Large-Scale Design Optimization of an Axial-Flux Vernier Machine With Dual Stator and Spoke PM Rotor for EV In-Wheel Traction

This article presents the optimization study targeting a specific drive cycle for a MAGNUS-type axial-flux permanent magnet vernier machine (AFPMVM). The proposed MAGNUS machine has a novel design with a dual-stator configuration, where only one stator is wound, with a high-polarity spoke permanent magnet (PM) rotor. The machine topology has a 3-D flux path, which necessitates the analysis of a large finite-element (FE) model. However, due to the computational complexity and time required for such a large FE model, a new approach was developed. This approach involves a computationally efficient FE analysis (CE-FEA) model combined with a single-point drive-cycle analysis and the differential evolution (DE) optimization algorithm. The targets of the optimization algorithm are derived by modeling the load operating cycle through a systematic k-means clustering method, identifying specific operating points representing high-energy zones within the drive cycle. The optimized design achieves a wide range of constant power operation, which is desirable for electric vehicle (EV) in-wheel traction. Experimental and numerical results demonstrate a higher torque density in the MAGNUS machine compared with commercially available EV traction motors. In addition, this article explores various flux-weakening methods for the MAGNUS machine, highlighting their respective benefits.