The invention discloses a
partial discharge signal denoising method based on a
wavelet adaptive threshold. The
partial discharge signal denoising method based on the
wavelet adaptive threshold comprises the following steps of (1) inputting a
partial discharge signal to be denoised, (2) carrying out
wavelet multi-scale
decomposition on the partial
discharge signal to obtain high-frequency coefficients of
decomposition scales and a low-frequency coefficient of a maximum
decomposition scale, (3) using a non-negative garrote
threshold function and a adaptive threshold
selection method based on
particle swarm optimization to carry out quantitative
processing on high-frequency coefficient components obtained in the step (2) so as to remove
noise components, storing the result to serve as new high-frequency coefficient components, (4) carrying out
signal reconstruction through the new high-frequency coefficient components and a low-frequency coefficient component, obtained in the step (2), of the maximum decomposition scale to obtain a partial
discharge signal without
noise, and (5) outputting the partial
discharge signal without the
noise. The partial discharge signal denoising method based on the wavelet adaptive threshold achieves wavelet coefficient threshold self-
adaptation selection on the premise that any priori knowledge does not exist, and is applicable to various actual partial discharge conditions and good in effect of removing
white noise, and the denoised partial discharge signal with higher quality can be obtained.