Performance Investigation of Quasi Z Source Inverter (QZSI) Fed Induction Motor (IM) Drive System with bGWO-RDF Technique

Abstract

In this dissertation, a hybrid strategy is intended for performance investigation of Quasi Z Source Inverter (QZSI) fed Induction motor (IM) drive system. The hybrid strategy is the fusion of both the binary version of grey wolf optimization (bGWO) and random decision forest (RDF) and commonly called as bGWO-RDF technique. The main contribution of the dissertation is to control the QZSI for enhancing the stability and the performance of IM drive system using the proposed hybrid strategy. The input of the proposed strategy is the speed and three phase current of IM. With the help of input parameters like actual speed and reference speed of IM, the PI controller gain parameters are optimized using bGWO technique. In order to train the RDF, the PI controller output like reference quadrature axis current and actual quadrature axis current is chosen. By utilizing the RDF, the reference three phase current is generated. By utilizing the actual and reference three phase current, the control pulses are generated at the final stage. Finally, the proposed technique is executed in the MATLAB/Simulink working platform. The effectiveness of the proposed is compared with various existing techniques such as ANN, PSO, PSO-SVM and GSA-SVM. The effectiveness of the proposed and other existing techniques is analyzed in terms of ripples and harmonics reduction, rise time, settling time and steady state error. Furthermore, the statistical analysis is also performed and compared with other existing techniques. The comparison result reveals that the total harmonic distortion (THD) is decreased and the oscillation period of the stator current, torque and speed are eliminated.