Comparison of fourier & wavelet transform methods for transmission line fault classification

Nowadays, power supply has become a business asset. The quality and reliability of power system needs to be maintained in order to obtain optimum performance. Therefore, it is extremely important that transmission line faults from various sources to be identified accurately, reliably and be corrected as soon as possible. In this paper, a new technique is discussed by which avoiding noise in fault detection in high voltage transmission lines is achieved. Later, a comparative study of the performance of Fourier transform and wavelet transform based methods combined with protective relaying pattern classifier algorithm Neural Network for classification of faults is presented. A new classification method is proposed for decreasing training time and dimensions of NN. The proposed algorithms are based on Fourier transform analysis of fundamental frequency of current signals in the event of a short circuit. Similar analysis is performed on transient current signals using multi-resolution Haar wavelet transform, and comparative characteristics of the two methods are discussed.