软件
軟件
연건
SOFT WARE
2014年
7期
13-17,22
,共6页
小波阈值去噪%EEMD%端点效应%ICA
小波閾值去譟%EEMD%耑點效應%ICA
소파역치거조%EEMD%단점효응%ICA
EMD%EEMD%end effect%ICA
针对非线性非平稳信号的去噪问题,提出了一种基于独立分量分析(Decomposition Components Analysis,简称ICA)算法的集合经验模态分解去噪方法。首先利用白噪声辅助数据分析方法——集合经验模态分解(Ensemble Empirical Mode Decomposition,简称EEMD)有效的抑制了经验模态分解(Empirical Mode Decomposition,简称EMD)中存在的端点效应和模态混叠现象,然后利用ICA算法对含噪信号经过EEMD分解后的有限个固有模态函数(Intrinsic Mode Function,简称IMF)进行去噪处理,有效的分离出若干个有效的语音信号分量,并对其进行语音重构,最后与小波阈值去噪方法进行比较,通过仿真可以看出,该方法对于信号去噪较为理想。
針對非線性非平穩信號的去譟問題,提齣瞭一種基于獨立分量分析(Decomposition Components Analysis,簡稱ICA)算法的集閤經驗模態分解去譟方法。首先利用白譟聲輔助數據分析方法——集閤經驗模態分解(Ensemble Empirical Mode Decomposition,簡稱EEMD)有效的抑製瞭經驗模態分解(Empirical Mode Decomposition,簡稱EMD)中存在的耑點效應和模態混疊現象,然後利用ICA算法對含譟信號經過EEMD分解後的有限箇固有模態函數(Intrinsic Mode Function,簡稱IMF)進行去譟處理,有效的分離齣若榦箇有效的語音信號分量,併對其進行語音重構,最後與小波閾值去譟方法進行比較,通過倣真可以看齣,該方法對于信號去譟較為理想。
침대비선성비평은신호적거조문제,제출료일충기우독립분량분석(Decomposition Components Analysis,간칭ICA)산법적집합경험모태분해거조방법。수선이용백조성보조수거분석방법——집합경험모태분해(Ensemble Empirical Mode Decomposition,간칭EEMD)유효적억제료경험모태분해(Empirical Mode Decomposition,간칭EMD)중존재적단점효응화모태혼첩현상,연후이용ICA산법대함조신호경과EEMD분해후적유한개고유모태함수(Intrinsic Mode Function,간칭IMF)진행거조처리,유효적분리출약간개유효적어음신호분량,병대기진행어음중구,최후여소파역치거조방법진행비교,통과방진가이간출,해방법대우신호거조교위이상。
In order to focus on the denoise of Nonlinear nonstationary signals, we first use the white noise assisted data analysis called EEMD (ensemble empirical mode decomposition) based on the ICA (decomposition Components Analysis) to suppress the end effect and effect and aliasing which appear in the empirical mode decomposition in the process of empirical mode decomposition(EMD). And then, we use ICA to denoise the noise signals which have been decomposed by EEMD into IMF, and we can efficiently decompose some effective speech signals components and reconstruct them. Finally, compared with the waved threshold method, the effect of this solution is ideal.