OPGW (Optical Fiber Composite Overhead Ground Wire) has large capacity, long service life and lack of spare fiber core in some sections. Its working state is directly related to whether it can ensure the safe and reliable transmission of electric energy. In this paper, the mechanism of OPGW fault was analyzed by data mining technology, and the fault type that was most associated with OPGW selection was found out by association rules, and the relationship between short-circuit current and short-circuit position of OPGW was obtained. In the process of line running current prediction under different methods, the actual line running current in the first hour was 119 A, the traditional method predicted the running current 132 A, and the method in this paper predicted the line current 123 A, which was closer to the actual situation. The large data analysis method in this paper is helpful to OPGW fault diagnosis and monitoring, and to find out the cause of OPGW cable fault.