To improve the performance (low torque ripple and high average torque) of the permanent magnet assisted switched reluctance motor (PMa-SRM)

a hybrid multi-objective optimization method is proposed

which combines the genetic algorithm optimizes support vector machine (GASVM) with non-dominated sorting genetic algorithm Ⅱ(NSGA-Ⅱ). Firstly

the PMa-SRM topology is introduced. Secondly

the experimental sample data is established by the Taguchi method

and the effects of the turn on angle

the turn off angle and the pole shoe angle on the PMa-SRM torque ripple and the average torque are analyzed by variance analysis. Thirdly

the torque ripple and the average torque predictive models between the turn on angle

the turn off angle and the pole shoe angle are established by the GASVM. Next

the NSGA-Ⅱ is used for global optimization of the predictive models

and the optimal design is selected from the Pareto front which is generated by NSGA-Ⅱ. The output performance of the optimized front and rear motors

and the prototype torque and vibration experiment confirm the effectiveness of the optimization design method with GASVM-NSGA-Ⅱ.