Multi-objective Optimization of Interior Permanent Magnet Synchronous Motor Based on Layered Response Surface Method and Improved NSGA-II

In order to simultaneously achieve high output torque, low torque ripple, and stable induced voltage of interior permanent magnet synchronous motor (IPMSM), this paper proposes a multi-objective layered optimization method based on a layered response surface method (LRSM) and an improved non-dominated sorting genetic algorithm-II(NSGA-II-M). Through comprehensive sensitivity index analysis, the optimization parameters are divided into top and bottom layers, and the response surface models are obtained using the central composite surface face (CCF) and the optimal Latin hypercube design (OPTLHD), respectively. By organically combining LRSM with NSGA-II-M, the final optimal parameters are obtained. A finite element analysis is conducted to compare the torque performance of the initial and optimized designs, verifying the feasibility and effectiveness of the proposed IPMSM multi-objective layered optimization method.