Research on Online Reliability Assessment Technology for IGBT Based on Driver-Side Measurement Fusion

Insulated Gate Bipolar Transistor (IGBT) have been widely adopted in energy conversion systems, and their reliability has become a critical focus in industrial applications. Temperature is the primary cause of failure in power electronic converters, and accurate junction temperature estimation of IGBT modules serves as a fundamental basis for reliability assessment. This paper proposes an online junction temperature evaluation method for IGBT based on multi-drive monitoring quantity fusion. By integrating gate current (I<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">g</inf>), Miller voltage (V<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GP</inf>), and plateau duration during turn-off (t<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MP,off</inf>), a three-domain Kalman filter spanning voltage, time, and current is constructed. Furthermore, adaptive Kalman filtering is employed to achieve dynamic noise suppression. Experimental results demonstrate that this method achieves a fusion weight adaptation accuracy of ±1.5%, enabling precise measurement of junction temperature.