Improved fault classification in series compensated EHV transmission line using Wavelet transform and Artificial Neural Network

摘要

The paper presents a novel technique proposed for protection of multi bus Extra High Voltage (EHV) transmission system having multiple series compensation in line sections. It is found that conventional methods fail to protect such system under wide range of operating scenarios involving faults in series compensated transmission lines. The protective scheme presented here is in the algorithm which utilizes single end post fault current data of three phases of the line. Here, a combination of Multi resolution Wavelet transform along different structures of Artificial Neural Networks (ANN) has been tested to attain an Adaptive Protection with high accuracy. A statistical learning scheme is proposed involving features extracted from Multi resolution Wavelet transformation. The features are divided in to various sets to provide learning of the ANN, which are later tested on unseen patterns to derive the Classification of faults. With detailed and extensive simulations, the feature extraction is provided which significantly captures the characteristics of the system under study during its advert operations. It is shown that the proposed methodology using Pattern Recognition scheme gives accurate and more reliable results in Identifying and Classifying faults as opposed to classical methods.