High impedance fault location - case study using wavelet transform and artificial neural networks

Abstract

This paper presents a high impedance fault (HIF) location method in distribution systems. The estimation of the HIF location is carried out by using an artificial neural network (ANN). The ANN inputs are both three phase currents and its energy of the Wavelet Coefficients obtained during the HIF. The Wavelet mother function used was Daubechies 48. The ANN was trained and the performance of the proposed HIF location was evaluated by using a HIF modeled with two series time-varying resistances controlled by the Transient Analysis of Control Systems (TACS) of the Alternative Transient Program (ATP). Actual data obtained from several staged HIFs in different soil types on Energisa utility (a Brazilian distribution system) were used to adjust the HIF model. (4 pages)