Non-probabilistic information fusion technique for structural damage identification based on measured dynamic data with uncertainty

Based on measured natural frequencies and acceleration responses,a non-probabilistic information fusion technique is proposed for the structural damage detection by adopting the set-membership identification(SMI) and twostep model updating procedure.Due to the insufficiency and uncertainty of information obtained from measurements,the uncertain problem of damage identification is addressed with interval variables in this paper.Based on the first-order Taylor series expansion,the interval bounds of the elemental stiffness parameters in undamaged and damaged models are estimated,respectively.The possibility of damage existence(PoDE) in elements is proposed as the quantitative measure of structural damage probability,which is more reasonable in the condition of insufficient measurement data.In comparison with the identification method based on a single kind of information,the SMI method will improve the accuracy in damage identification,which reflects the information fusion concept based on the non-probabilistic set.A numerical example is performed to demonstrate the feasibility and effectiveness of the proposed technique.     

P-CS不确定性量化模型与其性能数据驱动更新方法

在实际工程中, 广泛存在大量的不确定性信息, 直接或间接影响着工程结构形式设计、结构性能评估与预测以及在役结构损伤识别等工作的开展与决策. 这些多源不确定性信息往往需要用多种不同的不确定性量化模型加以描述; 与此同时, 不确定性变量在使用过程中可能随时间变化且难以直接测量, 需要间接根据性能测试信息在使用工程中更新不确定性量化模型. 为兼顾上述两个问题, 本文基于等概率变换原则提出了一种P-CS (probability-convex set) 不确定性量化模型, 该模型将不确定性变量用概率随机变量与非概率凸集变量组合表征, 可统一表达概率模型、非概率模型以及非精确概率模型, 实现多源、多类型不确定性的统一量化. 本文进一步基于贝叶斯理论提出了一种针对该P-CS不确定性量化模型的性能数据驱动更新方法. 该更新方法根据性能测试数据信息更新P-CS不确定性量化模型参数取值的信度分布, 从而根据后验信度分布计算得出当前P-CS不确定性量化模型参数集合. 通过数值算例详述了P-CS不确定性量化模型的构建方法与其概率、非概率特性, 并验证了性能数据驱动更新P-CS模型方法的适用性.      

基于振动信号神经网络层合板分层损伤检测研究

基于振动信号应用神经网络研究层合板分层损伤的检测方法. 对层合板分层损伤区域, 采 用相同坐标不同节点建立了分层损伤处的有限元模型; 通过数值模拟提取结构无损和 不同程度面积分层损伤的全局振动标识量; 重点研究神经网络对层合板分层损伤位置 和损伤程度的检测技术. 研究表明, 用结构全局振动标识量作为人工神经网络的输入, 对 层合板结构分层损伤检测是一种很有效的工程实用技术, 可应用于实际结构的在线损伤检测.     

基于不完备频响函数的结构损伤统计识别研究

基于不完备频响函数数据,结合概率统计方法,提出了一种能同时考虑模型参数不确定性和测试噪声影响的结构损伤统计识别方法.首先,基于频响函数某一行向量在不同频率下的幅值数据,利用矩阵拉直运算建立了关于损伤系数的确定性识别方程.其次,假设模型参数误差和测试噪声为零均值的高斯随机变量,根据摄动理论,推导了损伤后结构刚度参数的前二...     

结构损伤的非概率区间识别方法

基于结构时域加速度响应,利用区间分析方法对具有外界激励和测量数据不确定性的结构系统进行损伤识别.不确定性量被处理为有界区间数,基于参考有限元模型和被测加速度响应,经过提出的两步模型参数的区间修正方法分别得到了未损伤和损伤结构区间模型,进而通过定义的PoDE(Possibility of Damage Existence)确定了结构各单元的损伤可能性.最后以十杆平面桁架系统为例,对区间方法与概率方法的损伤识别结果进行对比,讨论了损伤程度和不确定度对识别结果的影响,结果表明了本文方法的可行性与有效性.     

模态参数不确定性分析的贝叶斯方法研究

在结构损伤诊断和参数识别中,实测结构模态参数不可避免地存在误差。本文将模态参数视为随机变量,采用贝叶斯方法对模态参数的不确定性进行分析。分析中选用高斯联合概率密度函数作为先验密度函数,通过多次独立的模态参数测试,得到传递函数的条件概率密度函数和模态参数的后验估计表达式,再利用拉普拉斯渐近方法求解边缘概率密度函数,得到模态参数的最大后验估计。在钢筋混凝土框架结构的模态试验中,利用本文方法给出了结构模态参数的估计值,结果表明,本文方法具有良好的收敛性。     

