In this paper an approach based on Neural Network sing Discrete wavelet transform is proposed to locate the fault of the three phase transmission line. The different test cases are carried out on the basis of extensive simulation studies using MATLAB toolbox for different fault operating conditions. The proposed method requires the three phase post fault currents to be measured (fault signals) at one end (sending end).In order to accomplish the task, features of the post fault currents are extracted in terms of Wavelet Energy using Discrete wavelet transform (DWT) first and then fed to train with feed forward Neural Network (FNN) for each case. The FNN estimates the location of the fault with a very high degree of accuracy which results in, an improvement over the existing iterative method of fault location, for a wide variation of fault operating conditions hence the method is reliable and robust.
1 December 2018 in Journal of Electrical Systems and Information Technology