A DT-CWT and Data mining based approach for High Impedance Fault Diagnosis in Micro-grid System

Abstract

The high impedance faults (HIFs) detection in the micro-grid is very difficult for over-current protection owing to the less fault magnitude. As this is a practical problem in the distribution feeder, the detection of HIFs plays a vital role in industry 4.0. A dual-tree complex wavelet transform(DT-CWT) and data mining approaches have been used for fault detection and fault classification. For the extraction of wavelet features by DT-CWT, the residual voltage is used as input and fed to various data mining approaches to separate the HIF events and the created confusion matrix gives the best results. The correctness of the method implemented is investigated with other data mining approaches like support vector machine(SVM), k-nearest neighbor(KNN), and ensemble classifiers. As correlated to further methods, the proposed technique is 100% accurate. To execute the proposed method, the interconnection of DG with a wind-integrated micro-grid system is designed using Power System Computer-Aided Design (PSCAD) software.