Structural damage detection using the optimal weights of the approximating artificial neural networks

This work presents a novel neural network‐based approach to detect structural damage. The proposed approach comprises two steps. The first step, system identification, involves using neural system identification networks (NSINs) to identify the undamaged and damaged states of a structural system. The partial derivatives of the outputs with respect to the inputs of the NSIN, which identifies the system in a certain undamaged or damaged state, have a negligible variation with different system errors. This loosely defined unique property enables these partial derivatives to quantitatively indicate system damage from the model parameters. The second step, structural damage detection, involves using the neural damage detection network (NDDN) to detect the location and extent of the structural damage. The input to the NDDN is taken as the aforementioned partial derivatives of NSIN, and the output of the NDDN identifies the damage level for each member in the structure. Moreover, SDOF and MDOF examples are presented to demonstrate the feasibility of using the proposed method for damage detection of linear structures. Copyright © 2001 John Wiley & Sons, Ltd.     

Application of clan member signal method in structural damage detection

It is well known that in most cases, a reference is necessary for structural health diagnosis, and it is very difficult to obtain such a reference for a given structure. In this paper, a clan member signal method (CMSM) is proposed for use in structures consisting of groups (or clans) that have the same geometry, i.e., the same cross section and length, and identical boundary conditions. It is expected that signals measured on any undamaged member in a clan after an event could be used as a reference for any other members in the clan. To verify the applicability of the proposed method, a steel truss model is tested and the results show that the CMSM is very effective in detecting local damage in structures composed of identical slender members.     

桥梁健康监测及诊断研究综述

本文分析了桥梁结构健康监测和诊断的必要性和紧迫性，对桥梁结构损伤探测的方法进行了详细的分类及论述。提出了桥梁健康监测和诊断的研究新领域——无线监测系统。最后指出了桥梁健康监测和诊断中存在的问题及发展趋势。     

基于EMD方法与RBF神经网络的结构损伤检测

针对结构损伤检测中损伤的识别、定位以及程度的标定这三个独立并按一定先后顺序进行的检测过程,提出了一种能将以上三者同时进行的联合检测方法。该方法首先利用经验模态分解(EMD)方法将三层钢筋混凝土剪切型结构在各种损伤工况下的顶层地震作用加速度响应分解为若干固有模态函数(IMF)分量,然后以此IMF分量和未经EMD分解的原始加速度响应数据来构造损伤标识量,作为特征参数依次输入到径向基函数神经网络(RBFNN)中进行损伤检测。给出了应用此方法的具体步骤,通过仿真实验证明了利用该方法进行结构损伤一次检测的可行性和有效性,结果表明,由加速度响应经EMD分解而得到的IMF分量输入到RBFNN中能够更为精确地一次检测出结构所有损伤信息,并且RBFNN在结构损伤损度大时具有更好的检测效果。     

应用人工神经网络技术的大型斜拉桥子结构损伤识别研究

本文应用人工神经网络技术对大型斜拉桥结构进行了子结构损伤识别研究。文中首先介绍了子结构损伤识别的基本方法，然后应用自组织竞争神经网络建立了对于大型桥梁结构识别子结构损伤情况的子结构损伤识别方法，并且应用BP网络进一步建立了大型桥梁结构各子结构内部的损伤位置和损伤程度的识别方法，数值模拟了一大跨度斜拉桥子结构损伤以及子结构内部损伤的识别过程，最后得出结论：(1)基于自组织竞争网络的子结构损伤识别方法能迅速准确地识别大型结构的损伤情况；(2)基于BP网络所建立的结构损伤识别方法，能对子结构中结构损伤的位置和程度进行进一步的识别；(3)基于人工神经网络技术的结构损伤识别方法是大型土木工程结构损伤识别的有效方法，可在工程结构损伤识别中广泛应用。     

基于BP神经网络的空间索杆结构节点损伤识别研究

针对某实际空间索杆结构的节点损伤现象,采用BP神经网络与基于振动的损伤识别两步法对其进行了识别研究,即首先确定可能发生节点损伤的子区域,在此基础上利用对应子区域的子网络识别出具体的损伤位置和程度。识别过程中采用两个杆单元模拟发生节点损伤的杆件,用抗弯刚度降低的端部短杆单元模拟节点损伤。研究表明,虽然空间索杆结构的动力性能较为复杂,但基于结构固有频率和模态位移的组合指标对节点损伤仍较为敏感,利用它们进行节点损伤识别是有效的。     

基于模态分析和神经网络的裂缝损伤识别

提出了裂缝损伤诊断的神经网络方法,探讨了用模态技术和神经网络对混凝土结构裂缝损伤进行识别与定位的方法。文中以一简支矩形截面梁为研究对象,通过完好结构和损伤结构的有限元分析,获取两者的损伤标识量,输入BP神经网络训练。以损伤位置和裂缝高度作为输出参数,对其进行单处损伤定位的研究。数值仿真结果表明,采用神经网络方法可以对裂缝做出较好的诊断。     

