基于位置和程度指标的结构损伤识别研究
|
摘要
为提高结构损伤识别的精度,提出基于动、静态数据融合的位置指标和完全基于频率的位置指标,并采用学习速率自适应调整的新型BP神经网络学习算法,其特点是在网络迭代过程中根据网络学习误差来调整学习速率的取值,该方法有效地克服了传统BP网络学习过程中容易陷入局部极小和收敛速度慢、学习效率不高的缺点,进一步讨论了参数输入方式对网络识别效果的影响,分别采用两步诊断法和一步诊断法进行损伤识别.结果显示,两步诊断法对损伤位置和程度的识别正确率较高,而一步诊断法识别效果却不令人十分满意;减少位置指标和程度指标的输入个数对损伤识别结果有显著的影响.
To increase the precision of structural damage identification,two kinds of new location indices based on the fusion of the static,dynamic data and natural frequency respectively were proposed.In view of the disadvantages of traditional BP neural networks,a self-adaptive algorithm which could adjust the value of learning rate according to output errors during the training process was introduced.The updated algorithm was proved to avoid oscillation phenomenon and improve learning rate effectively.The influence of input manner on identification effect was also discussed.Two-step-method and one-step-method using self-adaptive networks were adopted for damage detection respectively.It turned out that two-step-method made more accurate estimation for damage location and extent than one-step-method did.The number of location index and extent index influenced the identification effect greatly.
To increase the precision of structural damage identification,two kinds of new location indices based on the fusion of the static,dynamic data and natural frequency respectively were proposed.In view of the disadvantages of traditional BP neural networks,a self-adaptive algorithm which could adjust the value of learning rate according to output errors during the training process was introduced.The updated algorithm was proved to avoid oscillation phenomenon and improve learning rate effectively.The influence of input manner on identification effect was also discussed.Two-step-method and one-step-method using self-adaptive networks were adopted for damage detection respectively.It turned out that two-step-method made more accurate estimation for damage location and extent than one-step-method did.The number of location index and extent index influenced the identification effect greatly.
引文
[1]谢峻,韩大建.一种改进的基于频率测量的结构损伤识别方法[J].工程力学,2004,21(1):21-25.XIE Jun,HAN Da-jian.An improved method for struc-ture damage detection based on frequency measurement[J].Engineering Mechanics,2004,21(1):21 25.
[2]PANDEY P C,BARAI S V.Multilayer perceptron indamage detection of bridge structures[J].Computersand Structures,1995,94:597-608.
[3]WANG X,HU N,FUKUNAG A H,et al.Structuredamage identification using static test data and changesin frequencies[J].Engineering Structures,2001,23:610621.
[4]LEE U,SHIN J.A frequency response function-basedstructural damage identification method[J].Computersand Structures,2002,80:117 132.
[5]KIM Jeong-Tae,RYU Yeon-Sun,CHO Hyun-Man,etal.Damage identification in beam-typestructures:fre-quency-based method vs mode-shape-based method[J].Engineering Structures,2003,25:57 67.
[6]SHI Z Y,LAW S S.Structural damage localization frommodal strain energy change[J].Journal of Sound andVibration,1998,218(5):825 844.
[7]王柏生,倪一清,高赞明.框架结构连接损伤识别神经网络输入参数的确定[J].振动工程学报,2000,13(1):137142.WANG Bo-sheng,NI Yi-qing,KAO Zhan-ming.Inputparameters for artificial neural networks in frame con-nect ion damage identification[J].Journal of VibrationEngineering,2000,13(1):137 142.
[8]瞿伟廉,黄东梅.高耸塔架结构节点损伤基于神经网络的两步诊断法[J].地震工程与工程震动,2003,23(2):143 149.QU Wei-lian,HUANG Dong-mei.Two-step detection a-pproach to joint damage in high-rise tower structures byusing artificial neural network[J].Earthquake Engineer-ing and Engineering Vibration,2003,23(2):143 149.
[9]YUN Chung-Bang,YI Jin-Hak,BAHNG Eun-Young.Joint damage assessment of framed structures using a n-eural networks technique[J].Engineering Structures,2001,23:425 435.
[10]易伟建,刘霞.基于遗传算法的结构损伤诊断研究[J].工程力学,2001,18(2):64 71.YI Wei-jian,LIU Xia.Damage diagnose of structuresby genetic algorithms[J].Engineering Mechanics,2001,18(2):64 71.
[11]CAO C Y,HUNG Shin-Lin.Detection of structuraldamage via free vibration responses generated by appr-oximating artificial neural networks[J].Computersand Structures,2003,81:2631 2644.
[12]金虎,楼文娟,陈勇.基于自适应BP神经网络的结构荷载识别[J].浙江大学学报:工学版,2005,39(10):1596 1602.JIN Hu,LOU Wen-juan,CHEN Yong.Load identifica-tion for bridge structures based on self-adaptive BPneural networks[J].Journals of Zhejiang University:Engineering Science,2005,39(10):1596 1602.
[2]PANDEY P C,BARAI S V.Multilayer perceptron indamage detection of bridge structures[J].Computersand Structures,1995,94:597-608.
[3]WANG X,HU N,FUKUNAG A H,et al.Structuredamage identification using static test data and changesin frequencies[J].Engineering Structures,2001,23:610621.
[4]LEE U,SHIN J.A frequency response function-basedstructural damage identification method[J].Computersand Structures,2002,80:117 132.
[5]KIM Jeong-Tae,RYU Yeon-Sun,CHO Hyun-Man,etal.Damage identification in beam-typestructures:fre-quency-based method vs mode-shape-based method[J].Engineering Structures,2003,25:57 67.
[6]SHI Z Y,LAW S S.Structural damage localization frommodal strain energy change[J].Journal of Sound andVibration,1998,218(5):825 844.
[7]王柏生,倪一清,高赞明.框架结构连接损伤识别神经网络输入参数的确定[J].振动工程学报,2000,13(1):137142.WANG Bo-sheng,NI Yi-qing,KAO Zhan-ming.Inputparameters for artificial neural networks in frame con-nect ion damage identification[J].Journal of VibrationEngineering,2000,13(1):137 142.
[8]瞿伟廉,黄东梅.高耸塔架结构节点损伤基于神经网络的两步诊断法[J].地震工程与工程震动,2003,23(2):143 149.QU Wei-lian,HUANG Dong-mei.Two-step detection a-pproach to joint damage in high-rise tower structures byusing artificial neural network[J].Earthquake Engineer-ing and Engineering Vibration,2003,23(2):143 149.
[9]YUN Chung-Bang,YI Jin-Hak,BAHNG Eun-Young.Joint damage assessment of framed structures using a n-eural networks technique[J].Engineering Structures,2001,23:425 435.
[10]易伟建,刘霞.基于遗传算法的结构损伤诊断研究[J].工程力学,2001,18(2):64 71.YI Wei-jian,LIU Xia.Damage diagnose of structuresby genetic algorithms[J].Engineering Mechanics,2001,18(2):64 71.
[11]CAO C Y,HUNG Shin-Lin.Detection of structuraldamage via free vibration responses generated by appr-oximating artificial neural networks[J].Computersand Structures,2003,81:2631 2644.
[12]金虎,楼文娟,陈勇.基于自适应BP神经网络的结构荷载识别[J].浙江大学学报:工学版,2005,39(10):1596 1602.JIN Hu,LOU Wen-juan,CHEN Yong.Load identifica-tion for bridge structures based on self-adaptive BPneural networks[J].Journals of Zhejiang University:Engineering Science,2005,39(10):1596 1602.
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