Research on Extraction of Partial Discharge Signals Based on Wavelet Analysis

The suppression of noise is crucial prior to any partial discharge (PD) data analysis in on-line PD measurement. Recent research shows that the Wavelet transform (WT) has achieved good effect in noise rejection in on-line PD detection. This paper presents the vital issues of WT implementation in PD signals extraction, including optimal mother wavelet selection, decomposition level selection, threshold rule determination and thresholding function determination. The noise rejection effect of WT de-noising method is proved by simulated and field detected data.