基于独立分量和扩展通道的振动信号降噪方法

英文篇名：Noise Reduction of Beam Vibration Signals Based on Independent Component Analysis and Expanding Channels
中文刊名：武汉理工大学学报
英文刊名：Journal of Wuhan University of Technology
作者：赵丽芝 ; 袁海庆 ; 杨燕
英文作者：ZHAO Li-zhi ; YUAN Hai-qing ; YANG Yan (School of Civil Engineering and Architecture ; Wuhan University of Technology ; Wuhan 430070 ; China)
中文关键词：独立分量分析 ; 同测点多通道 ; FastICA ; 降噪
英文关键词：independent component analysis ; the same position and many tunnels ; FastICA ; noise reduction
出版日期：2008-06-25
机构：武汉理工大学土木工程与建筑学院,武汉理工大学土木工程与建筑学院,武汉理工大学土木工程与建筑学院武汉430070,武汉430070,武汉430070
年：2008
期：06
出版单位：武汉理工大学学报

摘要

针对ICA降噪技术中引入噪声通道的局限性,采取以扩展通道替代虚拟噪声通道的方法,将ICA的思想引入到结构同测点多通道振动信号处理中。叙述了ICA原理,FastICA算法的实现及其运用于振动信号降噪的思路,通过对实验梁同测点两通道振动信号的处理,并引入自相关函数和FFT谱衡量降噪效果,结果表明,采用扩展通道运用ICA进行结构振动信号降噪的可行性。
In view of the limitation to noisy channels based on the ICA,hypothesized noise channels were replaced by expanding channels in this paper.The ICA method was introduced to de-noise the same position and many channels structural vibration signals.The principles of ICA and FastICA algorithm were elaborated and the same position and two channels vibration signals of beam were de-noised by ICA.This method was proved effectually by the beam vibration signals in lab.Autocorrelation function and FFT spectrum were adopted to weigh the noise reduction effect.It was showed that it was feasible to adopt expanding channels and ICA to reduce noise of structural vibration signals.

引文

[1]王济.MATLAB在振动信号处理中的应用[M].北京:中国水利水电出版社,2006.
    [2]赵浩.独立分量分析在生物医学信号处理中的应用[J].生物医学工程研究,2004,22(4):56-60.
    [3]Lee Teiwon.Independent Component Analysis:Theory and Application[M].Boston:Kluwer Academic Publishers,1998.
    [4]季忠,金涛.虚拟噪声通道在基于ICA消噪过程中的应用[J].中国机械工程,2005,16(4):350-364.
    [5]杨福生,洪波.独立分量分析的原理与应用[M].北京:清华大学出版社,2006.
    [6]吕文彪,尹成.利用独立分量分析法去除地震噪声[J].石油地球物理勘探,2007,42(2):132-136.
    [7]胡广书.数字信号处理理论、算法与实现[M].北京:清华大学出版社,2006.