A method to detect high impedance faults in distribution feeders

Abstract

Detection of High Impedance Faults (HIFs) facilitates predictive maintenance in distribution systems. In this paper, a methodology has been developed for detecting HIF in distribution feeders. The method uses multi-resolution analysis with the help of Discrete Wavelet Transform to extract features of currents to be monitored on a feeder. These features are then used to train a Support Vector Machine for classification, i.e., to differentiate between HIF and switching events. Test results based on data obtained simulations on a prototype feeder show that the method is effective in identifying HIFs.