The Fault Detection and Protection Setting Optimization Method for Photovoltaic-Storage Dc Grids

DC fault detection is a critical challenge in DC grids. To address detection errors arising from fixed protection thresholds influenced by grid conditions in photovoltaic-storageDC distribution networks, this study proposes an optimized protection threshold setting and fault detection algorithm based on Artificial Neural Networks (ANN). The approach aims to minimize fault isolation time, maximize protection sensitivity, and reduce fault impact on loads. An optimization model is developed to achieve these objectives. The algorithm's innovation lies in its adaptive threshold determination, enhancing detection reliability. By configuring relays locally without requiring highspeed communication, the method enables rapid DC fault detection. MATLAB/Simulink simulations validate that the proposed scheme effectively resolves protection mismatch issues, adapts to diverse network structures, and significantly improves distribution network protection performance.