On-line Estimation Method of Junction Temperature Based on SiC Module Aging Parameter Identification

Abstract

Accurate evaluation of junction temperature is extremely important to improve the safety and reliability of SiC modules. However, it is difficult to obtain the internal junction temperature of SiC modules in real time, limited by sensors and testing methods. The junction temperature estimation based on the thermal network model combined with an integrated negative thermal coefficient (NTC) thermistor is an effective evaluation method. The existing thermal network models omit effects of the solder layer fatigue aging due to thermal stress, resulting in junction temperature estimation error. In this paper, a junction temperature estimation method is proposed, which establish the relationship between NTC sensor temperature and the thermal resistances of chip solder layer and bottom layer. The junction temperature and NTC sensor temperature of the SiC module is expressed using the equation of state analysis to obtain a discrete-time system of temperature. The aging thermal resistance parameters of the thermal network model can be identified online, and the accurate online estimation of different aging conditions inside the module can be realized.