High-impedance fault detection in power distribution grid systems based on support vector machine approach

Today, microgrids are used increasingly in different types because of its several financial and environmental benefits for customers, societies and nations. Its implementation, however, makes significant theoretical and practical challenges, such as fault detection that could make crucial damage to the utility grid and microgrids. This paper, accordingly, developed a novel, fast and accurate method based on Support Vector Machines (SVM) approach for High Impedance Fault (HIF) detection. The proposed method applied to a typical distributed generation system for detecting single line, double line and triple line HIFs. Also, the behavior of current signals of the other phases are investigated during faults occurrence. The simulation results show how this algorithm can separate faults condition among the other fault-like phenomena and gets a better response in comparison to present methods like Wavelet Transformation (WT) and Artificial Neural Networks (ANN). This method will boost the development of renewable energy usage by reducing fault detection delays and operational risks.

Automated summary

This study proposes a novel method for High Impedance Fault (HIF) detection based on Support Vector Machines (SVM). It can quickly and accurately detect faults in distributed generation systems. Compared to other methods like Wavelet Transformation (WT) and Artificial Neural Networks (ANN), this method provides better response and ability to differentiate faults from other fault-like phenomena, which helps reduce fault detection delays and operational risks.