Multiobjective Optimization of Yokeless Axial-Field Flux-Switching Permanent Magnet Motor Using the Hybrid Taguchi Genetic Algorithm for Expanded Speed Range

This paper proposes a multiobjective hybrid Taguchi genetic algorithm (HTGA) to optimize the speed range of a yokeless axial-field flux-switching permanent magnet (YASA-AFFSPM) motor. HTGA combines Taguchi’s local optimization with the global optimization of traditional genetic algorithms (GAs), facilitating faster and more accurate solutions. The Taguchi method is employed to generate offspring individuals within GA; it inherits parameter characteristics from stronger offspring, saving considerable computation time. The objective is to achieve a motor with low cogging torque, high average torque, and an expanded speed range in the field weakening area. Various parameters of the motor, such as the split ratio, stator axial length, pole angles, PM arc, and number of conductors per slot, are selected as optimization variables. The optimization constraints include the field-weakening rate, saliency rate, cogging torque, and average torque. The optimized motor parameters are determined, and the speed range before and after optimization is evaluated. Cosimulation analysis using a 3-D finite element method (FEM) is performed under no-load and full-load conditions to compare the motor’s speed regulation range. The optimized motor exhibits a maximum speed that is almost 1.5 times higher than the initial design, with improvements of 11.3% in average torque and 9% in cogging torque. Experimental results compared to 3-D FEM simulations demonstrate the superior performance of the optimized motor in terms of speed, torque, power, and efficiency.