基于独立分量分析和支持向量机的结构损伤识别集成算法
|
摘要
针对模式识别最关键的两个环节:特征提取和分类器设计,提出了基于独立分量分析(ICA)和支持向量机(SVM)的损伤识别集成算法,首先应用ICA方法计算独立源信号和混合矩阵[A],利用混合矩阵与模态振型的对应关系,得到振型矩阵[Φ],将模态振型的变形矩阵[Φ]*作为特征参数输入至SVM分类器进行损伤识别,在冲击载荷作用下,对钢框架结构模型进行了振动试验,结果表明:ICA方法提取的模态振型是一种高效的损伤特征参数,基于ICA和SVM的集成算法能够成功识别结构损伤、损伤位置和损伤程度,从而为结构健康监测提供了一种行之有效的损伤识别方法。
In order to extract the sensitive feature and design the effective classifier,this paper presents a novel approach to detect structural damage based on combining independent component analysis(ICA)extraction of modal shape and support vector machine(SVM).A new modal shape[Φ]*from ICA technique,represented by reshaping the mixing matrix[A],is employed as extracted feature to build multiple SVMs.The methodology is applied to steel frame structure with measured time histories from 13 accelerometers.The experiment results show that the proposed method can succeed not only in detecting the structural damage but also in estimating its location and extent,consequently providing an effective technology for structural damage detection.
In order to extract the sensitive feature and design the effective classifier,this paper presents a novel approach to detect structural damage based on combining independent component analysis(ICA)extraction of modal shape and support vector machine(SVM).A new modal shape[Φ]*from ICA technique,represented by reshaping the mixing matrix[A],is employed as extracted feature to build multiple SVMs.The methodology is applied to steel frame structure with measured time histories from 13 accelerometers.The experiment results show that the proposed method can succeed not only in detecting the structural damage but also in estimating its location and extent,consequently providing an effective technology for structural damage detection.
引文
[1]HyvrinenA,Oja E.Independent component analysis:algorithms and applications[J.]Neural Networks.2000,13(4-5):411-430
[2]Choi S,Cichocki A,Park H Me,t al.Blind Source Separation and Independent Component Analysis:A Review[J].Neural Informa-tion Processing.20056,(1):1-57
[3]McNeill S I,Zimmerman D C.A framework for blind modal identification using joint approximate diagonalization[J.]MechanicalSystems and Signal Processing.20082,2(7):1526-1548
[4]Kerschen G,Poncelet F,Golinval J C.Physical interpretation of independent component analysis in structural dynamics[J].Me-chanical Systems and Signal Processing.20072,1(4):1561-1575
[5]张晓丹,姚谦峰,刘佩.基于快速独立分量分析的模态振型识别方法研究[J].振动与冲击2,0092,8(7):158-162
[6]静行,袁海庆,赵毅.基于独立分量分析的结构模态参数识别[J].振动与冲击2,0102,9(3):137-142
[7]Widodo A,Yang B S,Han Tian.Combination of independent component analysis and support vector machines for intelligent faultsdiagnosis of induction motors[J].Expert Systems with Applications.2007,32(2):299-312
[8]张育智,何伟,李乔.用于结构损伤识别的神经网络输入选取规则探究[J].四川建筑科学研究,20103,6(1):63-67
[9]闫桂荣,段忠东,欧进萍.基于结构振动信息的损伤识别研究综述[J].地震工程与工程振动,20072,7(3):95-104
[10]焦卫东,杨世锡,吴昭同.复合ICA-SVM机械状态模式分类[J.]中国机械工程,2004,15(1):62-65
[11]田景文,高美娟.人工神经网络算法研究及应用[M.]北京:北京理工大学出版社.2006.238-241
[12]Song H Z,Zhong L,Han B.Structural damage detection by integrating independent component analysis and support vector machine[J].International Journal of Systems Science.2006,37(13):961-967
[2]Choi S,Cichocki A,Park H Me,t al.Blind Source Separation and Independent Component Analysis:A Review[J].Neural Informa-tion Processing.20056,(1):1-57
[3]McNeill S I,Zimmerman D C.A framework for blind modal identification using joint approximate diagonalization[J.]MechanicalSystems and Signal Processing.20082,2(7):1526-1548
[4]Kerschen G,Poncelet F,Golinval J C.Physical interpretation of independent component analysis in structural dynamics[J].Me-chanical Systems and Signal Processing.20072,1(4):1561-1575
[5]张晓丹,姚谦峰,刘佩.基于快速独立分量分析的模态振型识别方法研究[J].振动与冲击2,0092,8(7):158-162
[6]静行,袁海庆,赵毅.基于独立分量分析的结构模态参数识别[J].振动与冲击2,0102,9(3):137-142
[7]Widodo A,Yang B S,Han Tian.Combination of independent component analysis and support vector machines for intelligent faultsdiagnosis of induction motors[J].Expert Systems with Applications.2007,32(2):299-312
[8]张育智,何伟,李乔.用于结构损伤识别的神经网络输入选取规则探究[J].四川建筑科学研究,20103,6(1):63-67
[9]闫桂荣,段忠东,欧进萍.基于结构振动信息的损伤识别研究综述[J].地震工程与工程振动,20072,7(3):95-104
[10]焦卫东,杨世锡,吴昭同.复合ICA-SVM机械状态模式分类[J.]中国机械工程,2004,15(1):62-65
[11]田景文,高美娟.人工神经网络算法研究及应用[M.]北京:北京理工大学出版社.2006.238-241
[12]Song H Z,Zhong L,Han B.Structural damage detection by integrating independent component analysis and support vector machine[J].International Journal of Systems Science.2006,37(13):961-967
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