A New Fault Line Selection Method for Resonant Distribution Network based on WDMI and Hausdorff Distance in Double Selected Frequency Bands

Abstract

Aiming at the problems of insufficient utilization of fault features and single criterion in the existing single-phase grounding fault line selection(FLS) methods of resonant distribution network, a FLS method for distribution network based on wavelet decomposition modulus integral(WDMI) algorithm and Hausdorff distance algorithm is proposed. Firstly, the transient zero-sequence current of each line is decomposed by coif5 wavelet function to remove the lowest frequency band containing the power frequency component and the two high frequency bands of the noise signal. The wavelet reconstruction coefficients in the remaining frequency bands are calculated and the modulus integral matrix is constructed. According to the principle of maximum sum of modulus integral, the double selected frequency band(D-SFB) with obvious fault characteristics is determined. The preliminary FLS result is determined by the amplitude relationship satisfied by the fault in the D-SFB. Then the Hausdorff distance matrix in the D-SFB is constructed by the Hausdorff distance algorithm and the line is selected again according to the waveform correlation. Finally, the final FLS is realized according to the equal weight voting. A large number of simulation results show that the FLS method can achieve reliable FLS under different fault situations.