FFT与小波分析应用于抑制局部放电窄带干扰  

Suppressing Periodic Narrowband Noise in Partial Discharge(PD) Signal Using FFT and Wavelet Analysis

作　　者：李天云[1] 程思勇[1] 杨梅[1] 王爱凤[1] 

机构地区：[1]东北电力大学电气工程学院,吉林132002

出　　处：《高电压技术》2007年第12期71-74,123,共5页High Voltage Engineering

摘　　要：局部放电极易被噪声污染,众多干扰中载波通讯、高频保护、无线电广播等周期干扰尤其严重,而窄带干扰严重时可能完全淹没局部放电信号,这给局部放电的检测工作带来了巨大困难。为了从强的窄带干扰背景中有效提取局部放电信号,在基于FFT与小波分析理论的基础上,根据窄带干扰与局部放电信号具有不同频谱的特性,提出了抑制局部放电信号窄带干扰的新方法。考虑FFT处理方法的频谱泄漏性,再用小波变换对FFT处理后的局部放电信号进行进一步消噪,由此能够取得很好的消噪效果。仿真结果表明,此方法对窄带干扰去噪要比FFT、小波及小波模极大值去噪效果好,且得到的局部放电信号失真小。It is difficult to measure the partial discharge（PD） signals because the signals may be submerged or flooded when the surrounding noises are severe.In order to extract PD signals efficiently from the violent narrowband noise,a new method is put forward to suppress periodic narrowband noise in PD signals based on the FFT and wavelet theory.Due to the distinct spectrum feature of PD signals and the periodic narrowband noise,the spectra of PD signals riding in periodic narrowband noise are obtained by FFT,it is a threshold to suppress periodic narrowband noise.However,FFT has leakage of spectrum,above processes cannot achieve the ideal effect.Then the PD signals that are disposed by FFT are de-noised by wavelet analysis,it can achieve much better results. Simulation proves that the method proposed in this paper can effectively dispose the periodic narrowband noise,and the de-noising effect is superior to FFT,wavelet and the wavelet module maximum method.Moreover,the distortion of PD signals is very small.

关 键 词：FFT 小波分析 局部放电 窄带干扰 消噪 阈值处理 

分 类 号：TM85[电气工程—高电压与绝缘技术]