Multiobjective Optimization With Hybrid 2-D Equivalent Model on Axial Flux Permanent Magnet Machine for Special Electric Vehicles

This article proposes a new high-efficiency multiobjective optimization method for an axial flux permanent magnet (AFPM) machine. First, based on the torque production theory, a high-precision 2-D equivalent model for the AFPM machine is proposed, which can improve the accuracy of the traditional 2-D equivalent model and reduce the computation cost of the optimization design stage. Then, the hierarchical optimization strategy is proposed based on the single-parameter scanning (SPS) method and surrogate model, which improves the optimization efficiency. In addition, an improved gray wolf optimization (IGWO) algorithm is employed by integrating Tent chaotic mapping, nonlinear control parameters, and Levy flight, which enhances the global search capabilities and convergence rate. Afterward, the electromagnetic performances of the initial and optimal models are compared and analyzed based on the 3-D model and the proposed 2-D equivalent model. Finally, a 25-kW prototype is manufactured and tested to verify the effectiveness of the hybrid 2-D equivalent model and the proposed multiobjective optimization method.