ABSTRACTSAMPLE AND HOLD FUNCTIONS AND NON-MINIMUM-PHASE SYSTEMSByYingxu WangWith the same initial condition, a continuous-time system and the corresponding DE system will have the same system responses at sampling times due to switched input and zero-order hold input, respectively. A previous work had been done to show that using the Square Pulse Sample and Hold Function (SPSHF) as the switched input, the corresponding Discrete Equivalent (DE) system of a continuous-time non-minimum phase (NMP) system could have minimum phase (MP) behavior. In this thesis, the switched input in the definition of DE system was expended with another two sample and hold functions, Forward Triangle Sample and Hold Function (FTSHF) and Backward Triangle Sample and Hold Function (BTSHF). The DE systems and its Discrete Time (DT) systems equations were developed according to different sample and hold functions. A simulation case study was proceeded to indicate the feasible regions for selecting the sample and hold parameters for each switched input, which would lead the resulting DE systems MP. The results of simulation case study implied that it was possible to have an MPDT system with a smaller sampling period than discretizing the system directly using zero-order hold (ZOH) method. In order to study the robustness of each sample and hold function, the q-Markov cover system identification with Pseudo Random Binary Signal (PRBS) was then studied for the purpose of doing controller-in-the-loop (CIL) simulation. The CIL simulation used the dSPACE auto-box to simulate the system dynamics. A resistant-capacitor (RC) filter was used to simulate actuator dynamics using different capacitors. At last the characteristics of the three sample and hold functions were compared.
Read
