Three-phase Inverter Open Circuit Fault Classification Using Mean to RMS Current Ratio

This paper proposes an efficient and reliable method for diagnosing open-switch faults in three-phase inverters. In contrast to conventional signal-based diagnostic approaches that require additional measurement devices or are highly sensitive to load variations, the proposed method classifies fault types by using the mean-to-RMS (mean/RMS) ratio of current signals as a key indicator. The mean/RMS ratio quantitatively demonstrates the distortion level of current signals and remains insensitive to load variations, thereby ensuring consistent diagnostic results under diverse load conditions. Designed as an algorithm-based solution, this method minimizes computational complexity and memory consumption while maintaining accurate fault classification. The proposed approach categorizes fault conditions into six distinct classes based on the RMS values and mean/RMS ratios of three-phase currents. The effectiveness of the proposed diagnostic method is validated through simulations and experiments with reliable performance under varying load conditions and fault scenarios. This study presents a practical diagnostic solution to improve the stability and reliability of power conversion systems, with the potential for experimental validation and broader application in various fields.