A LSTM Line Selection Method of Single-phase Ground Fault Based on Big-data Platform

Due to the complex structure of the distribution network, a large number of power supply branches, and various neutral grounding modes, there is a problem with inaccurate fault line selection results. To identify single-phase faults quickly and accurately when they occur and reduce the risk of system operation, this paper analyzes the conventional electrical quantities and proposes a single-phase ground fault line selection method based on Long Short-Term Memory (LSTM) network. The simulation example in this paper collects actual operation data through the dispatching platform and creates an LSTM fault line selection model based on Keras. The comparison between LSTM and other algorithms shows that this method has the characteristics of high prediction accuracy.