Discrete Optimization of Weighting Factor in Model Predictive Control of Induction Motor

Abstract

Tuning the weighting factor is crucial to model predictive torque and flux control. A finite set of discrete weighting factors is utilized in this research to determine the optimum solution. The Pareto line optimization technique is implemented to prevent the occurrence of local optimum solutions. By conducting an accuracy analysis, the number of discrete weighting factors is optimized, and the number of iterations is reduced. The stator current distortion minimization criterion is used to obtain the ultimate global optimal solution from the Pareto line. This study compares the results of the proposed optimization method and the particle swarm optimization method based on experimental data from a 4 kW induction motor drive test bench. The proposed technique can achieve the global optimum weighting factor in a shorter computational duration while maintaining a slightly lower total harmonics distortion and torque ripple.