基于BP神经网络的空间索杆结构承重索损伤识别研究

针对某实际空间索杆结构的承重索预应力松弛现象,采用BP神经网络与基于振动的损伤识别方法,分别对单榀承重索和双榀承重索的预应力松弛进行了识别研究。研究表明,虽然空间索杆结构的动力性能较为复杂,但归一化后的结构前10阶固有频率变化比和归一化后的对应结构第一阶模态的部分节点的损伤信号指标对承重索的损伤位置较为敏感,利用其进行损伤定位是可行的。在此基础上,再增加考虑结构的第一阶同有频率平方变化比即可进一步有效识别出承重索的损伤程度。     

基于模糊神经网络和证据理论的结构损伤识别新方法

为了有效利用结构健康监测系统中的多源不确定数据,提高损伤识别的正确率,通过构造模糊神经网络(FNN)分类器,提出了一种新的概率赋值函数构造方法和数据融合损伤识别新方法.该损伤识别方法先对数据预处理,提取有效的特征参数,接着将它作为FNN的输入,构造FNN分类器,最后运用数据融合中的D-S证据理论计算出融合决策结果.为了验证所提方法的有效性,通过一个七层剪切型框架结构的数值模型,分别用单一FNN分类器和数据融合损伤识别方法进行了损伤识别和比较.研究结果表明,本文所提方法比单一决策结果更准确,具有更高的可靠度。     

混沌噪声背景下检测微弱信号的神经网络方法分析

地震勘探资料的噪声许多呈现混沌现象，利用传统的去噪方法效果并不理想，如何根据混沌固有的性质，对地震勘探资料中的有效信号进行提取是许多科学工作者极为关注的问题，针对这种混沌噪声下的微弱信号检测，本文提出三种神经网络方法并对此进行比较，理论分析及仿真实验表明这三种神经网络在信噪比达到—37dB时，均能检测混沌噪声背景中的微弱信号。     

Earthquake-induced damage detection and localization in masonry structures using smart bricks and Kriging strain reconstruction: A numerical study

The intrinsic vulnerability of masonry structures to seismic events makes structural health monitoring of the utmost importance for the conservation of the built heritage. The development of piezoresistive bricks, also termed smart bricks, is an innovative technology recently proposed by the authors for the monitoring of such structures. Smart bricks exhibit measurable variations in their electrical properties when subjected to external loads or, alternatively, strain self-sensing capabilities. Therefore, the deployment of a network of smart bricks into a masonry structure confers self-diagnostic properties to the host structure. In this light, this paper presents a theoretical investigation on the application of smart bricks to full-scale masonry structures for seismic assessment. This includes the study of the convenience of providing electrical isolation conditions to the sensors, as well as the effectiveness of smart bricks when installed into either new constructions or in pre-existing structures. Secondly, numerical results are presented on the seismic analysis of a three-dimensional masonry building equipped with a network of smart bricks. Finally, in order to map the strain field throughout the structure exploiting the outputs of a limited number of sensors, an interpolation-based strain reconstruction approach is proposed.     

A neural network approach for structural identification and diagnosis of a building from seismic response data

This work presents a novel procedure for identifying the dynamic characteristics of a building and diagnosing whether the building has been damaged by earthquakes, using a back‐propagation neural network approach. The dynamic characteristics are directly evaluated from the weighting matrices of the neural network trained by observed acceleration responses and input base excitations. Whether the building is damaged under a large earthquake is assessed by comparing the modal parameters and responses for this large earthquake with those for a small earthquake that has not caused this building any damage. The feasibility of the approach is demonstrated through processing the dynamic responses of a five‐storey steel frame, subjected to different strengths of the Kobe earthquake, in shaking table tests. Copyright © 2002 John Wiley & Sons, Ltd.     

A frequency response function change method for damage localization and quantification in a shear building under ground excitation

A frequency response function change (FRFC) method to detect damage location and extent based on the change in the frequency response functions of a shear building under the effects of ground excitation was proposed in this paper. The damage identification equation was derived from the motion equations of the system before and after the occurrence of the damage. Efforts to make the FRFC method less model‐dependent were made. Intact system matrices, which could be estimated using the measured data without the need for an analytical model, and the frequency response functions were required for the FRFC method. The effects of measurement noise and model parameter error in the FRFC method were studied numerically. The proposed FRFC method was validated by experimental studies of a six‐story steel building structure with single and multiple damage cases. Copyright © 2012 John Wiley & Sons, Ltd.     

