Research On Denoising Method Of Partial Discharge Signal Of Cable Terminal

Partial discharge detection is one of the most effective methods of fault diagnosis of electrical equipment,and is widely used in cable accessory fault diagnosis.However,due to the presence of background noise,especially the presence of white noise,the sensitivity of the partial discharge on-line detection is greatly reduced.The frequency components of white noise are distributed throughout the frequency domain,and the time domain waveform is also spread over the whole time domain,which makes the white noise interference in the PD signal become a difficult and hot topic in the technology of PD signal denoising.Based on the above discussion,two kinds of PD signal denoising methods are proposed to remove the white noise interference under high SNR and low SNR,respectively,based on the partial discharge detection of cable termination.The main research contents are as follows:(1)The main causes of the failure of the cable terminal are analyzed.Four typical typical defect models of the cable terminal are designed as the test object to carry out the partial discharge test.The partial discharge signal of the typical defects of the four kinds of cable terminals is taken as the noise Method of test validation.(2)A method of denoising signal based on kurtosis and time domain energy is proposed to solve the problem that the ablation rate of the existing PD signal denoising method is low and the edge of the pulse is not clear after denoising.Firstly,the localization of the PD pulse is achieved by calculating the kerp of the noise signal,and then the position of the PD peak in the whole data window is determined to reduce the amount of subsequent pulse extraction.Secondly,by analyzing the white noise and the localized pulse signal,the time domain energy threshold and the corresponding window length are determined.Finally,the time-domain energy search method is used to detect and determine the local discharge pulse edge of the edge of the search pulse with the pulse peak as the center,and then the partial discharge Pulse extraction.(3)An adaptive sparse decomposition denoising method is proposed.At the same time,a matching atomic library corresponding to the PD signal is constructed to improve the calculation of the MP algorithm in the process of de-noising Efficiency and denoising effect.Based on the fast spectrum kurtosis and S transform of the signal,the time-frequency characteristic of the PD signal is obtained quickly,including the approximate position of the center frequency,the bandwidth,the starting point of the PD release and the extinguishing point;the partial frequency characteristic of the PD signal is used to optimize the PD signal The time-frequency parameters of the atomic pool are matched,and the matching of the atoms is selected by the selection of a small number of atoms in the MP calculation.The best matching atoms obtained by each iteration are sparse to the original Put the signal to the noise.The simulation and experimental results show that the two methods of denoising are effective to solve the problem of denoising in the process of de-noising of existing PD signal denoising.The calculation rate is slow and the denoising is not complete,Signal distortion after denoising.