Identification of wavefronts in Partial Discharge acoustic signals using discrete wavelet transform

This paper describes a method to discriminate the wavefront of the acoustic signal obtained from partial discharges. The technique used here takes advantage of the wavelet transform in order to extract, in a very precise way, the characteristics related to the desired wavefront. This information is used to locate the source of the partial discharge occurring in oil-insulated power transformers. The methodology to find the correct spot of the partial discharge uses genetic algorithm to solve a set of non-linear equations. Practical results are shown and the methodology is described in detail.