[P2-8] Motor Drives, Condition Monitoring and Diagnostics

[P2-8-01] Uncertainty-Aware IGBT Aging State Classification for Traction Inverters Using Bayesian Residual Networks

*Jiahong Liu², Subham Sahoo¹, Xing Wei¹, Bo Yao¹, Yichi Zhang¹, Yongjie Liu¹, Frede Blaabjerg¹, Huai Wang¹ (1. Aalborg University (Denmark), 2. State Grid Jiangsu Electric Power Company Ltd. (China))

Keywords:

Bayesian neural network,Condition monitoring,IGBT,Traction inverter,Uncertainty quantification,ResNet,Variational inference