Junction temperature estimation in IGBT power modules based on Kalman filter

Insulated gate bipolar transistor (IGBT) were widely used in energy field, its reliability has become the focus of industry. Temperature is the main reason for the failure of power electronic converter, and the accurate estimation temperature of IGBT module is an important basis for its reliability evaluation. Currently the main methods of extracting junction temperature are thermo sensitive electrical parameter method (TSEP) and the thermal impedance model method, due to the TSEP method requires additional auxiliary circuit and it is susceptible to electromagnetic interference, does not necessarily apply to the applications. In this paper, a junction temperature estimation method depends on thermal impedance model is proposed, it fuses Kalman filter and IGBT modules' NTC thermistor reflects indirectly junction temperature. Proposed method estimation procedure does not need to store large amounts of data, does not take up too much memory and accurate estimation of junction temperature, it is also easier to realize the IGBT reliability evaluation.