Authors: Bhavesh Bhalja, R.P. Maheshwari
Addresses: Department of Electrical Engineering, AD Patel Institute of Technology, 388121, Anand, Gujarat, India. ' Department of Electrical Engineering, Indian Institute of Technology Roorkee, Roorkee-388121, Uttrakhand, India
Abstract: Using the powerful analysing and decomposing features of wavelet transform, a new fault detection scheme is presented. Further, the same technique has been extended for fault classification and accurate fault location from the relaying point to the fault using a radial basis function neural network (RBFNN) which provides a more efficient approach for training and computation. The relaying scheme depends on the three line voltages and currents of the transmission line at each end. One of the key points of this paper is preprocessing module of the measured signals to extract the most significant features from the signals. Extracting distinctive features (frequency components) from the line currents and voltages and feeding them to a RBFNNs result in an accurate and fast transmission relaying scheme. The proposed wavelet based RBFNN scheme provides encouraging results for detecting, classifying and precisely locating fault events independent of different fault and system conditions.
Keywords: wavelet transform; WT modulus maxima; WTMM; radial basis function; RBF neural networks; RBFNNs; fault detection; fault classification; fault location; transmission lines; feature extraction.
International Journal of Power and Energy Conversion, 2011 Vol.2 No.4, pp.353 - 364
Accepted: 15 Nov 2010
Published online: 14 Aug 2011 *