An Identification Method of Fault Type Based on GWO-SVM for Distribution Network

To achieve the accurate and fast identification of fault type for distribution network, a fault type identification method based on grey wolf optimization algorithm (GWO) and support vector machine (SVM) is proposed in this paper. When faults occur, the zero-sequence voltage of fault signals are obtained, and then the low-frequency energy values of them are calculated. Then the wavelet singular entropy value is calculated by Singular value decomposition (SVD) method and entropy theory. The low-frequency energy values and singular entropy values are used as the input of GWO-SVM classifier to identify the fault type. The results of simulations show that the proposed method can identify all grounding faults and short-circuit faults correctly, and the prediction accuracy is better than it of SVM classifier.

I. Introduction

Distribution network is complicated and inevitably affected by various types of faults under actual working conditions. Grounding faults and short-circuit faults are becoming the common electrical faults. Therefore, fault type identification is the premise of accurate fault location and fast recovery of power supply.