A new method of single-phase ground fault line selection based on principal component analysis and support vector machine

Abstract

For the existing single-fault line selection methods that cannot accurately identify all single-phase grounding faults in a small current grounding system, this paper proposes a new method of small current single-phase grounding fault line selection based on the combination of principal component analysis and support vector machine. This method first uses wavelet transform to denoise and divide the transient zero-sequence current of the bus and each feeder; then uses the zero-sequence energy method to first detect the bus fault. If the detected fault is a bus fault, end of line selection; if the detected fault is not a bus fault, use principal component analysis to reduce the dimension of the fault zero sequence current signal. Select the three most influential factors, and enter the support vector machine classifier to finally get the fault line. Use Matlab/Simulink to build a simulation model and simulate fault experiments. The final experimental results show that this method has high stability in line selection and is not affected by noise, ground resistance, fault angle, fault distance, and the number of feeders. Identity bus faults and feeder faults, and select lines accurately.