[1]刘军香,王立新,姜慧,等.基于加速度二次协方差矩阵和神经网络的结构损伤识别[J].地震工程与工程振动,2019,39(03):214-221.[doi:10.13197/j.eeev.2019.03.214.liujx.022]
 LIU Junxiang,WANG Lixin,JIANG Hui,et al.Structural damage identification based on covariance of covariance of matrix of acceleration and neural network[J].EARTHQUAKE ENGINEERING AND ENGINEERING DYNAMICS,2019,39(03):214-221.[doi:10.13197/j.eeev.2019.03.214.liujx.022]
点击复制

基于加速度二次协方差矩阵和神经网络的结构损伤识别
分享到:

《地震工程与工程振动》[ISSN:/CN:]

卷:
39
期数:
2019年03
页码:
214-221
栏目:
论文
出版日期:
2019-06-30

文章信息/Info

Title:
Structural damage identification based on covariance of covariance of matrix of acceleration and neural network
作者:
刘军香1 王立新234 姜慧234 朱嘉健23 卢滔1
1. 防灾科技学院, 河北 三河 065201;
2. 广东省地震局 中国地震局地震监测与减灾技术重点实验室, 广东 广州 510070;
3. 广东省地震局 广东省地震预警与重大工程安全诊断重点实验室, 广东 广州 510070;
4. 深圳防灾减灾技术研究院, 广东 深圳 518003
Author(s):
LIU Junxiang1 WANG Lixin234 JIANG Hui234 ZHU Jiajian23 LU Tao1
1. Institute of Disaster Prevention, Sanhe 065201, China;
2. CEA Key Laboratory of Earthquake Monitoring and Disaster Mitigation Technology, Guangdong Earthquake Agency, Guangzhou 510070, China;
3. Guangdong Provincial Key Laboratory of Earthquake Early Warning and Safety Diagnosis of Major Projects, Guangdong Earthquake Agency, Guangzhou 510070, China;
4. Shenzhen Academy of Disaster Prevention and Reduction, Shenzhen 518003, China
关键词:
加速度响应白噪声激励二次协方差矩阵BP神经网络损伤识别
Keywords:
acceleration responsewhite noise excitationcovariance of covariance matrixBP neural networkdamage identification
分类号:
P315.9
DOI:
10.13197/j.eeev.2019.03.214.liujx.022
摘要:
为了能对结构早期损伤进行有效识别,本文提出了一种基于加速度响应二次协方差(CoC)矩阵和神经网络的结构损伤识别方法。首先通过数值模拟,以白噪声作为激励,获取结构在不同损伤位置和损伤程度下的加速度响应,并计算相应的二次协方差矩阵;然后,把二次协方差矩阵作为BP神经网络的输入特征向量,对网络进行训练并对损伤位置和损伤程度同时进行识别。本文以桁架为例,将二次协方差矩阵和BP神经网络结合,对结构单损伤和多损伤分别进行识别,同时采用模态频率和模态振型与BP神经网络结合作为对比指标。对比发现:相比于模态指标,基于加速度响应二次协方差矩阵和BP神经网络的损伤识别方法,能够较好的识别结构的单损伤和多损伤,且具有更好的稳定性和抗噪性。
Abstract:
In order to identify structural early damage effectively, this paper proposed a method for structural damage identification based on the Covariance of Covariance (CoC) matrix of acceleration responses and neural network method. Firstly, under the white noise excitation, the acceleration responses of the structure with different damages is obtained and used to calculate the corresponding CoC matrix. Secondly, the covariance of covariance matrix was used as the input feature vector for back propagation (BP) neural network to train the network. Thirdly, the trained network is used to identify the structural damage state. In this paper, a truss structure is used as an example. The CoC matrix of acceleration responses was combined with the BP neural network to identify the damage. In the meanwhile, modal frequency and modal shape combined with BP neural network were used as the comparison index. Compared with the modal index, the damage identification method based on CoC matrix and BP network can better identify the damages, and has better stability and noise resistance ability.

