Cogging Torque Minimization of a Dual-Type Axial-Flux Permanent Magnet Motor Using a Novel Optimization Algorithm

The design of an electric machine is a multivariable and multimodal problem that requires much time to find the optimal design result. To address this problem, we propose a novel optimization algorithm using a kriging meta-model, a contour face climb method, a reinterpolation method in an expected region, an advanced pattern search method, and a uniform search method in the entire problem space. A rapid, exact, and reliable optimization process for a multivariable and multimodal design problem is possible using the proposed novel algorithm. Its validity was confirmed by comparing the optimization results of a test function with evolutionary optimization methods. For verification of the application to an electric machine, a dual-rotor axial flux permanent-magnet synchronous motor was designed using the proposed algorithm in this paper.