Fast Optimization of a Switched Reluctance Motor for 2W-EV using Combined Surrogate Model and Multi-Objective Optimization

This paper deals with the design and optimization of switched reluctance motor (SRM) for a two-wheeler electric vehicle (2W-EV) through the process of screening, surrogate modelling combined with a multi-objective genetic algorithm (MO-GA) approach. The design variables are selected based on the objective of high average torque and reduced torque ripple. Sensitivity analysis of design variables is done with definitive screening design(DSD) that considers interaction of design variables and screened for significant variables. Next, surrogate models are constructed based on selected significant variables. Three different surrogate models quality and accuracy are compared and a combination of constructed models is selected for multi-objective optimization using genetic algorithm (MO-GA). The MO-GA based optimization process is combined with the combined surrogate model to generate the pareto front to assess the best designs within design space.