Torque prediction for condition monitoring of permanent magnet synchronous motor using machine learning algorithms

Abstract

Permanent Magnet Synchronous Motor (PMSM) is the widely used motor for electric vehicle applications. It is also used in industrial drives, robotics, elevators, and escalators. The wide range of application of this motor is due to its higher efficiency and torque performance in both high and low speed of operation. Torque is considered as one of the important features in PMSM as it is indicative of the motor performance. Fault diagnosis and identification of type of faults in PMSM is possible by predicting the torque. In an electric vehicle, it is necessary to monitor the motor performance for safety reasons. In this paper, torque is predicted for different motor parameters using Machine learning (ML) algorithms. Various ML algorithms such as Linear Regression, Support Vector Regression, Elastic Net Regression, Ridge Regression and Lasso Regression are implemented and a comparative study is done to find the best ML model for the torque prediction of a PMSM.