality power conversion for PV power system.

Q A dc-dc converter based PV simulator system is proposed and implemented
for test purpose. A new combined control method was applied to realize

functions. The experiment results are shown to verify the circuit.

9 The DSP based control unit for Z—source inverter with PV simulator system
was realized and tested. All the software parts include DSP code, CPLD

code, and lookup tables, are developed and implemented.

163

O The proposed topologies and control strategies can not only contribute to PV
power system, but also to other renewable energy sources, such as fuel cell,

wind power, TEG and so on.

7. 4. Recommended‘Future Works

Renewable energy is very diverse in resources used and conversion. With the
steadily increasing technologies, the PCSs of renewable energy sources are important
for DG applications. People are seeking the new topologies to gain much more
efficiency. This thesis presented a new power conditioning system based on
Z-Source inverter and PV simulator for renewable energy sources. For PV simulator,
it is still have margin to improve performance based on hardware and software
implementation. For grid interconnected PV Z-Source inverter system, battery can
be added as energy storage, also the according control can be studied.

In the future, inverter technology still needs advanced switching components,
improved capacitors, and the fewest interconnections, etc. The inverter lifetime has
better to be longer than 10 years. The design leads to a high integrated and fewer
component trend. Based on advanced technologies, such as DSP, and modular

power electronics, more and more new inverter topologies will arise.

164

APPENDIX 1. HARMONIC DISTORTION
FACTOR OF THE CURRENT RIPPLE

For a single canier switching period with the load between output phase legs a

and b, shown in Fig. A.1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Vdc/2 1 I
\\ I, Phase 3
\- l l l 1 l L
\\ Ar/4 AT/Z A3T/4 I,’ A r t
-Vdc/2 \ 7‘—
\ I
\ l
A \ I,
Vdc/2 x [.1
\ I Phase b
l \\ 1 _l’ l J k
AT/4- A‘T/IZ' A3174 AT t
-Vdc/2 V
Vdc . ------ eab Line-line output
voltage and ripple
\ current
A3T/4 AT '
T T
-Vdc 1 2 T3

 

Figure A. 1: Ripple current for two phase legs of three-phase inverter.

When OStSTI,

-5112,

Aiab = L

When T1 StST1+T2,

165

(Al)

, V — e 6
Arm = (—"C—L—“—2)(t — Ti) “ii—b“ (A2)

When T1+T2_<_tST1+T2+T3,

V —e e e AT—T
(Mm “QQTr - “b( 3) (A3)

A1 =
“b .L L L 2

Thus, one can obtain

<Ai2>= (V—dCJZL
ab 2L AT

2
(20%—8a)] tde
Vdc

[[ZVdc +;(eb - 9a)](t _ T1)
dc

+ [20% _ea)]T1]2d(t _ T)
Vdc

 

r ‘2
2 _

T, (wee—w.)

1 dc

O +

 

T3
+1 (ZVdc+2(:b‘ ea)]T 2 d(t——T1--T2)
+

[Lb- __ea)]T1
Vdc , ]

 

 

 

 

(A4)

 

Substitute “ —M COS 6’0 , ff—e b — —M 008(90 - git], thus obtain

dc dc

166

.2 _
Alab —

Vd

r92L2 7’ —2

2 AT2 1 m3
C

I
7r/3

i

 

t

0032 (6 + g) -— 3J3M 3 cos3 (9 + g)

— .5144 00309 + %)(cos3 ((9 — 3315)

— cos3 (9

167

M6

 

(A.5)

APPENDIX 2. SOME EXPERIMENT
HARDWARES FIGURES

 

Figure A3: Z—Sourcc inverter“.

168

 

Figure A.4: Part of grid-connected Z—Source inverter with PV simulator system.

169

APPENDIX 3. FUNCTIONAL BLOCK
DIAGRAM OF 2407A DSP CONTROLLER

Fig. A.5 shows the block diagram of the basic configuration for the LF2407A
DSP board. The major interfaces of the board include the target RAM, dual SRAM
memory, analog interface, CAN interface, RS 232 interface, 485 interface, SPI data
logging interface, analog signals interface, PWM out signal interfaces and other
functions.

