고조파 성분을 이용한 고저항 지락 사고 검출 기법에 관한 연구

A Study on High Impedance Fault Detection Method Using Harmonic Components

A high impedance fault on the multi-grounded three-phase four-wire distribution system can not be detected by conventional overcurrent sensing devices. In this paper, the neural network is used to detect high impedance faults. The proposed algorithm using back - propagation neural network is demonstrated by simulation with the staged fault test data. The harmonic components of current and the phase of voltage are used as the inputs of neural network. Results of the simulation can be used as a reference for the development of a high impedance fault detector.