Parameters Optimization of the Permanent Magnet Linear Synchronous Machine Using Kriging-based Genetic Algorithm

Abstract

This paper presents an efficient optimization design method for a permanent magnet linear synchronous motor (PMLSM) to suppress thrust ripple and maintain high average thrust. The effects of twelve design parameters on the performance of PMLSM are studied, and five key design parameters are selected through Taguchi method. To obtain the optimal configuration of five design parameters, an improved Latin hypercube sampling is adopted to generate samples for constructing the Kriging model, and the Kriging-based genetic algorithm is applied to increase the optimization efficiency. The mean square error function as a criterion of response accuracy is to make the optimization result more robust. A two-dimensional finite element analysis is used to evaluate performances of the original and optimized PMLSM. The results show that the optimized PMLSM has a 68% reduction in thrust ripple and a slight increase in average thrust compared with the original PMLSM. Furthermore, the optimized PMLSM with skewing pole will achieve better thrust characteristics.