A lasso regression with embedded analytical model (EAM), called EAM-LR, is proposed to quickly and accurately calculate the thrust performance of the permanent magnet synchronous linear motor (PMSLM) in this paper, and combined with the EAM-LR, the chaotic golden section search algorithm (CGA) was introduced to optimize the PMSLM structure to achieve high thrust density and low thrust ripple. First, the PMSLM thrust performance was analyzed by analytical model (AM) to determine the variation range of structural design parameters. Based on the variation range, a finite-element sample database was established. Then, combined with the finite-element sample database, the analytical mapping functions derived from AM, were integrated into Lasso regression to establish EAM-LR. Finally, CGA was introduced to optimize the performance of PMSLM, and simulation experiment comparison proves the effectiveness of the proposed method.

machine learning model， embedded analytic model， lasso regression， chaotic golden section search algorithm， permanent magnet synchronous linear motor (PMSLM).