Extraction of partial discharge pulse via wavelet shrinkage

It is an important step in the study of partial discharges within power equipment to extract the partial discharge (PD) pulses from the noise. The wavelet analysis shows its strong power in extracting tiny signals and eliminating noise, and a lot of de-noising algorithms using wavelets have been developed. Based on the comparison among different wavelet de-noising algorithms, a new method-wavelet shrinkage is used to extract PD pulses from the signal polluted by more than one type of noise. With the numerical simulation, different influencing factors, such as the class of wavelets, the scales of wavelet decomposition and the type of threshold are analyzed in detail to obtain the best results. The simulation results show that wavelet shrinkage is valid for extracting PD pulses from strong noise.