Optimization of a Consequent-Pole PMSM with a Sobol Sequence Based Multi-Objective Hybrid Optimization Algorithm

Yanzhe Shang,Deliang Liang,Shaofeng Jia,Yongtao Liang
DOI: https://doi.org/10.23919/ICEMS52562.2021.9634233
2021-01-01
Abstract:Consequent pole permanent magnet synchronous machine (CPPMSM) can greatly improve the permanent magnet utilization rate, save materials and decrease costs. In this paper, the parametric modeling of CPPMSM is carried out by using the finite element software. Aiming at core loss and torque ripple, the minimum loss and torque ripple are taken as optimization objectives, and relevant structural parameters of CPPMSM are selected as optimization variables. The sobol sequence method is used to sample combinations of optimization variables, and the sensitivity analysis and stratification are carried out. Through stratification, the algorithm combined with kriging surrogate model algorithm (Kriging-SMA) and particle swarm optimization algorithm (PSOA) is used for multi-objective optimization. Thus, a multi-objective hybrid optimization algorithm based on the sobol sequence method is formed, and optimal design of CPPMSM can be realized.
What problem does this paper attempt to address?