 

V I

Figure A-5: Functional block diagram of 2407A DSP board
The detailed information is listed below:
1. DRAM 544 Words
2. SRAM 2k Words
3. Other function
4. 10 bit ADC converter

5. Can

170

6. SCI

7. SP1

8. External memory

9. DSP core

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Event manager B

Event manager A

PLL/Reset circuit

Other I/O

CAP/QEP units

F lash/ROM 32k Words
Analog signal process units
Can transistors

SPI flash memory

Real time clock

Ram 4k

Pro/data memory space
Extending bus connector
Switching signals (Fiber Led)
Fault signals (Fiber detector)
Frequency signals processing circuit

Out PWM signals processing circuit

171

REFERENCES

[1]. R. C. Dugan, T. E. McDermott; “Distributed generation”, Industry Applications
Magazine, IEEE, Volume: 8 Issue: 2, March-Apri12002, pp. 19 -25.

[2]. Calais, M.; Myrzik, J .; Spooner, T.; Agelidis, V.G. “Inverters for single-phase
grid connected photovoltaic systems-an overview”Power Electronics Specialists
Conference, 2002. pesc 02. 2002 IEEE 33rd Annual, Volume: 4, 23-27 June 2002
Pages: 1995 — 2000.

[3]. Wu, T.-F.; Chang, C.-H.; Chen, Y.-K.; “A multi-function photovoltaic power
supply system with grid-connection and power factor correction features”Power
Electronics Specialists Conference, 2000. PESC 00. 2000 IEEE 3lst Annual, Volume:
3, 18-23 June 2000 Pages: 1185 - 1190 vol.3.

[4]. Jin Wang; Peng, F.Z.; Anderson, J .; Joseph, A.; Buffenbarger, R.; “Low cost fuel
cell converter system for residential power generation”, Power Electronics, IEEE
Transactions on, Volume: 19, Issue: 5, Sept. 2004 Pages: 1315 - 1322.

[5]. F. Z. Peng, “Z—Source Inverter,” IEEE Transactions on Industry Applications,
Vol. 39, No. 2, pp. 504-510, March/April 2003.

[6]. Fang Z. Peng, Miaosen Shen, and Zhaoming Qian, “Maximum Boost Control of
the Z-Source Inverter”, in Conf. Rec. Power Electronics Specialists Conference, 2004.
PESC 04. 2004 IEEE 35rd Annual, Volume: 1 Volume 1, pp: 255-260, Jun. 2004.

[7]. Y. C. Kuo, T. J. Liang, and J. F. Chen, “Novel Maximum-Power-Point-Tracking
Controller for Photovoltaic Energy Conversion System,” IEEE Transactions on
Industrial Electronics, Vol. 48, No. 3, June 2001, pp. 594-601.

[8]. Weidong Xiao, William G. Dunford, “A Modified Adaptive Hill Climbing MPPT
Method for Photovoltaic Power Systems,” 35th Annual IEEE Power Electronics
Specialists Conference, Aachen, Germany, 2004.

[9]. K.H.Hussein, I.Muta, T.Hoshino and M.Osakada., “Maximum Photovoltaic
Power Tracking: an Algorithm for Rapidly Changing Atmospheric Conditions”, IEEE
Proceeding Generation, Transmission and Distribution, vol.142, pp.59-64, Jan. 1995.

172

[10]. E.Koutroulis, K.Kalaitzakis and NC. Voulgaris, “Development of a
microcontroller-based, photovoltaic maximum power point tracking control system”,
IEEE Trans. power Electronics, vol.16, pp. 46-54, Jan.2001.

[11]. Sheri, M.; Joseph, A.; Wang, J.; Peng, F.Z.; Adams, D.J., “Comparison of
traditional inverters and Z-Source inverter”, in Conf. Rec. of IEEE Power Electronics
Specialist Conference, pp. 1692-1698, June, 2005.

[12]. Reference fiom mp://www.bigfrogmountain.com/solarhistoghtm.

[13]. Solar Electric Power, “The US Photovoltaic Industry Roadmap”, Prepared by
Energetics, Incorporated, Columbia, Maryland, under contract to Sandia National
Laboratories.

[14]. Amulf Jager-Waldau, “Status of PV research, solar cell production and
market implementation in Japan, USA and the European Union”.

