Enhanced Particle Filtering-Based Lifetime Prediction for IGBT in High-speed Trains

For the high-speed, high-power and long-attened-time applications, traction converters with IGBT in high speed train have become the most fragile component during the conventional operations. In consideration of the difficulty of monitoring junction temperature of IGBT, the fatigue failure analysis is used as the basis to fit the junction temperature mathematical model, and then the enhanced particle filtering algorithm is used to predict the lifetime of IGBT module. Case studies performed on a simulation example indicate that the presented approach can accurately predict thr lifetime of IGBT, which is helpful for improving the reliability of the device.