Faulty Feeder Detection Under High Impedance Faults for Resonant Grounding Systems Based on Empirical Wavelet Transform and Synthetic Voting

In order to solve the problem that the line cannot be accurately selected when a high-resistance fault (HIF) occurs in a resonant grounded distribution system, by analyzing the impedance of the zero-sequence network, it can be concluded that within the characteristic frequency band, the current amplitude of the healthy feeder is smaller than that of the faulty feeder, and the current phase of the healthy feeder is opposite to that of the faulty feeder. Therefore, a high-resistance fault line selection method for a resonant grounding system based on empirical wavelet transform (EWT) is proposed. Firstly, the EWT algorithm is used to extract the transient current in the characteristic frequency band accurately. Secondly, the characteristic energy ratio, characteristic mutual correlation coefficient, and characteristic integral area are constructed for the transient current to measure the energy, phase, and polarity of the transient current. Finally, the high-resistance fault line selection is accurately realized according to the principle of “the minority obeys the majority”. Different types of high-resistance arc fault conditions are tested in a radial resonant grounding system with distributed generations (DGs). The test shows that the method has good adaptability and can correctly select the line under high-resistance faults.