An experimental review on application of wavelet transform on partial discharge evaluation of power transformers

Partial Discharge (PD) measurement and analysis is now a popular method in condition monitoring and dielectric diagnosis of power transformers. Discrete Wavelet Transform (DWT) is a powerful and well known method to analyze PD data in both time and frequency domains. In this contribution, the application of this transform on de-noising of recorded PD signals, defect classification and separating PD pulses of simultaneous defects is investigated and reviewed experimentally. For this purpose, an unconventional ultra wide band acquisition system is used to detect and record PD pulses data in time domain. Experimental works including those made during on-site PD measurement on a power transformer and PD commissioning test of a dry type transformer as well as experiments on artificially made PD defects are investigated.