摘要
由于气体绝缘组合电器(GIS)的局部放电(PD)特高频检测虽然能够避开大多数窄带干扰,但是,还有部分窄带干扰会进入检测系统,从而湮没PD信号,因此在现场检测时必须加以抑制。根据PD信号与周期性窄带干扰小波变换后的不同特点,剖析了抑制周期性窄带干扰的常用方法--小波系数直接置零法的局限性,提出了可以用于抑制周期性窄带干扰的有效小波系数(EWC)阈值法,并以所构造的db4复小波对仿真信号与实测GIS特高频PD信号进行了对比去噪研究。研究结果表明:基于EWC阈值法的复小波变换能够有效地抑制GIS特高频PD信号中的周期性窄带干扰。
As periodic narrowband interference is inevitable in the gas insulated switchgear (CIS) partial discharge (PD) ultra-high-frequency (UHF) detecting system, such interference should be suppressed during field detection. The magnitude of periodic narrowband interference is often greater than that of the UHF PD signal. According to the different wavelet characteristics of PD and periodic narrowband interference, the limitations of the commonly used method of direct zero-setting of the wavelet coefficient are analyzed and the effective wavelet coefficient (EWC) threshold method for suppressing the periodic narrowband interference is proposed. By referring to this threshold, the complex wavelet constructed is used to compare with the UHF PD signals in GIS and in noise elimination. The results show that the EWC threshold method of complex wavelet transform can effectively suppress the periodic narrowband interference from GIS UHF PD signals.
出处
《电力系统自动化》
EI
CSCD
北大核心
2005年第7期43-47,共5页
Automation of Electric Power Systems
基金
国家自然科学基金资助项目(50377045)重庆市攻关项目资助(7795)