[15]. Tadao ISHIKAWA, “Grid-connected photovoltaic power systems: Survey of
inverter and related protection equipments”.

[16]. Muhammad H.Rashid, “Power Elctronics Handbook”.

[17]. S.Kumar, S.C. Bhattacharya and M. Augustus Leon, “A survey on PV
systems and accessories in ASIA”.

[18]. “IEEE Guide for Terrestrial Photovoltaic Power System Safety”, the Institute
of Electrical and Electronics Engineers, Inc.

[19]. Yu Chin Qin, Ned Mohan, Russell Bonn, “Status and Needs of Power
Electronics for Photovoltaic Inverters: Summary Document”.

[20]. P.Welter, More, “Better, cheaper-the current market survey: Grid connected
inverters “(Mehr, besser, billiger-Die aktuelle Marktiibersicht, Wechsehichter zur
Netzeinspersung, in German), PHOTON das Solarstrom-Magazin (German solar
electricity magazine), No.3, pp. 60-71, May-June 2000.

[21]. B.Lindgren, Topology for Decentralised Solar Energy Inverters with a Low
Voltage AC-Bus, Proceedings of the 8th European Conference on Power Electronics
and Applications, Lausanne 1999.

173

[22]. R.O.Caceres and I.Barbi, “A boost dc-ac converter: Analysis and
experimentation,” IEEE Trans. Power Electron, vol. 14, pp. 134-141, Jan. 1999.

[23]. N.Vazquez, J. Alvarez, C. Aguilar, and J .Arau, “Analysis and experimental
study of the buck, boost and buck-boost inverters,” in Proc. IEEE PESC’99,
Charleston, SC, June 27-July1 1999, pp. 801-806.

[24]. N.kasa, T.lida, and H.1wamoto, “An inverter using buck-boost type chopper
circuits for popular small-scall photovoltaic power system,” in Proc. IEEE IECON’99,
San Jose, CA, Nov. 1999, pp. 185-190.

[25]. S.B.Kjar and F. Blaabjerg, “A novel single stage inverter for the ac-module
with reduced low-frequency ripple penetration,” in proc. 10th EPE European Conf.
Power Electronics and Applications, Toulouse, France, Scp. 2-4,2003.

[26]. CM. Wang, “A novel single stage full-bridge buck-boost inverter,” in Proc.
IEEE APEC’03, Miami Beach, FL, Feb.9-l3, 2003.

[27]. M. Nagao and K. Harada, “Power flow of photovoltaic system using
buck-boost PWM power inverter,” in Proc. IEEE PEDS’97, Singapore, May 26-29,
1997.

[28]. M.Kusakawa, H. Nagayoshi, K. Kamisako, and K. Kurokawa, “Further
improvement of a transformerless, voltage-boosting inverter for ac modules,” Solar
Energy Mater. Solar Cells, vol.27, pp.379-387, Mar. 2001.

[29]. T. Boutot and L. Chang, “Development of a single-phase inverter for small
wind turbines,” in Proc. IEEE Electrical and Computer Engineering Canadien Conf,
Waterloo, ON, Canada, May24-28, 1998, pp. 305-308.

[30]. S. Saba and V. P. Sundarsingh, “Novel grid-connected photovoltaic
inverter.” Proc. Inst. Elec. Eng, vol.143, pp. 219-224, Mar. 1996.

[31]. S.Funabiki, T. Tanaka, and T. Nishi, “ A new buck-boost-operation-based
sinusoidal inverter circuit,” in Proc. IEEE PESC’OZ, Cairns, Australia, June 23-27,
2002, pp. 1624-1629.

[32]. T. Shimizu, K. Wada, and N. Nakamura, “ A flyback-type single phase utility
interactive inverter with low-frequency ripple current reduction on the dc input for an

174

ac photovoltaic module system,” in Proc. IEEE PESC’02, Cairns, Australia, June
23-27, 2002, pp. 1483-1488.

[33]. B. K. Bose, P. M. Szczesny, and KL. Steigerwald, “Microcomputer control
of a residential photovoltaic power conditioning system,” IEEE Trans. Ind. Applical.,
vol. IA-21, pp.1182-1191, Sep.1985.

