Research on IGBT Fault Diagnosis Method Based on Characteristic Parameters Identification of Time-Domain Reflection Signal

The malfunction of an insulated gate bipolar transistor (IGBT) device can result in the failure and damage of both itself and the system in which it is installed, and various types of faults in the device typically lead to changes in its impedance characteristics. This paper leverages the advantages of travelling wave reflectometry in the field of identifying characteristic impedance parameters of devices, and conducts device-level research on the transmission and reflection response characteristics of sequence/spread spectrum time domain reflectometry (STDR/SSTDR) signals for IGBT. On the one hand, based on a clear understanding of the aliasing patterns and their impacts on reflected signals, this paper proposes a signal pre-processing algorithm using a time-domain signal transmission analytical model to reduce identification errors introduced by aliasing phenomena. On the other hand, based on establishing the mapping relationship between the characteristic parameters of various IGBTs and their different fault states, this paper proposes an IGBT fault diagnosis method based on characteristic parameters identification of time-domain reflection signals. Compared to traditional methods, the proposed method can achieve non-intrusive fault detection for both IGBT and anti-parallel diode simultaneously, and has the flexibility to adjust the testing conditions. Finally, relevant experiments were conducted to verify the proposed method, and the expected results were achieved.