Discrete Wavelet Transform Based Analysis of Transformer Differential Current

Abstract The conventional differential relay has been used to detect the internal faults within the transformer. But, in field certain mal-functioning of differential relay has been reported which results in unnecessary tripping of differential relay. This leads to the need for development of improved classifier technique. But, for that all possible operating conditions of transformer has to be studied. In this paper, the Authors have used discrete wavelet transform to analyze the transformer differential current during various operating conditions, since wavelet transform gives good information about frequency and time domain simultaneously. Certain statistical features have been extracted from the decomposed signals and they can be used as an input to the improved classifier algorithm Daubechies 6 is used as mother wavelet for the analysis using DWT and signal has been decomposed up to level 4