Research on Single-phase High-impedance Ground Fault Based on Neural Network

Abstract

Because of the complicated operation environment surrounding the power grid and the influence factors, the single phase fault often occurs through the nonideal conductor, such as the branch of a tree, sand, ponds and other media grounding, the grounding resistance of such faults is non-linear and resistance value is large, known as high impedance fault (High impedance fault, referred to as HIF). High resistance ground fault if not timely processing, May result in phase failure or short circuit failure, which is disadvantageous to the safety and stability of power system., serious will also be a threat to personal safety, therefore, the fault should occur in a timely manner to identify and take appropriate measures to remove the fault line. However, if there is a serious resistance fault, the fault time will be very long, and the power outage will be intermittent, which is difficult to detect. At the same time, the traditional protection measures are unreliable, because the fault point is usually combined with intermittent current, so the fault flow is random, and determining and detecting large faults is the focus and challenge of secondary energy protection.. In this paper, we propose a neural network-based single-phase high-resistance grounding failure recognition method and study it.