基于局部递归率分析的振动信号非平稳评价
|
摘要
针对土木工程结构环境振动信号非平稳程度评价的问题,提出一种新的基于递归量化分析的评价环境振动信号非平稳程度指标。首先,研究递归图中反映信号非平稳特征的模式,根据递归量化分析中已有的信号非平稳程度评价指标即递归趋势应用于土木工程结构环境振动信号时的不足,将质量损失函数理论引入递归量化分析,分析递归图中局部递归率的变化情况,并以最小局部递归率为基准,提出一个新的衡量信号非平稳程度的指标——递归损失量;然后,以几种典型的非平稳信号为例,分别计算2种非平稳评价指标并进行对比分析,讨论其差异及各自不同的适用条件;最后,将指标用于桥梁结构环境振动信号的非平稳程度评价。计算结果表明:提出的递归损失量比递归趋势更能适应环境振动信号的特点,能够对环境振动信号的非平稳程度进行更客观的评价。
According to non-stationary assessment for operational vibration signals,a novel index was proposed based on the recurrence quantification analysis.The non-stationarity patterns in the recurrence plots were studied.To fill up the deficiency of recurrence trend,which was the commonly used non-stationarity measurement in the recurrence plot and was not suitable for assessing non-stationarity of structural operational vibration signals,the variation of local recurrence density was discussed,the minimal local recurrence density was used as a reference,and then a new non-stationarity measurement named recurrence loss value was proposed in which the quality loss function theory was introduced.Some typical non-stationary signals were used to validate it.Comparison of recurrence trend and recurrence loss value was made,and the result illustrates their differences and similarities.For reinforcing proposed measurement,the measurements were used for operational vibration signals.The results show that recurrence loss value has a good applicability for operational vibration signals.
According to non-stationary assessment for operational vibration signals,a novel index was proposed based on the recurrence quantification analysis.The non-stationarity patterns in the recurrence plots were studied.To fill up the deficiency of recurrence trend,which was the commonly used non-stationarity measurement in the recurrence plot and was not suitable for assessing non-stationarity of structural operational vibration signals,the variation of local recurrence density was discussed,the minimal local recurrence density was used as a reference,and then a new non-stationarity measurement named recurrence loss value was proposed in which the quality loss function theory was introduced.Some typical non-stationary signals were used to validate it.Comparison of recurrence trend and recurrence loss value was made,and the result illustrates their differences and similarities.For reinforcing proposed measurement,the measurements were used for operational vibration signals.The results show that recurrence loss value has a good applicability for operational vibration signals.
引文
[1]Doebling S W,Charles R,Farrar M B,et al.Damageidentification and health monitoring of structural and mechanicalsystem s from changes in their vibration characteristics:aliterature review[R].Los Alamos:Los Alamos NationalLaboratory Report,1996,1-153.
[2]REN Weixin,SUN Zengshou.Structural damage identificationby using wavelet entropy[J].Engineering Structures,2008,30(10):2840-2849.
[3]Phillips P C,Perron P.Testing for a unit root in time seriesregression[J].Biometrika,1988,75(2):335-346.
[4]Bendat J S,Pierso A G.Random data:analysis and measurementprocedure[M].New York:John Wiley&Sons,2000:135-247.
[5]PAN Qin,Grimmelsman K A,Prader J,et al.Evaluation of thequality of ambient vibration monitoring data from the HenryHudson Bridge[C]//Aaron A.Nonintrusive Inspection,StructuresMonitoring,and Smart Systems for Homeland Security.Philadelphia:SPIE,2006:6178-6180.
[6]陈隽,徐幼麟,李杰.Hilbert-Huang变换在密频结构阻尼识别中的应用[J].地震工程与工程振动,2003,23(4):34-42.CHEN Jun,XU Youlin,LI Jie.Hilbert-Huang transform fordamping ratio identification of structures with closely spacedmodes of vibration[J].Chinese Journal of EarthquakeEngineering and Engineering Vibration,2003,23(4):34-42.
[7]Eckman J P,Oliffson K S,Ruelle D.Recurrence plots ofdynamical systems[J].Europhys Lett,1987,4(9):973-977.
[8]March T K,Chapman S C,Dendy R O.Recurrence plot statisticsand the effect of embedding[J].Physica D,2005,200(1/2):171-184.
[9]Marwan N,Romano W C,Thiel M,et al.Recurrence plot for theanalysis of complex systems[J].Physics Reports,2007,438(5/6):237-329.
