Abstract:

In order to suppress the white noise of partial discharge on-line detection,this paper proposed an Improved Threshold Algorithm of Lifting Wavelet Transform.The noisy signals are first processed by lifting wavelet decomposition to get the scale coefficients on the highest decomposition scale and wavelet coefficients on each decomposition scale.Then the scale coefficients are reserved and the wavelet coefficients are processed with an improved threshold selection method and a derivative threshold function.Finally,the de-noised signal can be obtained by reconstructing the scale coefficients and processed wavelet coefficients.The de-noising results of simulated partial discharge signals and measured partial discharge signals show that:compared with the traditional soft threshold function method,the proposed method has a better effect on removing the white noise.

Key words: partial discharge, on-line detection, white noise, lifting wavelet, de-noising