Research on Single Phase Grounding Fault Detection Technology for Railway 10 kV Distribution Lines

The use of a single electrical threshold judgment method can only roughly determine whether there are abnormal electrical fluctuations in the circuit, and the reliability of fault detection is poor. To solve the above problems, research on single-phase grounding fault detection technology for railway 10 kV distribution lines will be carried out. Firstly, sensors are installed on the 10 kV distribution line of the railway to collect signals. The Haar wavelet decomposition is used to calculate the sub signal energy, standard deviation, and other parameters, which are combined into feature vectors to extract the characteristics of the 10 kV distribution line of the railway; afterwards, the support vector machine algorithm is used to preprocess the feature vectors and input them into the model. Based on the output results, faults are determined to achieve automatic detection of single-phase grounding faults in railway 10 kV distribution lines. The experiment showed that the research method maintained a fault detection accuracy of over 90% even when the grounding transition resistance reached 2000 Ω, significantly reducing the number of misjudgments compared to the comparative method, and improving the reliability of single-phase grounding fault detection for railway 10 kV distribution lines.