Investigation on fault identification in a power system under critical conditions using wavelet based techniques

Abstract

In electrical power system fault identification under critical conditions is one of the major protection issues which are investigated with various techniques Digital Signal Processing is one among them which widely employs Wavelet Packet Transform WPT based methods for fault detection In this work protection issues in Distribution and Transmission system are discussed Distributed Generation is widely employed nowadays due to the increasing demand of electrical energy The protection issues are increasing in Distribution systems due to the incorporation of Distributed Generation These issues occur during critical conditions such as islanding loss of generation grid reconnected condition and addition of generation The short circuit current during islanding condition and loss of generation is lesser than that of the fault during grid connected condition Hence there is a possibility of not tripping the relay for fault during islanding condition and loss of generation During grid reconnected condition and addition of generation the current increases and the protective relay identifies this non fault condition as fault Similarly the occurrence of symmetrical fault and its identification during power swing is also a critical protection issue in transmission line Hence in this work overcurrent protection schemes with three techniques namely Direct axis quadrature axis Wavelet Singular Entropy dqWSE Improved Weighted Wavelet Packet Entropy WWPE and Fisher Asymmetry Coefficient Wavelet Packet Transform FAC WPT are proposed to identify the occurrence of fault under the above mentioned critical conditions These techniques use WPT as pre processing technique newline newline newline