High impedance fault detection in Microgrid system using Discrete wavelet transform and Support vector machine

Abstract

Over-current relaying scheme is unable to identify high impedance faults (HIFs) in the micro-grid system due to the low current magnitude. In this paper, the realistic nature of the HIF arc is discussed and the proposed scheme separates the faulty events. The features are extracted from Discrete wavelet transform (DWT) by taking residual power as input and fed to support vector machine(SVM) classifier for fault categorization. The confusion matrix clearly shows the variation of true positive rate(TPR) and false-negative rate(FNR) of faulty event and non-faulty event respectively. Also, it shows the variation between the predicted class and the true class. The accuracy of the technique is tested on a 5 bus micro-grid system integrated with a wind generator which is simulated using Power system computer-aided design (PSCAD) software.