基于AVMD-自适应小波包法的电缆局部放电去噪研究  

Research on cable partial discharge denoising based on an AVMD-adaptive wavelet packet method

作　　者：吴童 孙抗[1] 师文文 WU Tong;SUN Kang;SHI Wenwen(College of Electrical Engineering and Automation,Henan Polytechnic University,Jiaozuo 454000,China;Chinalco Logistics Group Zhongzhou Co.,Ltd.,Jiaozuo 454000,China)

机构地区：[1]河南理工大学电气工程与自动化学院,河南焦作454000 [2]中铝物流集团中州有限公司,河南焦作454000

出　　处：《电力系统保护与控制》2020年第14期95-103,共9页Power System Protection and Control

基　　金：国家自然科学青年基金项目资助(61403130);河南省科技攻关项目资助(202102210092)。

摘　　要：为准确提取检测到的局部放电信号,针对高压电力电缆的噪声抑制问题,提出了自适应变分模态分解(AVMD)结合自适应小波包分解的方法提取纯净的局放信号。首先运用AVMD将周期性窄带干扰、白噪声和局放信号分解在不同的基本模态分量中,将周期性窄带干扰滤出,得到仅含有白噪声的局放信号。再运用自适应小波包分解,将信号分解在高中低频的分量中,根据阈值法将不含局放信号的分量滤出,得到较为纯净的局放信号,并将所提方法分别与其中单独一种算法进行去噪比较分析。仿真结果表明,所提方法抑制噪声效果更明显,与仿真信号的相似度最高。In order to extract the detected PD signal accurately and solve the problem of noise suppression of a high voltage power cable,AVMD combined with adaptive wavelet packet decomposition is proposed to extract a pure PD signal.First,the periodic narrow-band interference,white noise and PD signals are decomposed into different basic modal components by AVMD.The periodic narrowband interference is filtered out to obtain the partial discharge signal with white noise only.Then adaptive wavelet packet decomposition is used to decompose the signal in the high,middle and low frequency components.According to the threshold method,the components without a PD signal are filtered out to get a pure PD signal.The method is compared with one of the other methods.The simulation results show that the noise suppression effect of this method is more apparent,and the similarity with the simulation signal is the highest.

关 键 词：局部放电 AVMD 自适应小波包分解 周期性窄带干扰 白噪声 

分 类 号：TM855[电气工程—高电压与绝缘技术]