Multiphysics Dimension-Reduced Optimization of a PM Hub Motor Based on Electromagnetic-Thermal Mesh-Shared Network Model

This article proposes a multiphysics dimension-reduced optimization design approach to comprehensively improve the electromagnetic performances and temperature characteristics of the hub motors. First, an electromagnetic-thermal mesh-shared network (ET-MSN) model is proposed, which achieves a high integration between the electromagnetic analysis model (E-AM) and the thermal analysis model (T-AM). The significant advantage of the ET-MSN model is its ability to establish the mapping between the electromagnetic performances and temperature characteristics. Based on the ET-MSN model, two electromagnetic-thermal dimension-reduced objectives that simultaneously represent electromagnetic performances and temperature characteristics are defined, thereby facilitating efficient multiphysics dimension-reduced optimization. Finally, a prototype is fabricated. The proposed optimization approach is thoroughly validated by simulation and experimental results, which effectively address challenges presented in traditional multiphysics optimization methods.