Research on Fault Detection and Localization in Meshed DC Microgrids Based on Local Information Measurements

DC microgrids have emerged as a research hotspot due to the demand for efficient integration of renewable energy. However, their low inertia and absence of zero-crossing points render traditional AC protection strategies ineffective. This paper proposes a fault detection method based on local measurements and event classification, which achieves rapid detection by analyzing the first-order/second-order derivative characteristics of fault currents. A Fault Current Limiting Offset (FCLO) is introduced to enhance localization accuracy. A ring-type DC microgrid model is established in Matlab/Simulink, and simulation results demonstrate that the proposed method completes fault detection and isolation within 0.2–0.85 ms without relying on communication channels, significantly improving system reliability.