Research on single-phase ground fault identification method based on support vector machine

Abstract

Abstract Single-phase grounding faults seriously affect the stable operation of the distribution network system. Timely and accurate identification of early fault characteristics is essential to achieve early warning of faults and reduce the occurrence of permanent faults. In single-phase grounding faults, the current is small and the fault characteristics are complex, which leads to existing identification methods having low accuracy and poor reliability. This research presents a method for identifying single-phase grounding faults using the Support Vector Machine (SVM) framework. Firstly, the process begins with an in-depth examination of the waveform characteristics typical to these faults, followed by extracting critical data features necessary for fault identification. Then the fault affiliation degree is calculated as the input to the SVM model, and the fault category is the output. Finally, the algorithm model is simulated and analyzed by using PSCAD/EMTDC and the results indicate that the proposed algorithm accurately detects single-phase earth faults in low-current earth systems.