小波系数软硬阈值折中方法在故障定位消噪中的应用  

The Application of De-noise Based on Compromise Algorithm Between Soft-thresholding and Hard-thresholding of Wavelet Coefficient for Fault Location

作　　者：周静[1] 陈允平[1] 周策[2] 梁劲 

机构地区：[1]武汉大学电气工程学院,湖北省武汉市430072 [2]国家电网公司,北京市100031 [3]北京电力公司,北京市100031

出　　处：《电力系统自动化》2005年第1期65-68,共4页Automation of Electric Power Systems

摘　　要：强加在行波上的噪声造成了故障定位不可接受的误差,因此去除噪声是提高故障定位精度的关键因素。文中提出了一种软硬阈值折中消噪的方法,即在两种方法上施加不同的权重,可以有效地克服两种方法的缺点,与单纯用软阈值或硬阈值去噪相比,可以获得更高的信噪比增益。Noise superimposed on signal may cause unacceptable errors of fault location. For accurate fault location based on traveling wave, effective de-noise is very important and a key technique to ensure high accuracy of fault location. This paper proposes a de-noising approach to detect singularity of signal. That is a compromising algorithm between soft-thresholding and hard-thresholding of wavelet coefficient. Simulation results show that the de-noise method proposed in the paper is feasible and effective.

关 键 词：消噪 折中方法 小波变换 故障定位 

分 类 号：TM711[电气工程—电力系统及自动化]