Interval dependent wavelet de-noising technique for high frequency transient signals analysis during impulse testing of transformers

In transient phenomena of transformer, one of the important research areas of interest is removing the noise in the measured signals during impulse test and to find if there exists a relation between the high frequency transient signals to declare about pass criterion or fail criterion of the transformer at manufacturing location. In this paper, to remove noise and achieve better SNR, interval dependent wavelet de-noising technique is utilized against multivariate de-noising technique. After proper selection of de-noising technique to identify the winding fault in transformer, nonlinear interpretation technique is employed. To prove the usefulness of the proposed technique, impulse testing data of 66.7 MVA and 250 MVA transformers are used.