An Intelligent Fault Detection and Faulty Line Identification Scheme for Hybrid Microgrid using Ensemble of kNN approach

Abstract

Traditional power is reliant on fossil fuels, which are slowly depleting and causing environmental concerns. As a result, energy dependency has gradually moved towards renewable distributed generation (DG) resources in the current power distribution system. With the more penetration of DGs, protection of hybrid microgrid network is becoming increasingly complex. In this work, a protection scheme is developed for hybrid microgrid with Discrete wavelet transform (DWT) and Ensemble of k-nearest neighbor (kNN) to perform the dual task of fault detection and faulty line identification. Testing of the proposed scheme has been performed against various internal and external fault scenarios. In terms of accuracy, the proposed scheme outperforms single kNN and linear support vector machine (SVM) classifiers.