Rotor temperature estimation for permanent magnet synchronous motors in electric vehicles

Aiming at solving the problems of the demagnetization for the permanent magnet due to the rotor high temperature, on-line temperature estimation of the permanent magnet synchronous motor is crucial. Considering the real heat flow in the motor and on-line temperature estimation requirements, a five-node simplified equivalent thermal network is proposed based on the lumped parameter thermal network, and the state equations of the motor temperature estimation are derived. Based on the linear functions of the node temperatures derivative, node temperature, and heat loss in the discretized state equations, the multiple linear regression method is proposed to identify the parameters in the motor state equation. Then, the rotor temperature can be predicted in real-time based on the state equation. The bench experiment was conducted under the locked-rotor, low-speed and high speed conditions, and the temperature predicted by the proposed model was compared with the temperature from the sensor. The experimental results show that the proposed thermal model can accurately predict the rotor temperature with the temperature error of 6℃. The performance of experimental results verifies the accuracy of the proposed thermal model and the temperature estimation algorithm, and the lower-order model is easy to be implemented in the embedded system of automobile for the estimation of the rotor temperature in real time. © 2021, Harbin University of Science and Technology Publication. All right reserved.