Multi-Objective Optimization Design of Dual-Stator Low-Speed High-Torque Motor based on BP Neural Network-Improved Genetic Algorithm

This paper explores the multi-objective optimization design issue of the dual-stator hybrid rotor low-speed high-torque motor (DS-LSHTM).The torque, torque ripple and motor efficiency are selected as the optimization targets. Firstly, by simplifying the motor structure, the representative parameters are selected for sensitivity analysis, and then the back propagation neural network improved genetic algorithm (BP-GA) and single factor scanning method are used for multi-objective optimization of parameters with different sensitivities to determine the overall design of the motor. Finally, the target performance changes of the motor before and after optimization are compared by finite element simulation.