振动与冲击
振動與遲擊
진동여충격
JOURNAL OF VIBRATION AND SHOCK
2014年
10期
162-166
,共5页
王怀光%张培林%吴定海%李兵%范红波
王懷光%張培林%吳定海%李兵%範紅波
왕부광%장배림%오정해%리병%범홍파
形态滤波器%提升小波%数据压缩%机械振动%状态监测
形態濾波器%提升小波%數據壓縮%機械振動%狀態鑑測
형태려파기%제승소파%수거압축%궤계진동%상태감측
morphological filter%lifting wavelet transformation%data compression%mechanical vibration%condition mo-nitoring
针对大型复杂机电设备状态分布式监测面临的海量数据传输问题,对比研究基于提升小波与形态提升小波的振动信号数据压缩方法。对实测浮点型振动信号,利用小波稀疏分解特性提出基于提升小波变换的机械振动信号数据压缩方法,通过变换后小波系数优化组合及改进编码方法,可提高阈值处理数据压缩效果。针对实测数据含大量冗余信息缺点,结合状态监测目的提出基于形态提升小波的数据压缩方法,利用形态学滤波器非线性分析特性,在振动信号网络监测数据压缩传输过程中实现信号预处理,预制噪声干扰,较好保留有用信息。所提最大区分度准优化选择分解层数,通过对比分析两种数据压缩方法表明,形态提升小波方法具有计算量小、分析速度快、压缩比高等优点。
針對大型複雜機電設備狀態分佈式鑑測麵臨的海量數據傳輸問題,對比研究基于提升小波與形態提升小波的振動信號數據壓縮方法。對實測浮點型振動信號,利用小波稀疏分解特性提齣基于提升小波變換的機械振動信號數據壓縮方法,通過變換後小波繫數優化組閤及改進編碼方法,可提高閾值處理數據壓縮效果。針對實測數據含大量冗餘信息缺點,結閤狀態鑑測目的提齣基于形態提升小波的數據壓縮方法,利用形態學濾波器非線性分析特性,在振動信號網絡鑑測數據壓縮傳輸過程中實現信號預處理,預製譟聲榦擾,較好保留有用信息。所提最大區分度準優化選擇分解層數,通過對比分析兩種數據壓縮方法錶明,形態提升小波方法具有計算量小、分析速度快、壓縮比高等優點。
침대대형복잡궤전설비상태분포식감측면림적해량수거전수문제,대비연구기우제승소파여형태제승소파적진동신호수거압축방법。대실측부점형진동신호,이용소파희소분해특성제출기우제승소파변환적궤계진동신호수거압축방법,통과변환후소파계수우화조합급개진편마방법,가제고역치처리수거압축효과。침대실측수거함대량용여신식결점,결합상태감측목적제출기우형태제승소파적수거압축방법,이용형태학려파기비선성분석특성,재진동신호망락감측수거압축전수과정중실현신호예처리,예제조성간우,교호보류유용신식。소제최대구분도준우화선택분해층수,통과대비분석량충수거압축방법표명,형태제승소파방법구유계산량소、분석속도쾌、압축비고등우점。
In order to solve the problem of mass data transmission faced by the distributed condition monitoring of large-scale electromechanical equipments,a method for data compression of vibration signals based on lifting wavelet transformation and that based on morphological lifting wavelet transformation were studied contrastively.For actual floating-point vibration signals,a method for data compression of mechanical vibration signals based on lifting wavelet transformation was put forward by taking the advantage of wavelet sparse decomposition characteristic.The effects of data compression with threshold processing were improved greatly through optimal combination of the transformed wavelet coefficients and code improvement.To overcome the shortcoming of the measured data containing much redundant information,a method for data compression based on morphological lifting wavelet transformation was proposed for condition monitoring.The pretreatment of network monitoring data of vibration signals was realized in the process of data compression transmission by adopting the non-linear analysis characteristic of a morphological filter,then the noise was reduced and the useful information of signals was reserved.A maximum differentiating capacity criterion was proposed for selecting the number of decomposition.By comparing these two methods for data compression,it was shown that the morphological lifting wavelet transformation has the advantages of simple calculation,fast analysis speed and high compression ratio.