RSM Assisted MOGA for SRM EV Drive Control Factors Optimization

This work deals with the multiobjective optimization of control factors of a switched reluctance motor (SRM) drive for an electric vehicle (EV) application using an optimization algorithm with reduced computation effort. In motor operation using conventional torque sharing function (TSF), input control parameters must be optimized to achieve desired drive response. These control factors are sensitive to motor operating speed and load condition. The sensitivity analysis of control factors, followed by the effect of operating conditions, is utilized to compose the objective function using response surface method. Response surface model (RSM) combined with a multiobjective genetic algorithm (MOGA) is implemented to obtain the optimum value of control variables with faster computation to meet the minimum torque ripple and maximum torque per ampere criterion. The low computation method computes the optimum result for varying operating applications without continuously calculating the desired objective and evaluates the result based on the fitted model. The input reference torque to TSF also plays a vital role in reducing the torque ripple. Therefore, torque loop is established to track TSF reference torque accurately. Experimental and simulation results validate performance of presented combined technique to optimize TSF input parameters.