摘要
由Mallat的极大值去噪算法出发,与利用尺度间小波变换相似系数的阈值去噪方法结合,提出了一种新的改进去噪算法。该方法克服了模极大值法在信号重构方面计算复杂,重构信号产生偏差的缺陷。通过对局放信号的去噪仿真计算,证明了该方法运算简单,适合局放信号的处理,处理后的局放信号不失真,剔除噪声效果好。
This paper introduces a new denosing algorithm based on Mallat's local maximum denosing algorithm. It combines the threshold method using similarity coefficients of the wavelet transform with scales and modular maximum algorithm, and it overcomes the disadvantage of calculation complexity and reconstruction deviation of revaluate signal in local maximum denosing algorithm. The simulation shows that the denosing method is very effective.
出处
《电力系统及其自动化学报》
CSCD
北大核心
2005年第2期9-12,共4页
Proceedings of the CSU-EPSA
关键词
小波变换
模极大值
阈值法
去噪
局部放电
wavelet transform
local maximum
thresholding
denosing
partial discharge
