Optimization of Permanent Magnet Surface Shapes of Electric Motors for Minimization of Cogging Torque Using FEM

A novel algorithm which minimizes the cogging torque of surface-mounted permanent magnet (SPM) brushless motors by optimizing the surface profile of PM is presented. An efficient strategy for using finite element method (FEM) is proposed to calculate the cogging torque of PM motors with different PM shape designs. Compared with the use of general FEM, the computing time with this proposed method is significantly reduced, which is only 0.017% of that required by the former. Because of the considerable reduction of computing time in the magnetic field analysis of PM motors, it becomes feasible to use genetic algorithm (GA) to optimize the PM shape in order to realize cogging torque minimization. A numerical experiment is used to validate the proposed method.