Wavelet base selection for de-noising and extraction of partial discharge pulses in noisy environment

Wavelet-based de-noising is used to separate partial discharge (PD) signals from the noises resulting from measurement circuits or the surrounding environment. PD de-noising by using the wavelet shrinkage method is capable of separating the noise component to some extent, but the selection of the wavelet base has a remarkable effect on the de-noising results. The wavelet base is directly related to the distortion of the PD waveform and quality of the de-noising process. Although there are applications on PD noise separation in the literature, the selection of the wavelet base, which affects the evaluation of the PD characteristics, is still challenging. Instead of using correlation-based wavelet base selection for de-noising PD data, in this study a novel wavelet base selection method based on the most informative sub-band energy and entropy for separating noise from PD pulses is introduced and successfully applied to raw data obtained from the PD measurement set-up. The advantage of the proposed method is that the wavelet base selection solution is automatic and independent of the original noise-free pulse waveform. This study shows that the proposed method is useful for the extraction of noisy PD pulses by describing the basic discharge parameters such as discharge amplitude and the duration and time of occurrence more clearly.