Feature extraction based on EMD for Insulated Gate Bipolar Transistor

In this paper, empirical mode decomposition (EMD) is used in feature extraction on Insulated Gate Bipolar Transistor (IGBT). Dynamic changes of the degradation parameters such as collector-emitter voltage (V_CE) and collector-emitter current (I_CE) were got by temperature cycling test. Having conducted data compression and preliminary processing, we proceed the decomposition by using the EMD method. Then IGBT degradation trends can be found through the curve after treatment and the degenerate point can be found through energy analysis. The results show that EMD method can be well suited for feature extraction of IGBT data, and lay a foundation for further study of the data.