Optimization of Secondary Iron of Homopolar Linear Synchronous Motor for Traction Application Based on Finite Element Method and Regression Model

Due to the transverse flux, homopolar linear synchronous motor (H-LSM) is generally analyzed by three-dimensional finite element method (FEM). The computational effort is large when using multi-objective optimization algorithm to optimize the structural parameters based on FEM. In this paper, a new method is presented. Structural parameters of the secondary iron are sampled, and calculated the corresponding performance of H-LSM using FEM. The regression model between structural parameters and performance of H-LSM was trained via machine learning algorithms. Based on this regression model, the Pareto solutions are calculated by using the multi-objective optimization algorithm, taking thrust and normal force as the optimization targets. Finally, the Pareto solutions are verified by FEM. In one case, after optimization, the thrust increases by more than 30%. It is proved that this method is accurate and effective, and can greatly reduce computational effort.