[34]. Soeren Baekhoej Kjaer, John K. Pedersen, and Frede Blaabjerg, “A Review
of Single-Phase Grid-Connected Inverters for Photovoltaic Modules”, IEEE trans. On
Industry Applications, vol. 41, No.5, Scp/Oct 2005.

[35]. Hang-seek Choi, Y.J.Cho, J.D.Kim and RH. Cho, Grid-connected
Photovoltaic Inverter with Zero-current-switching.

[36]. M.Meinhardt, G. Cramer, B. Burger, and P.Zacharias, Multi-string-conveiter
with reduced specific cost and enhanced functionality, Conference Proceedings of the
Eurosun 2000, Kopenhagen, Denmark, June 2000.

[37]. Hudson, R.M.; Behnke, M.R.; West, R.; Gonzalez, 8.; Ginn, J.; Design
considerations for three-phase grid connected photovoltaic inverters, Photovoltaic
Specialists Conference, 2002. Conference Record of the Twenty-Ninth IEEE 19-24
May 2002 Page(s): 1396—1401.

[38]. S. Saba and VP. Sundarsingh, “Grid connected photovoltaic inverter as an
industrial product,” Eur. Polymer Fed, pp.46-51, Oct. 1996.

[39]. Beristairr, J. Bordonau, A. Gilabert, and G. Velasco, “Synthesis and
modulation of a single phase dc/ac converter with high frequency isolation in
photovoltaic energy applications”, in Proc. IEEE PESC’03, Aca pulco, Mexico, June
15-19, 2003, pp. 1191-1196.

[40]. Naoto Kikuchi, Souichirou Shigeeda, Hiroshi Watanabe, Tokuo Ohnishi,
Fumio Harashima, Single Phase Amplitude Modulation Inverter for Utility interactive
Photovoltaic System.

[41]. G. R. Walker, and PC. Sernia, “Cascaded DC-DC Converter Connection of
Photovoltaic Modules.

175

[42]. Raymond M. Hudson, Michael R. Behnke, Rick West, Sigifredo Gonzalez
and Jerry Ginn, Design Considerations for Three-Phase Grid Connected Photovoltaic
Inverters.

[43]. Source: National renewable Energy Laboratory.

[44]. Source: WWW-Ballardcom.

[45]. Source: www.bgiconpower.com.

[46]. Source: www.phoenixtec.com.

[47]. Source: www.fronius.com.

[48]. Source: www.studer.com.

[49]. Source: www.sm_a.com.

[50]. Source: www.magnetek.com.

[51]. Source: www.pvpowered.com.

[52]. Source: www.solectrli_a.com.

[53]. Source: www.xantrex.com.

[54]. Kentaro Hayashi, Takae Shimada, Hirotaka Koizumi, Yasuo Ohashi, A novel
cascaded PV inverter by utilizing ready-made Ics for Digital Audio Amplifier.

[55]. Nobuyuki Kasa, Takahiko Iida, and Liang Chen, F lyback Inverter Controlled
by Sensorless Current MPPT for Photovoltaic Power System, IEEE transactions on
Industrial Electronics, vol. 52, No. 4, August 2005.

[56]. Dan Ton, and Ward Bower, Summary Report on the DOE High-tech Inverter
Workshop.

176

[57]. Thomas Surek, Photovoltaics: energy for the new millennium, National
Renewable Energy Laboratory, 1999.

[58]. R. H. Bonn, Developing a next generation PV inverter, 29‘h IEEE
Photovoltaic Specialists Conference, pp. 1352-1355, May 2002.

[59]. M. E. Ropp, Design issues for grid-connected Photovoltaic systems, Georgia
Institute of Technology, 1998.

[60]. R. West, PV Inverter Products Manufacturing and design Improvements for
Cost Reduction and Performance Enhancements, NCPV and Solar program Review
Meeting 2003.

[61]. G. Keller, T.Krieger, M. Viotto, and U. Krengel, Module orientated
photovoltaic inverters-a comparison of different circuits, IEEE First World
Conference on Photovoltaic Energy Conversion, IEEE Photovoltaic Specialists
Conference, pp. 929-932, 1994.

[62]. S H Lloyd, G A Smith, D G Infield, “Design and construction of a modular
electronic photovoltaic simulator”, Power Electronics and Variable Speed Drives, the
Eighth International Conference, 2000.

