Thermal Modeling of Tubular Permanent Magnet Linear Synchronous Motor Based on Random Forest

Abstract

This paper proposed a novel thermal modeling analysis method of tubular permanent magnet linear synchronous motor(PMLSM) based on machine-leaning method. Firstly, the structure and main parameters and the finite element (FE) thermal modeling of motor are introduced. A small sample about the average temperature rise of permanent magnet, overall average temperature rise of PMLSM and coil temperature rise are obtained by FE method, corresponding to different heat source inputs. Based on the sample dataset, a powerful machine learning algorithm called Random Forest(RF) is employed to fit the function relationship between output design objectives and input sources parameters. The accuracy of thermal prediction model is verified by the remaining group of sample data. Comprehensive performance comparison shows that the motor thermal prediction model was established by RF is better than other methods such as KNN and SVM.