Extraction of partial discharge signals using wavelet transform

The application of Wavelet Transform to extract partial discharge signals from narrow-band disturbance such as sinusoidal interference is described in this paper. Mallat's algorithm of dyadic wavelet transform was used to filter out PD signals. Two kinds of numerical simulating signals and some on-site experimental data were processed by this method. The results showed that the noise could be cut down efficiently when it was the high frequency sinusoidal noise. Thus the purpose of extracting PD pulses and reducing noises could be realized. Comparing with previously developed methods, its processing speed is faster. Some: factors that effected the extracted results were also discussed.