Research on Surrogate Model-Based Optimization Method for Dual-Phase Magnetic Material Permanent Magnet Synchronous Motors

Improving power density to reduce the volume and weight of electric motors has become a prominent research focus for high-speed motors. However, in-creasing rotational speed results in excessive centrifugal forces, which may lead to rotor cracking. To address this challenge and enhance rotor structural integrity, this paper proposes an optimized design method for an interior permanent magnet synchronous motor (IPMSM) employing dual-phase magnetic materials. Considering the structural complexity and manufacturing challenges associated with rotors composed of dual-phase magnetic materials, rotor geo-metric parameters are selected as optimization variables. A surrogate model of the motor is established by training an artificial neural network (ANN), with electromagnetic performance and manufacturability defined as optimization objectives. The NSGA-II algorithm is subsequently employed to derive Pareto-optimal solutions, enabling efficient optimization of the dual-phase magnetic rotor structure. Finally, the validity and effectiveness of the proposed optimization method are verified through finite element analysis.