基于贝叶斯方法的模态参数识别的不确定度分析

振动模态分析和模态参数识别是动态测试的一个重要研究方向。模态参数在模型的修正、响应的预测、系统的健康检测及控制等方面有着重要的作用。但动态测试的不确定度分析,尤其是模态参数的不确定度的研究还十分缺乏。本文主要基于贝叶斯方法,通过傅立叶变换(FFT)建立时域数据和频域数据之间的对应关系。根据共振频带内的多个模态的响应数据得到相对应的模态参数,优化得到模态参数的最佳估计值,评定模态参数识别的不确定度。在固支梁的模态实验中,加速度传感器采集环境激励中的振动数据,运用贝叶斯法进行处理得到模态参数的最佳估计值。在此基础上,通过模态参数的最佳估计值,以及仪器的检定报告数据,结合合成不确度分析方法,系统分析了模态参数识别的不确定度。     

基于凸集模型的结构地震多维易损性分析

将凸集模型应用于结构的地震多维易损性分析。建立钢筋混凝土框架结构模型,选择最大层间位移角和最大层加速度两种参数建立多维性能极限状态方程。通过平均信息熵理论,获得两种参数的区间估计。考虑椭球模型和区间模型两种形式的凸集模型,在标准空间内通过拉丁超立方抽样生成样本点,通过矩阵变换将其映射到凸集空间内,建立结构地震响应的凸集模型。将凸集变量样本点代入极限状态方程,进行了易损性分析。采用概率模型进行对比计算,研究表明,与概率模型相比,当PGA较小时,凸集模型的破坏概率较大,而PGA较大时,凸集模型的破坏概率较小;椭球模型和凸集模型的分析结果差距较小,在各个PGA下破坏概率差值仅为0.05~0.1,因此可以不考虑凸集类型不同对易损性分析结果的差异。     

基于L1正则化的随机梁式结构静力损伤识别方法

提出了一种结合摄动法和L1正则化方法的随机梁式结构静力损伤识别方法。考虑初始模型误差和测量误差的影响,建立了关于随机损伤指数的控制方程,并将摄动法和L1正则化方法相结合,对随机损伤指数的控制方程进行求解,进而从概率的角度对结构的损伤进行识别。损伤试验结果表明,和传统的最小二乘求解法相比,本文方法能够更为准确地识别多处局部损伤的位置及大小,对实际结构损伤检测具有较好的参考价值。     

砖石古塔模型振动台试验的离散元模拟分析

砖石古塔块体间黏结强度较低,受地震作用易产生裂缝后发生破坏,为研究砖石古塔在地震作用下开裂及裂缝发展机制,以玄奘塔1/8缩尺模型结构为对象,建立离散元模型,计算了地震波激励下结构的加速度及位移反应,与振动台试验结果进行对比,分析了塔体开裂破坏全过程。结果表明,数值计算所得结构的加速度及位移反应与振动台试验结果一致,当地震烈度较低时,两者顶层位移变化曲线基本一致;烈度提高后,塔体开裂导致结构动力响应的计算值与试验结果出现差异,但两者的变化规律相同。在地震波激励下,塔体首层先开裂,随地震烈度的提高延伸至中部塔层,块体第2层出现阶梯状错动,模型损伤过程与试验裂缝发展状况基本吻合。研究结果为砖石古塔地震损伤及破坏过程分析提供了参考。     

基于位移响应协方差参数的数值仿真和实测损伤识别研究

本文使用结构位移响应协方差参数进行结构损伤识别,首先推导和建立位移响应协方差参数的解析公式,它是结构频率、振型和阻尼等模态参数的函数,结构物理参数的改变会导致该协方差参数的改变;对一个七层框架结构进行数值模拟分析来演示该方法的有效性,通过比较结构不同状态下各单元位移响应协方差参数CoD的分布曲线,研究各单元CoD与损伤程度的关系曲线,发现损伤位置处的CoD改变最大,其次是对称和附近单元,通过单损伤和多损伤工况研究分析,表明基于结构损伤前后CoD的改变,能成功判定损伤发生和识别出损伤位置,最后把该方法应用于一个实验室简支钢梁的损伤识别,通过对锤击振动下的加速度响应进行二次积分得到位移响应,并比较钢梁损伤前后的CoD,得到损伤概率向量,成功识别出损伤位置。该方法具有较好的噪声鲁棒性,无需结构分析模型,计算简便,具有较好的工程应用性。     

