A Data-Driven Fault Line Selection Method For Shore Power System

The reliable operation of the shore power supply system greatly affects the reliable power supply of the entire shore power system. However, due to the complex topological structure of the shore power supply system, the changeable operation mode, and the numerous lines, fault detection is difficult. The correct rate of line selection of various fault line selection devices currently used is not high. This paper proposes a single-phase-to-ground data driven fault line selection method for small-current grounding systems based on big data. First, the algorithm accesses and stores the big data related to the failure of the onshore power supply system. The raw data is processed into feature data using preprocessing and feature extraction techniques. Then, the big data mining algorithm is used to mine the mapping association between data features and fault lines. Finally, the mapping relationship is used for line selection, and an advanced big data mining algorithm based on massive and multi-dimensional fault-related data is proposed to improve the accuracy of line selection.