Investigation of suppressing PD white noise by compound information of complex wavelet transform
Ju Tang, XU Zhong-rong, Caixin Sun · European Transactions on Electrical Power · 2008
Abstract According to the distinct properties between the complex wavelet coefficients of partial discharge (PD) signals and white noises, this paper presents the specific compound information (CI) of WTRInseries for complex wavelet transform (CWT), which are utilized to suppress white noises. Through the simulation case, the influence of n‐index value on denoising effect is discussed and the optimal CI of corresponding to different PD waveforms is analyzed by the simulation investigation. The parameters of variable tendency parameter (VTP), normalized correlation coefficient (NCC), and signal‐to‐noise ratio (SNR) are defined to synthetically evaluate the denoising effect. The db4 complex wavelet is constructed to suppress the white noise in ultra high frequency (UHF) PD detection system. The results indicate that the denoising effect of the CI for CWT is better than the simple information (SI) for CWT and the CI of WTRInseries have the optimal denoising effect to suppress white noise for UHF PD signals. Copyright © 2008 John Wiley & Sons, Ltd.