Transformer partial discharge signal adaptive denoising method based on white shark optimization optimized successive variational mode decomposition

To effectively solve the problems of white noise and periodic narrowband interference in partial discharge detection, this paper proposes a partial discharge denoising method that combines the optimization of successive variational mode decomposition (SVMD) and wavelet threshold denoising. Compared to variational mode decomposition, SVMD does not require the pre-setting of the number of decomposition modes. However, it is affected by the balance parameter. To solve this problem, the white shark optimization algorithm is proposed to search for the optimal balance parameter. The kurtosis criterion is adopted for the decomposition modes to screen out the mode dominated by the amplifier signal, thereby eliminating the influence of periodic narrowband interference. The wavelet threshold denoising method is utilized to eliminate the residual small amount of white noise in the mode, and finally, the mode is reconstructed to obtain the denoising completion signal. Through analysis of the simulation and experimental signals, and by comparing with the complete ensemble empirical mode decomposition with adaptive noise combined wavelet threshold denoising method and the sym8 wavelet threshold denoising method. The results show that the denoising effect of the method proposed in this paper is better, and the characteristics of the partial discharge waveform are well-retained.