Optimal Design of PMa-SynRM for Electric Vehicles Using Grain-Oriented Electrical Steel and Surrogate Model Based on Stacking Ensemble

Abstract

In this paper, grain-oriented electrical steel sheet (GOES) and the surrogate model based on the stacking ensemble method of machine learning are proposed to improve the performance of the permanent magnet-assisted synchronous reluctance motor (PMa-SynRM). The application of the GOES in the stator teeth increased the average torque by 7.94% with −0.32% less core loss. Furthermore, the six-dimensional optimal design of the PMa-SynRM with was conducted with the stacking ensemble method and derived 16.99% improved average torque, 1.28% decreased core loss, and 0.7% increased driving efciency.