Fault cause Identification of Distribution System Based on Association Rule

Identifying the fault cause in the distribution system is very challenging. In the face of massive fault information, accurately identifying the cause of faults is the key to accurate fault warning. Aiming at five fault causes in the distribution systems, this paper analyzes the correlation between different fault causes with text features and electrical features. FP-Growth algorithm is used to mine association rules, and a comprehensive database containing all historical fault events and a knowledge base containing effective strong association rules are constructed. The Jaccard similarity coefficient is used to identify fault cause. Finally, the accuracy of the proposed algorithm is verified by field measured data.