Research on IGBT Junction Temperature Prediction Method Based on Extended Kalman Filtering

The monitoring of insulated gate bipolar transistor(IGBT) junction temperature is crucial for the health management of power modules. The commonly used temperature sensitive electrical parameters(TSEPs) methods and thermal network modeling methods currently have problems such as insufficient measurement accuracy and the need for regular updates of model parameters. This article proposes a method for predicting the junction temperature of IGBT based on extended Kalman filtering(EKF). The on-state voltage drop is used as TSEPs to obtain the junction temperature, a thermal network model is established, and EKF is used to adaptively modify the junction temperature. At the same time, model parameters such as thermal resistance and heat capacity can be identified and updated to achieve optimal estimation of the junction temperature throughout the entire life cycle. The simulation results show that this method exhibits high prediction accuracy for both junction temperature and thermal network parameters during the aging process of IGBT, and has important value for practical engineering applications.