Fault diagnosis of SiC MOSFETs based on time-frequency analysis of acoustic emission signals

In the high-frequency switching applications of SiC MOSFETs, traditional techniques are faced with challenges in achieving non-invasiveness measurement, real-time evaluation, and online monitoring. An effective condition monitoring approach for SiC MOSFETs is crucial to ensure the safe operation of power electronic systems. A solution based on the acoustic emission (AE) detection technique is proposed in this paper. Through the utilization of time-frequency analysis, the time-frequency characteristics of the AE signals generated by SiC MOSFETs under double-pulse test conditions have been explored. Three multi-dimensional characteristic parameters, namely time-frequency variance, energy entropy, and time-frequency center of gravity, have been used, which can achieve quantitative characterization of the dynamic behavior from SiC MOSFETs. By systematically comparing the time-frequency characteristics of AE signals from SiC MOSFETs under normal operating conditions and typical faulty states, the experimental results indicate that the combination of the three characteristics can be employed to identify different types of faults.