A Simple and Fast Unified Threshold Fault Diagnosis Method for DC Microgrids

Threshold-based fault diagnosis methods are widely used in DC microgrid line protection owing to their rapid response. However, existing threshold settings still rely primarily on simulation tuning and empirical trials, which limit interpretability and general applicability. To overcome these limitations, this article proposes a unified threshold-based fault diagnosis framework for pole-to-pole (PP) and pole-to-ground (PG) faults in DC microgrids. First, a multiple mean skewness deviation (MMSD)-based fault feature extraction algorithm is developed to capture the statistical and distributional characteristics of fault filter capacitor (FC) currents without requiring high-frequency sampling. Then, faults are rapidly detected by comparing MMSD feature differences with a unified threshold, simplifying the diagnostic process while enhancing reliability and accuracy. Furthermore, a single-ended differential fault classification strategy based on bipolar current features is introduced, enabling accurate fault identification for a wide range of fault resistances within 1 ms. Simulation and experimental results validate the proposed method’s effectiveness, robustness, and adaptability, offering a practical solution for fast and reliable DC microgrid protection.