Find in Library

The small-current grounding fault in distribution network is hard to be located because of its weak fault features. To accurately locate the faults, the transient process is analyzed in this paper. Through the study we take that the main resonant frequency and its corresponding component is related to the fault distance. Based on this, a fault location method based on double-end wavelet energy ratio at the scale corresponding to the main resonant frequency is proposed. And back propagation neural network (BPNN) is selected to fit the non-linear relationship between the wavelet energy ratio and fault distance. The performance of this proposed method has been verified in different scenarios of a simulation model in PSCAD/EMTDC.