[63]. Hiroshi Nagayshi, “I-V curve simulation by multi-module simulator using
l-V magnifier circuit”, Solar Energy Materials & Solar Cells 82, pp. 159-167, 2004.

[64]. Olilla, J ., “A medium power PV-array simulator with a robust control
strategy”, Tampere, Finland, Tampere University of Technology, 1995, IEEE, pp.40.

[65]. Nagayoshi. H., Orio. S., Kono. Y., Nakajima. H., “Novel PV array/module
I-V curve simulator circuit”, IEEE 29th Photovoltaic Specialists Conference,
pp.1535-1538, May 2002.

[66]. IEEE P929/Dll, Draft Recommended Practice for Utility Interface of
Photovoltaic (PV) Systems, Nov. 1999.

[67]. T]. Liang, Y. C. Kou, J. R.Chen, “single-stage photovoltaic energy
conversion system,” proc. Inst. Elect. Eng, vol. 4, no. 148, pp.339-344, 2001.

177

[68]. Wu Libo; Zhao Zhengming; Liu Jianzheng; Liu Shu; Yuan Liqiang;
“Modified MPPT strategy applied in single-stage grid-connected photovoltaic
system,” Electrical Machines and Systems, ICEMS 2005 Proc. of the Eighth
International Conference, vol. 2, pp. 1027-1030, Sept. 2005.

[69]. Guo-Kiang Hung; Chih-Chang Chang; Chem-Lin Chen; “Automatic
phase-shift method for islanding detection of grid connected photovoltaic inverters,”
IEEE Transactions on Energy Conversion, vol. 18, pp. 169-173, March 2003.

[70]. J .-M. Kwon, K. H. Nam, B.-H. Kwon, “Photovoltaic Power Conditioning
System with Line Connection,” Trans. on Industrial Electronics, vol. 53, no. 4, pp.
1048- 1054, Aug 2006.

[71]. SJ. Chiang, K.T. Chang, C.Y. Yen, "Residential photovoltaic energy storage
system," Trans. on Industrial Electronics, vol. 45, no. 3, pp. 385-394, Jun.1998.

[72]. Jain, S.; Agarwal, V.; “A Single-stage Grid Connected Inverter Topology for
Solar PV Systems with Maximum Power Point Tracking,” IEEE Transactions on
Power Electronics, vol. 22, pp. 1928-1940, Sep. 2007.

[73]. Liserre, M.; Teodorescu, R.; Blaabjerg, F.; “Stability of photovoltaic and
wind turbine grid-connected inverters for a large set of grid impedance values,”, IEEE
Transactions on Power Electronics, Volune 21, issue 1, Jan. 2006, pages: 263-272.

[74]. H. Shokrollah Timorabadi, C. Li, and F. P. Dawson, “Application of a Fast
Synchronization Systems in Real Time Power System Monitoring and Control”, Proc.
of 22"d Biennial Symp. On Communications, Queen’s University, Kingston, Canada,
pp. 371-376, May, 2004.

[75]. K. E. Martin, “Precise timing in electric power systems”, Frequency Control
Symp., 47‘h proc. of IEEE Int’l, pp. 15-22, June, 1993.

[76]. G. T. Volpe, “A phase-locked loop control system for a synchronous motor”,
IEEE Trans. Automat. Contr., vol. AC-15, pp.88-95, Feb. 1970.

[77]. A. J. Goldstein, “Analysis to the phase controlled loop with a sawtooth
comparator”, Bell Syst. Tech., pp. 603-633, 1963.

[78]. J. L. Brown, “A digital phase and frequency-sensitive detector”, Proc. IEEE,
vol. 59, pp. 717, Apr. 1971.

178

[79]. R. C. E. Thomas, “Frequency comparator performs double duty”, EDN, pp.
29-32, Nov., 1970.

[80]. Chihchiang Hua, Jongrong Lin, and Chihming Sheri, “Implementation of a
DSP-Controlled Photovoltaic System with Peak Power Tracking,” IEEE Transactions
on Industrial Electronics, vol. 45, no.1, pp. 99-107, Feb. 1998.

