Cogging Torque Prediction Model and Analysis of Permanent Magnet Motor Based on GA-PSO-BP

Due to the complex nonlinear relationship of permanent magnet motors, it is difficult to effectively construct the optimal design model of permanent magnet motors by analytical method and finite element analysis (FEA). in order to solve this problem, this paper uses a hybrid algorithm combining genetic algorithm (GA) and particle swarm optimization algorithm (PSO) to optimize the BP neural network, so as to construct a cogging torque prediction model for permanent magnet motor. Firstly, the 8-poles and48-slots surface-mount PM magnet brushless DC motor (SPMBLDC) model is established by the FEM, and four design parameters and cogging torque of the motor are selected as research objects. Secondly, 380 sets of data were extracted by Latin Hypercube (LHS), 380 sets of cogging torque values were obtained by FEM, and the data are randomly divided into training samples and test samples in a ratio of 8:2, and the GA-PSO-BP algorithm was repeatedly trained and tested to construct a cogging torque prediction model. The final results verify the feasibility and effectiveness of the cogging torque prediction model constructed by GA-PSO-BP.