机械与电子
機械與電子
궤계여전자
MACHINERY & ELECTRONICS
2014年
10期
12-16
,共5页
崔心瀚%马立元%魏忠林%王天辉
崔心瀚%馬立元%魏忠林%王天輝
최심한%마립원%위충림%왕천휘
阈值去噪%经验模态分解%希尔伯特%黄变换%特征提取
閾值去譟%經驗模態分解%希爾伯特%黃變換%特徵提取
역치거조%경험모태분해%희이백특%황변환%특정제취
threshold de noising%EMD%HHT%feature extraction
为了在信号瞬时特征提取过程中有效降低噪声干扰影响,提出一种基于小波阈值降噪和经验模态分解(EMD)的信号瞬时特征提取方法。根据信号特征选择适合的小波阈值函数进行降噪处理,然后对降噪信号进行 EMD 分解,以互相关系数作为判别依据,保留含有信号瞬时特征的本征模函数(IMF),并进行 Hilbert 时频谱图和边际谱图分析,最终完成信号瞬时特征的提取。
為瞭在信號瞬時特徵提取過程中有效降低譟聲榦擾影響,提齣一種基于小波閾值降譟和經驗模態分解(EMD)的信號瞬時特徵提取方法。根據信號特徵選擇適閤的小波閾值函數進行降譟處理,然後對降譟信號進行 EMD 分解,以互相關繫數作為判彆依據,保留含有信號瞬時特徵的本徵模函數(IMF),併進行 Hilbert 時頻譜圖和邊際譜圖分析,最終完成信號瞬時特徵的提取。
위료재신호순시특정제취과정중유효강저조성간우영향,제출일충기우소파역치강조화경험모태분해(EMD)적신호순시특정제취방법。근거신호특정선택괄합적소파역치함수진행강조처리,연후대강조신호진행 EMD 분해,이호상관계수작위판별의거,보류함유신호순시특정적본정모함수(IMF),병진행 Hilbert 시빈보도화변제보도분석,최종완성신호순시특정적제취。
In order to reduce the impact caused by noise pollution upon the signal processing of instanta-neous feature extraction,a method combining wavelet threshold de noising and empirical mode decomposi-tion (EMD)to extract the instantaneous features of signals is proposed here.The method selects a suitable wavelet threshold which depends on the feature of signal to make pretreatment to the signal,and then the intrinsic mode functions (IMFs)of de noising signals are obtained with EMD.Basing on the cor-relation degrees,the real IMFs are extracted to an-alyze the Hilbert spectrum and the marginal spec-trum.Finally we complete the extraction of signals instantaneous feature.