SESB'S 275 kV transmission line fault classification using wavelet-based neural networks

Abstract

Fault in transmission lines might risks and contributes to the biggest impacts on power system security. Hence, early fault diagnosis is a prominent area of investigation in power utility with relevant intelligent system applications. The recent infrastructure for fault and event monitoring in transmission power systems using Disturbance Fault Recorders (DFR) becomes a common practice as it provides a high amount of data not only for grid operator for normalization purposes but beneficial to scholars in extending their progress research on fault classification studies. DFR is proven reliable in observability, but the large amounts of data and waveform may time consuming for the engineers to analyse the details. Therefore, there is a need to classify type of fault occurrence as both current and voltage waveforms contain significant high frequency transient signals during disturbance. Many studies revealed a lot of techniques and hybrid approaches for fault detection and classification. Though, among the various techniques reported, a combination of wavelet-based and neural networks techniques has been proven as the best approach to determine the correct fault type and classification with the approximation of 91% of accuracy. The objective of this study is to perform fault identification by using Discrete Wavelet Transform (DWT) module to extract feature of the transient signal and gain wavelet coefficient. The obtained data will be classified using Artificial Neural Networks (ANN) architecture by categorizing the types of fault based on details wavelet analysis. The feasibility of Discrete Wavelet Transform and Artificial Neural Network algorithm is tested on SESB’s 275kV overhead transmission line using MATLAB software. The said techniques are simple and accurate in fault detection and classification. The algorithm produces good results in simulations, indicating that it will be highly beneficial in the future operational of power system.