Temperature Prediction for PMSM Motors Using Deep Learning Algorithms

This study examines the impact of heat dissipation and cooling techniques on the temperature increase in permanent magnet synchronous motors (PMSM). Accurate assessment and prediction of motor temperature, torque, stator yoke, stator tooth, and stator winding temperatures play a crucial role in ensuring the safety and reliability of permanent magnet synchronous motors. In this work, the impact of these three temperature factors is investigated to prevent motor insulation aging, insulation burnout, demagnetization of the permanent magnets, and additional issues resulting from elevated temperatures in the stator windings. A computational model used for predicting the temperature of permanent magnet synchronous motors is presented. The model uses ambient temperature, coolant temperature, torque, motor speed, longitudinal voltage, transverse voltage, longitudinal current, cross current, and surface temperature of the permanent magnet as inputs, and stator winding temperature, stator yoke temperature, and temperature of the stator teeth as outputs. A convolutional neural network (CNN) model was constructed for PMSM temperature prediction.