基于EMD的局部放电去噪方法的研究  

Study on De-noising Method of Partial Discharge based EMD

作　　者：孙博[1,2] 张建文[1] 潘磊落[1] 

机构地区：[1]中国矿业大学信息与电气工程学院,江苏徐州221008 [2]广东省建科建筑设计院,广州510500

出　　处：《绝缘材料》2014年第3期89-93,共5页Insulating Materials

摘　　要：针对局部放电信号在线检测中通常存在噪声污染、小波变换受小波基的选择影响较大,能量泄露及脉冲信号畸变等问题,在EMD的基础上,采用基于自相关函数的改进时空去噪和阈值去噪相结合的EMD去噪方法,在MATLAB的环境下进行仿真验证,并对电晕放电信号进行去噪分析。结果表明:采用该方法去噪性能在信噪比和波形畸变率两方面都较db4和db8小波去噪更好,有良好的去噪效果。In view of the problems of noise pollution, wavelet transform is affected by the selection of base wavelet, leakage of energy and pulse signal distortion during online detection of partial discharge signal, based on Empirical Mode Decomposition（EMD）, a de-noising method was proposed by combination of the characteristics of autocorrelation function and threshold, and the method was validated in the simulation environment of MATLAB, and the corona discharge signal was de-noising analyzed. The results prove that the de-noising effect of the de-noising method based on EMD is better than that of the wavelet de-noising method.

关 键 词：EMD分解 局部放电 白噪声 自相关函数 

分 类 号：TM835[电气工程—高电压与绝缘技术]