Predicting the service life of IGBT modules is an effective way to assess their health and reliability. Based on the IGBT aging experiment data, This paper constructs an IGBT state detection index including a two-dimensional parameter of the saturation voltage VCE (ON) and the junction temperature Tj. For the normalized data, this paper introduces a way of segmented processing to remove the large index fluctuation caused by the break of the IGBT bond wire. Then, it performs the regression analysis on the change considering only the aging of the solder layer and find the fitted aging data curve as well as approximate function expression. To find the relationship between the parameters of IGBT module, this solution builds a prediction model of IGBT aging based on BP neural network. The accuracy of the neural network model in the prediction of remaining life is analyzed for the same sample with different channels and different experimental conditions.
