Wavelet-ANN Based Fault Location Scheme for Bipolar CSC-Based HVDC Transmission System

Wavelet transform (WT) is a widely used method for analyzing fault transients in the faulty signals to achieve fault diagnosis. It enhances the efficiency and accuracy in the operation of Artificial Neural Network (ANN) when used as a companion. The proposed work introduced an accurate way to predict the fault location for bipolar Current Source Converter (CSC) based High Voltage DC transmission system using ANN with the combination of WT. The proposed algorithm uses the extracted features of AC sinusoidal voltage, DC voltage and current signals achieved at both poles of the rectifier end in the line. The input signals are sampled at 1 kHz which is less in comparison with the other existing techniques. The numerous simulation results validate the efficiency and stability of the proposed ANN-based strategy to deviation in the fault resistance and different fault locations. The proposed strategy estimates the location of fault with mean percentage error of 0.0289%. The advantage of this strategy is that it does not necessitate any telecommunication system as it utilizes the measurements of only rectifier end.