Novel Adaptive Fault Detection Strategy in DC Microgrid Utilizing Statistical-Based Method

Abstract

The DC microgrid emerges as a proficient solution for anonymously evolving DC loads and their accompanying applications. Overcurrent-based relaying schemes can't provide such systems with the desired sensitivity and selectivity. The differential scheme has become a prominent solution with an enriched data processing technique and advanced communication framework in real-grid scenarios. The faults may remain undetected with high fault impedance where the current direction doesn't change at fault inception in a differential scheme. The conventional local protection schemes result in limited performance with fixed relay settings. The threshold selection is multifaceted in various protection schemes and largely depends on system topology, leading to catastrophic failure with any system or operating conditions alteration. This article proposes an adaptive statistical Fano Factor tool-based scheme to detect and classify faults with enhanced sensor tolerance capability. The scheme utilizes the current data at line ends. The performance of the proposed method is tested under various operating scenarios, including instantaneous switching operations of sources or loads and an evolving case, where fault impedance varies during fault. The method is fast, effective with high impedance faults, and immune to system disturbances. Adaptability, robustness, sensitivity, and high efficacy are its strategic features even with different system topologies. The method's performance is tested using data obtained from PSCAD/EMTDC simulations for numerous cases.