High impedance fault detection and isolation in power distribution networks using support vector machines

Abstract

This paper proposes an accurate High Impedance Fault (HIF) detection and isolation scheme in a power distribution network. The proposed schemes utilize the data available from voltage and current sensors. The technique employs multiple algorithms consisting of Principal Component Analysis, Fisher Discriminant Analysis, Binary and Multiclass Support Vector Machine for detection and identification of the high impedance fault. These data driven techniques have been tested on IEEE 13-node distribution network for detection and identification of high impedance faults with broken and unbroken conductor. Further, the robustness of machine learning techniques has also been analysed by examining their performance with variation in loads for different faults. Simulation results for different faults at various locations have shown that proposed methods are fast and accurate in diagnosing high impedance faults. Multiclass Support Vector Machine gives the best result to detect and locate High Impedance Fault accurately. It ensures reliability, security and dependability of the distribution network.

A Literature Review on Methodologies of Fault Location in the Distribution System with Distributed Generation

“…The approach shows good accuracy but requires retraining of SVM whenever there is a change in system topology. Another SVM‐based method for high‐impedance fault detection and isolation in a power distribution network is proposed in a previous study …”

“…Another SVM-based method for high-impedance fault detection and isolation in a power distribution network is proposed in a previous study. [95] The main challenges in regard to the possible implementation of intelligent techniques for fault location are as follows: 1) The choice of the mother wavelet, support size, and noise in the signal affect the performance of wavelet-based methods. 2) The ANN-based method highly depends on the amount and quality of training data provided to the ANN algorithm.…”

Fault Diagnosis of Rotating Electrical Machines Using Multi-Label Classification

“…In contrast to conventional signal processing based fault detection techniques [65], recently a few attempts were made for the application of intelligent algorithms [66,67] including new approaches to Fault Detection and Isolation (FDI) [68] based on fuzzy logic, decision trees, neural networks, and further machine learning techniques [69][70][71][72][73]. However, most of them rely on the measurement and processing of vibration signals, which require at least one vibration sensor, which demands extra costs for its proper installation and maintenance [74][75][76][77]. In addition, a technician needs knowledge and a good amount of experience to correctly use such sensors [78][79][80][81][82][83][84].…”

Multi-Label Classification for Fault Diagnosis of Rotating Electrical Machines

“…In contrast to conventional signal processing based fault detection techniques [65], recently a few attempts are made for the application of intelligent algorithms [66,67] including new approaches to fault detection and isolation (FDI) [68] based on fuzzy logic, decision trees, neural networks, and further machine learning techniques [69][70][71][72][73]. However, most of them rely on the measurement and processing of vibration signals, which require at least one vibration sensor, which demands extra costs for its proper installation and maintenance [74][75][76][77]. In addition, a technician needs knowledge and a good amount of experience to correctly use such sensors [78][79][80][81][82][83][84].…”