One‐year monitoring of the Z24‐Bridge: environmental effects versus damage events

When using the analysis of vibration measurements as a tool for health monitoring of bridges, the problem arises of separating abnormal changes from normal changes in the dynamic behaviour. Normal changes are caused by varying environmental conditions such as humidity, wind and most important, temperature. The temperature may have an impact on the boundary conditions and the material properties. Abnormal changes on the other hand are caused by a loss of stiffness somewhere along the bridge. It is clear that the normal changes should not raise an alarm in the monitoring system (i.e. a false positive), whereas the abnormal changes may be critical for the structure's safety. In the frame of the European SIMCES‐project, the Z24‐Bridge in Switzerland was monitored during almost one year before it was artificially damaged. Black‐box models are determined from the healthy‐bridge data. These models describe the variations of eigenfrequencies as a function of temperature. New data are compared with the models. If an eigenfrequency exceeds certain confidence intervals of the model, there is probably another cause than the temperature that drives the eigenfrequency variations, for instance damage. Copyright © 2001 John Wiley & Sons, Ltd.     

基于小波分析的结构损伤检测研究进展

近10几年来,在土木和机械领域结构损伤识别方法已引起不同领域的相关学者的极大研究兴趣。小波分析是一种新的数学分析和信号处理工具,可以对非稳态信号进行详细的时频分析,是传统傅里叶分析所不能及的,已广泛应用于土木、机械和航空工程领域中,特别是在结构损伤识别和健康监测中的应用尤为突出。本文回顾和总结了小波分析理论及其在结构损伤识别、损伤定位和损伤程度确定中的应用,对今后的研究进行了讨论和展望。     

A Bayesian probabilistic approach for structure damage detection

A Bayesian probabilistic approach is presented for the damage detection of multistorey frame structures. In this paper, a Bayesian probabilistic approach is applied to identify multiple damage locations using estimated modal parameters when (1) the measurement data are potentially corrupted with noise, (2) only a small number of degrees of freedom are measured, and (3) a few fundamental modes are estimated. To reduce the potentially intensive computational cost of the proposed method, a branch-and-bound search scheme is proposed and a simplified approach for the modelling of multistorey frame structures is employed. A six-storey shear frame example and two multistorey frame examples, with multiple damage locations, are presented to illustrate the applicability of the proposed approach. © 1997 John Wiley & Sons, Ltd.     

高耸塔架结构节点损伤基于神经网络的两步诊断法

本文针对高耸钢塔架结构的损伤特点，建立了具有节点损伤的有限元模型，提出了一种分层神经网络两步诊断法：第一步，由基于区域残余力理论的第一层神经网络进行结构损伤区域的初诊；第二步，由基于应变模态理论的第二层神经网络进行损伤区域内的具体损伤节点位置和程度的诊断。对一平面塔架结构的数值仿真分析表明：本文提出的损伤诊断方法的结果是令人满意的。     

Bayesian probabilistic damage detection of a reinforced‐concrete bridge column

A Bayesian probabilistic approach for damage detection has been proposed for the continuous monitoring of civil structures (Sohn H, Law KH. Bayesian probabilistic approach for structure damage detection. Earthquake Engineering and Structural Dynamics 1997; 26 :1259–1281). This paper describes the application of the Bayesian approach to predict the location of plastic hinge deformation using the experimental data obtained from the vibration tests of a reinforced‐concrete bridge column. The column was statically pushed incrementally with lateral displacements until a plastic hinge is fully formed at the bottom portion of the column. Vibration tests were performed at different damage stages. The proposed damage detection method was able to locate the damaged region using a simplified analytical model and the modal parameters estimated from the vibration tests, although (1) only the first bending and first torsional modes were estimated from the experimental test data, (2) the locations where the accelerations were measured did not coincide with the degrees of freedom of the analytical model, and (3) there existed discrepancies between the undamaged test structure and the analytical model. The Bayesian framework was able to systematically update the damage probabilities when new test data became available. Better diagnosis was obtained by employing multiple data sets than just by using each test data set separately. Copyright © 2000 John Wiley & Sons, Ltd.     

光滑地表面毁伤检测方法研究

光滑地面毁伤程度评估非常重要,目前评估方法主要是基于光学图像,可是在恶劣的气候或环境条件下,很难获得光学图像,合成孔径雷达（SAR）克服了这个缺点. SAR图像是地表结构和电特征等地球物理参数的映射,通过SAR图像反演地表结构参数,可以推测出光滑地面毁伤程度. 本文以几何光学模型（GOM）为基础,建立神经网络反演模型,以获得光滑地表面受损后的粗糙度参数：表面均方根高度（σ）和表面相关长度（l）,并进一步评估光滑地表面受损程度. 实验结果表明该方法可行.     

工作状态下梁及桥梁模态与损伤识别的新方法

工作状态下桥梁结构的模态参数识别是桥梁损伤识别的重要环节,考虑桥梁检测的实用性,桥梁检测一般应建立在环境激励的基础上,已有的环境激励下模态参数识别的方法对模态频率的识别的精度较高,而对位移模态的识别则误差较大。提出了一种利用移动质量块在不同位置时对桥梁的模态频率进行多次测量,用各次测得的频率值确定位移模态的新方法,使得位移模态识别的精度接近频率识别的精度,建立了该方法的初步模型,推导了频率与位移模态关系的理论公式,并通过数值模拟对该方法的有效性进行了说明。