[81]. D. P. Hohm, M. E. Ropp, “Comparative study of maximum power point
tracking algorithms using an experimental, programmable, maximum power point
tracking test bed,” IEEE 2000, pp.1699-1702.

[82]. E. Koutroulis, KKalaitzakis and NC. Voulgaris, “Development of a
microcontroller based photovoltaic maximum power point tracking control system”,
IEEE Trans. Power Electronics, vol. 16, pp.46-54, Jan. 2001.

[83]. Xuanyuan Wang; Kazerani, M.; “A novel maximum power point tracking
method for photovoltaic grid-connected inverters,” Industrial Electronics Society,
2003, IECON’03, the 29th Annual Conference of the IEEE, Volume 3, 2-6 Nov. 2003.
Pages: 2332-2337.

[84]. Shen,M.; Jin Wang; Joseph, A.; Peng, F. Z.; Tolbert, L. M.; Adams, D. J.;
“Maximum constant boost control of the Z-Source inverter,” Industry Applications
Conference, 2004. 39th IAS Annual Meeting, Conference Record of the 2004 IEEE,
Volume 1, 3-7 Oct. 2004.

[85]. Boost, M. A.; Ziogas, P.D.; “State of the art canier PWM techniques: a
critical evaluation,” IEEE Transactions on Industry Applications, vol. 24, pp. 271-280,
March-April 1988.

[86]. Menniti, D.; Picardi, C.; Pinnarelli, A.; Sgro, D.; “Grid-connected inverters
for alternative energy sources with a combine voltage and current control strategy,”
Clean Electrical Power, 2007, ICCEP’07 International Conference, pp. 223-228,
21-23 May 2007.

[87]. Arruda, L.N., Sliva, S.M., Filho, B.J.C, “PLL structures for utility connected
systems,” Industry Applications conference, 2001. Thirty-sixty IAS Annul Meeting.
Conference Recordof the 2001. Volume. 4, pp. 2655-2660, 2001.

[88]. Da Silva, Sergio A. Oliveira; Tomozaki, Edgar; Novochadlo, Rhodolfo;
Antonia, Emane; Coelho, Alves; “PLL Structures for Utility Connected Systems

179

under Distorted Utility Conditions,” IEEE Industrial Electronics, IECON 2006-32nd
Annual Conference, pp. 2636-2641, Nov. 2006.

[89]. H. Koizunri, T. Mizuno, T. Kaito, Y. Noda, N. Goshima, M. Kawasaki, K.
Nagasaka, K. Kurokawa, “A Novel Microcontroller for Grid-Connected Photovoltaic
Systems,” Trans. on Industrial Electronics, vol. 53, no. 6, pp. 1889-1897, Dec 2006.

[90]. Code Composer Studio Development Tools v3.3 Getting Started Guide (Rev.
H), w. TI. com.

[91]. Code Composer Studio’s Command Window, www.TI.com.

[92]. Programmable power supply ALSlSOVlZA for photovoltaics generator
simulation. Technical note 402545-E1 and appendices. Ainelec.

[93]. Elgar Solar Array Simulators. www.clg_ar.com.

[94]. F. G. Shinskey, Feedback Controllers for the Process Industries. New York,
McGraw-Hill, 1994.

[95]. M.G.Jaboori, M.M.Saied, and AA. Hanafy, “ A contribution to the
simulation and design of photovoltaic systems”, IEEE TransEnergy Conv., vol 6,
pp.401-406, Sep. 1991.

[96]. Chihchiang Hua, Chihming Sheri, “Comparative study of peak power
tracking techniques for solar storage system”, Applied Power Electronics Conference
and Exposition, 1998. APEC ’98. Conference Proceedings 1998., Thirteenth Annual,
pp. 679-685., vol.2.

[97]. International Energy Agency. Trends in Photovoltaic Applications: Survey
Report of Selected [EA Countries.

[98]. P. D. Maycock, “World PV Cell/Module Production”, PV News, Vol. 24, No.
2, Feb. 2005.

[99]. http://en.wikipedia.org/wiki/Solar cell.

[100]. Renewable energy sources, http://iamest.irc.it.

180

[ 10 l ]. http ://www.nextenergynews .com/news 1/next-energv—new312. 19d.html.

181