Multi-objective optimization of a permanent magnet synchronous motor based on an automated design and analysis procedure

Abstract

Since traction motors for electric vehicles require a variety of features to be used simultaneously, research on the design optimization of permanent magnet synchronous motors (PMSMs) has been increasing. To apply design optimization in the industrial field, it must be accurate, simple, and the design time should be minimized. This study proposes an automated design and analysis procedure to optimize the shape of a PMSM. ANSYS Maxwell and PIAnO were combined to implement automation, which are finite element analysis and design optimization software programs, respectively. The average torque and total harmonic distortion of the back electromotive force were set as a multi-objective function. In addition, the efficiency and torque ripple were set as constraints with five design variables to satisfy them. To verify the superiority of the proposed design optimization procedure, three optimization algorithms were reviewed and compared with metamodel-based design optimization. These algorithms include parallel algorithm discovery and orchestration, a progressive quadratic response surface method, and sequential two-point diagonal quadratic approximate optimization Moreover, because it is a design optimization problem with a multi-objective function, the changes in the design optimization results according to the changes in the weighting ratio were examined. Finally, the validity of the design results was verified through experiments. The research results demonstrate that the proposed design optimization procedure is more accurate, faster, and better than the conventional method.