New Methods for Detecting Partial Discharges Using Adaptive Fast Noise Reduction

Partial discharge (PD) signals of switchgear are interspersed with various noises such as white noise, narrowband interference, etc. PD signals with small amplitude may be submerged in the noise, which has an impact on the identification of PDs. In order to solve the problems of poor noise reduction effect, slow noise reduction speed and poor adaptivity of PD signal, the article proposes a noise reduction method for PD signal based on Improved Snow Ablation Algorithm Optimized Adaptive Modal Decom-position (ISAO-ICEEMDAN). The algorithm introduces chaotic mapping and Gaussian variation to improve the snow ablation algorithm (SAO); then, the improved snow ablation algorithm (ISAO) is used to adaptively select the parameters of ICEEMDAN and re-construct the appropriate modal components according to the principle of minimum envelope entropy, and then the noise-reduced PD signals are obtained, which realizes the fast and effective adaptive noise reduction of the PD signals. Finally, the article carries out experimental verification of the noise reduction effect on the simulated PD signal and the measured PD signal, respectively. The experimental results show that ISAO-ICEEMDAN has obvious superiority in noise reduction effect, noise reduction speed and adaptivity.