Multi‐Objective Optimization of <scp>FT</scp> ‐ <scp>PMLSM</scp> Based on <scp>GPR</scp> ‐ <scp>SVR</scp> and <scp>NSGA</scp> ‐ <scp>II</scp>

To suppress the harmonic components of detent force in the flat‐type permanent magnet linear synchronous motor (FT‐PMLSM), this paper utilizes Gaussian process regression (GPR) and support vector regression (SVR) to evaluate the comprehensive performance of the 12‐slot/14‐pole FT‐PMLSM. Firstly, the topology structure of the 12‐slot/14‐pole FT‐PMLSM is introduced. The Fourier decomposition theory and the equivalent air gap length theory are employed to analyze the model, and a finite element model is established with several structural parameters. Secondly, the Taguchi method experimental design is carried out, and the Pearson correlation coefficient (PCC) is introduced to perform global sensitivity analysis of structural parameters to screen out the key structural parameters and reduce the dimensionality of the design variables. Thirdly, a high‐precision GPR‐SVR combined surrogate model is established to map structural parameters and performance, and training is carried out for the sample space. Based on this, NSGA‐II is introduced to conduct an iterative solving process to obtain the optimal structural parameter combination. Finally, based on the optimization results, the thrust performance of the motor is predicted using the finite element method (FEM), which verifies the effectiveness of the proposed optimization approach. © 2026 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.