Research on partial discharge signal denoising of transformer based on improved CEEMD and adaptive wavelet threshold

Because of the interference of white noise, the traditional denoising method is not ideal for partial discharge signal of transformer. To solve this problem, a fusion denoising method combining sample entropy with improved Complementary Ensemble Empirical Mode Decomposition (CEEMD) and Adaptive Wavelet Threshold is proposed. In this method, the original noisy signal is decomposed by CEEMD, and the modal of the Intrinsic Mode Fiction (IMF) is determined by introducing sample entropy, and the IMF component of noise is suppressed by using adaptive wavelet threshold denoising. The simulation results show that the fusion denoising method can effectively suppress the noise in partial discharge signal, and the denoising signal with high signal-to-noise ratio and low root mean square error can be obtained.