Comparison of CWT & DWT based algorithms in combination with ANN for protection of power transformer

This paper mainly presents comparison between intelligent algorithms based on Continuous Wavelet Transform (CWT) and Discrete Wavelet Transform (DWT) in combination with Artificial Neural Network to discriminate the magnetizing inrush current signals from the internal fault current signals of the power transformer. This work also includes development of CWT and DWT based preprocessing units to extract distinguishing attributes from inrush and internals fault signals, which are quicker, completely independent from the traditional second harmonic restraining methodologies. Extracted attributes are fed to ANN based post processing unit to classify inrush current and internal fault current of power transformer. Proposed scheme achieves proper classification with high discrimination rate and least error, avoiding mal tripping of power transformer.