[10]Zbilut J P,Webber J,Charles L.Embeddings and delays asderived from quantification of recurrence plots[J].PhysicsLetters A,1992,171(3/4):199-203.
[11]Marwan N,Meinke A.Extended recurrence plot analysis and itsapplication to ERP data[J].International Journal of Bifurcationand Chaos,2004,14(2):761-771.
[12]Taguchi G.Introduce to off-line quality control[M].Nagoya:Central Japan Quality Control Association,1979:18-90.
[13]Marwan N.Encounters with neighbours-current developments ofconcepts based on recurrence plots and their applications[D].Potsdam:University of Potsdam.Institute of Physics,2003:22-34.
[14]Marwan N,Wessel N,Meyerfeldt U,et al.Recurrence-plot-based measures of complexity and their application toheart-rate-variability data[J].Physical Review E,2002,66(2):026702.1-026702.8.
[15]GAO Jianbo.Detection nonstationary and state transitions in atime series[J].Physical Review E,2001,63(2):066202.1-066202.8.
[16]YAN Runqiang,ZU Yiqing,ZHU Yishen,Automatic speechsegmentation combining an Hmm-based approach andrecurrence trend analysis[C]//Castanie F.InternationalConference on Acoustics,Speech,and Signal Processing 2006,Touluse,France:IEEE.2006:797-800.
[17]Artiles-leon N.A pragmatic approach to multi-responseproblems using loss function[J].Quality Engineering,1996,9(2):213-248.
[2]REN Weixin,SUN Zengshou.Structural damage identificationby using wavelet entropy[J].Engineering Structures,2008,30(10):2840-2849.
[3]Phillips P C,Perron P.Testing for a unit root in time seriesregression[J].Biometrika,1988,75(2):335-346.
[4]Bendat J S,Pierso A G.Random data:analysis and measurementprocedure[M].New York:John Wiley&Sons,2000:135-247.
[5]PAN Qin,Grimmelsman K A,Prader J,et al.Evaluation of thequality of ambient vibration monitoring data from the HenryHudson Bridge[C]//Aaron A.Nonintrusive Inspection,StructuresMonitoring,and Smart Systems for Homeland Security.Philadelphia:SPIE,2006:6178-6180.
[6]陈隽,徐幼麟,李杰.Hilbert-Huang变换在密频结构阻尼识别中的应用[J].地震工程与工程振动,2003,23(4):34-42.CHEN Jun,XU Youlin,LI Jie.Hilbert-Huang transform fordamping ratio identification of structures with closely spacedmodes of vibration[J].Chinese Journal of EarthquakeEngineering and Engineering Vibration,2003,23(4):34-42.
[7]Eckman J P,Oliffson K S,Ruelle D.Recurrence plots ofdynamical systems[J].Europhys Lett,1987,4(9):973-977.
[8]March T K,Chapman S C,Dendy R O.Recurrence plot statisticsand the effect of embedding[J].Physica D,2005,200(1/2):171-184.
[9]Marwan N,Romano W C,Thiel M,et al.Recurrence plot for theanalysis of complex systems[J].Physics Reports,2007,438(5/6):237-329.
[10]Zbilut J P,Webber J,Charles L.Embeddings and delays asderived from quantification of recurrence plots[J].PhysicsLetters A,1992,171(3/4):199-203.
[11]Marwan N,Meinke A.Extended recurrence plot analysis and itsapplication to ERP data[J].International Journal of Bifurcationand Chaos,2004,14(2):761-771.
[12]Taguchi G.Introduce to off-line quality control[M].Nagoya:Central Japan Quality Control Association,1979:18-90.
[13]Marwan N.Encounters with neighbours-current developments ofconcepts based on recurrence plots and their applications[D].Potsdam:University of Potsdam.Institute of Physics,2003:22-34.
[14]Marwan N,Wessel N,Meyerfeldt U,et al.Recurrence-plot-based measures of complexity and their application toheart-rate-variability data[J].Physical Review E,2002,66(2):026702.1-026702.8.
[15]GAO Jianbo.Detection nonstationary and state transitions in atime series[J].Physical Review E,2001,63(2):066202.1-066202.8.
[16]YAN Runqiang,ZU Yiqing,ZHU Yishen,Automatic speechsegmentation combining an Hmm-based approach andrecurrence trend analysis[C]//Castanie F.InternationalConference on Acoustics,Speech,and Signal Processing 2006,Touluse,France:IEEE.2006:797-800.
[17]Artiles-leon N.A pragmatic approach to multi-responseproblems using loss function[J].Quality Engineering,1996,9(2):213-248.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心