参考文献/References:

[1] 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, 2007, 27(3):95-103. YAN Guirong, DUAN Zhongdong, OU Jinping.Review on structural damage detection based on vibration data[J].Earthquake Engineering and Engineering Dynamics, 2007, 27(3):95-103.(in Chinese)
[2] BARAI S V, PANDEY P C.Vibration signature analysis using artificial neural networks[J].Journal of Computing in Civil Engineering, 1995, 9(4):259-265.
[3] ELKORDY M F, CHANG K C, LEE G C.Neural networks trained by analytically simulated damage states[J].Journal of Computing in Civil Engineering, 1993, 7(2):130-145.
[4] 段晨东, 刘义艳.基于多测点小波包特征矢量的神经网络损伤诊断方法[C]//第11届全国设备故障诊断学术会议.南京:中国振动工程学会, 2008:31-33. DUAN Chendong, LIU Yiyan.A structure damage diagnosis neural network method based on feature vectors composed of wavelet packets from multi-sensors[C]//The 11th National Conference on Equipment Fault Diagnosis.Nanjing:China Vibration Engineering Society, 2008:31-33.(in Chinese)
[5] 刘义艳, 巨永锋, 段晨东, 等.基于小波包变换和神经网络的损伤诊断方法[C]//2010年第三届计算智能与工业应用国际会议, 2010, 168-171. LIU Yiyan, JU Yongfeng, DUAN Chendong, et al.A structure damage diagnosis method based on wavelet packet transform and neural network[C]//Proceedings of 2010 the 3rd International Conference on Computational Intelligence and Industrial Application, 2010, 168-171.(in Chinese)
[6] XIE Donghai, TANG Hongwei.Application of wavelet packet transform and neural network to detect damage of elastic thin plate[J].Advanced Materials Research, 2012, 11(594):1105-1108.
[7] 徐菁, 曲丽敏, 卢翠萍.大跨度空间网格结构的健康监测系统[J]. 兰州理工大学学报, 2016, 42(4):128-133. XU Jing, QU Limin, LU Cuiping.Healthy monitor system with large-span space-netted structure[J].Journal of Lanzhou University of Technology, 2016, 42(4):128-133.(in Chinese)
[8] LI Xueyan, LAW S S.Matrix of the covariance of covariance of acceleration responses for damage detection from ambient vibration measurements[J]. Mechanical Systems and Signal Processing, 2010, 24(4):945-956.
[9] LI Xueyan, LAW S S, WANG Lixin.Health monitoring of in-service bridge deck by covariance of covariance matrix of acceleration[J]. Applied Mechanics and Materials, 2011, 7(27):4808-4814.
[10] LI Xueyan, WANG Lixin, LAW S S.Damage detection for structures under ambient vibration via covariance of covariance matrix and consistent regularization[J].Advances in Structural Engineering, 2013, 16(1):77-86.
[11] WANG Lixin, LI Xueyan, TAN Y, et al.Long-term health monitoring of in-service bridge deck by covariance of covariance matrix of acceleration responses[J].Advances in Structural Engineering, 2015, 18(12):2129-2149.
[12] HUI Y, LAW S S, KU C J.Structural damage detection based on covariance of covariance matrix with general white noise excitation[J].Journal of Sound and Vibration, 2017, 389:168-182.
[13] 王立新, 李雪艳, 姜慧, 等. 基于加速度二次协方差矩阵参数变化比法的环境振动下结构损伤识别[J].振动与冲击, 2016, 35(8):143-147+158. WANG Lixin, LI Xueyan, JIANG Hui, et al.Damage identification under ambient vibration based on change ratios of covariance of covariance matrix components of structural acceleration reponses[J].Journal of Vibration and Shock, 2016, 35(8):143-147+158.(in Chinese)

相似文献/References:

[1]王宗纲,孙宁,高象波,等.汽轮发电机框架式基础1:10模型动力分析[J].地震工程与工程振动,2005,25(01):033.
 Wang Zonggang,Sun Ning,Gao Xiangbo,et al.Modal analysis for a 1:10 model of turbine-dynamotor frame foundation[J].EARTHQUAKE ENGINEERING AND ENGINEERING DYNAMICS,2005,25(03):033.
[2]耿万里,刘敦宇,蔡永恩,等.预测北京地铁16号线振动对北京大学精密仪器的影响[J].地震工程与工程振动,2014,34(06):019.[doi:10.13197/j.eeev.2014.06.19.gengwl.003]
 GENG Wanli,LIU Dunyu,CAI Yongen,et al.Prediction of the influence of the proposed Beijing metro line 16 on a precise instrument of Peking University[J].EARTHQUAKE ENGINEERING AND ENGINEERING DYNAMICS,2014,34(03):019.[doi:10.13197/j.eeev.2014.06.19.gengwl.003]

备注/Memo

备注/Memo:
收稿日期:2018-11-10;改回日期:2019-1-8。
基金项目:地震科技星火计划项目(XH16031);广东省科技计划项目(2015A020217007);中央高校基本科研业务费(ZY20180303)
作者简介:刘军香(1993-),女,硕士研究生,主要从事结构健康监测与损伤识别研究.E-mail:1211354454@qq.com
通讯作者:王立新(1976-),男,研究员,主要从事结构健康监测与损伤识别研究.E-mail:wlxustc@hotmail.com
更新日期/Last Update: 1900-01-01