Remaining useful lifetime prediction of thermally aged power insulated gate bipolar transistor based on Gaussian process regression

Abstract

This paper proposes a new approach for power electronic converters faults prognosis under-insulated gate bipolar transistor (IGBT) failures. Power electronic converters such as inverters and rectifiers are crucial parts of most renewable energy systems. Usually, the power converters are subjected to a high failure frequency rate and lead to a power system shut down. These faults, generally, start with thermal overstress, leading to IGBT wear-out over time. In the considered application, the power electronic converters are supplied by a generator. The unbalance IGBT degradation has been artificially created through high amplitude thermal overstress utilizing a DC at the gate. The proposed prognosis approach is based on the computation of the time-domain features to extract the degradation behavior of the IGBT device. Then, the Gaussian process regression technique is applied to the remaining useful life estimation. Prognostic results in three IGBT cases illustrate the effectiveness of the proposed prognostic approach, leading to an effective prognostic procedure for IGBT faults in power electronic converters.