Owing to the bad condition of the transmission line, it is difficult to accurately capture the equipment operating status, especially the battery status. In this paper, a method for predicting the state of charge of the transmission line fault location device based on RBF neural network is proposed. Firstly, the simulation test platform of transmission line is built in the laboratory, thereby the operation data under constant current discharge are obtained. On this basis, the neural network training samples are constructed and substituted into the algorithm model for training, with the test samples are used to verify the algorithm model. Tests show that the RBF neural network algorithm considering the battery terminal voltage and discharge time can effectively identify the state of charge of the battery capacity of the transmission line fault location device, which provide a reference for the health and device operation status analysis of the transmission line fault location device.