高拱坝动力特性与动力响应模型试验研究

高拱坝的动力破坏特性越来越受到关注,动力特性和动力响应是其研究的一个重要方面。通过模型试验的方法研究高拱坝动力特性和动力响应的特征,动力特性试验比较了完好结构与有施工横缝结构的基阶动力特性,表明当结构存在施工横缝时,结构基阶频率降低,阻尼增大。结构动力响应采用振动台模拟地震波激励,试验结果表明,完整结构与有施工缝结构动力响应具体数据不同,但规律相同。加速度最大响应沿拱顶水平向在坝中拱顶的值最大,向两边的坝肩处,加速度最大响应逐渐减小。沿拱坝垂向加速度最大响应从坝顶到基础,响应均逐渐减小。试验结果可为高拱坝结构动力安全研究提供技术依据。     

Post-buckling analysis of a thin stiffened plate with uncertain initial deflection via interval analysis

The post-buckling behavior of the thin stiffened plates with uncertain initial deflection under uni-direction compression load is investigated based on Von-Karman large deflection theory. The uncertain initial deflections are considered to be unknown except that they belong to a given set in the interval range. Interval analysis model for computing the bounds of curves of the post-buckling deflection versus load of the plate is presented. Interval analysis model is compared with the stochastic model, which is taken as the benchmarks of accuracy for judgment. The results indicate that the non-probabilistic and stochastic methods will produce similar results for a large deviation of uncertainty. If the probabilistic information is unavailable, one should not propose the probabilistic method based on an arbitrary assumption on the distribution of the deflection coefficients. Rather, one should use the non-probabilistic method to uncertainty.     

基于模态柔度曲率的砖石古塔损伤识别研究

损伤识别是结构性能评估的有效途径之一,砖石古塔的模态柔度曲率是进行损伤识别的有效参数之一。为研究古塔结构损伤机制与模态柔度曲率的关系,依据兴教寺窥基塔结构模型振动台试验及数值模拟结果,分析周期性荷载作用下塔体动力响应及破坏特征,建立了塔体损伤程度与模态参数的关系模型。结果表明,古塔结构开裂破坏后,刚度减小,模态柔度曲率增大;数值计算与试验测试所得加速度及频域响应规律一致,塔体模型结构试件的开裂损伤随加载工况增多而持续累积,加载前期损伤参数变化缓慢,而后保持稳定,试验及模拟所得塔体的破坏过程一致。因此,塔体模态柔度曲率变化与古塔结构损伤累积具有一致性,可用于有效可靠识别砖石古塔结构损伤识别。     

基于贝叶斯理论的恢复力模型参数识别方法

提出了基于贝叶斯理论的恢复力模型参数识别方法,该方法考虑了模型误差的影响,结合实测滞回曲线数据,不仅可以得到模型参数的最有可能值,而且可以得到模型参数的定量的不确定性。以密肋复合墙体在低周反复荷载作用下所得滞回曲线为例,提出了可考虑刚度降低、捏拢滑移及极限荷载后强度降低现象的恢复力模型,建立了基于贝叶斯理论的恢复力模型参数识别计算框架,推导得到了模型参数的负对数似然函数,据此可得到模型参数的最有可能值及协方差矩阵。对标准密肋复合墙体预制试件和现浇试件的恢复力模型参数进行了识别,将根据模型参数最有可能值得到的滞回曲线及根据模型参数最有可能值及协方差矩阵得到的骨架曲线,与相应的实测值进行了对比,验证了所提方法的可行性及识别结果的合理性,更新的模型参数概率分布可用于后续的抗震风险评估。     

Test analysis of detection of damage to a complicated spatial model structure

A two-stage damage detection approach is proposed and experimentally demonstrated on a complicated spatial model structure with a limited number of measurements. In the experiment,five known damage patterns,including 3 brace damage cases and 2 joint damage cases,were simulated by removing braces and weakening beam-column connections in the structure. The limited acceleration response data generated by hammer impact were used for system identification,and modal parameters were extracted by using the eigensystem realization algorithm. In the first stage,the possible damaged locations are determined by using the damage index and the characteristics of the analytical model itself,and the extent of damage for those substructures identified at stage I is estimated in the second stage by using a second-order eigen-sensitivity approximation method. The main contribution of this paper is to test the two-stage method by using the real dynamic data of a complicated spatial model structure with limited sensors. The analysis results indicate that the two-stage approach is ableto detect the location of both damage cases,only the severity of brace damage cases can be assessed,and the reasonable analytical model is critical for successful damage detection.     

随机梁式结构静力损伤识别的一种改进方法

针对已有的损伤识别方法会出现损伤误识别的问题,本文在已有方法的基础上发展了一种随机梁式结构静力损伤识别的改进方法。假定静力荷载下梁式结构初始模型参数（如弹性模量和几何尺寸等）及测量误差为随机量,给出已有的基于随机有限元模型的梁式结构静力损伤识别方法,并进一步提出了一种改进方法。该方法通过设定损伤概率指标的阈值和反复迭代对结构损伤识别进行改进。数值算例和简支梁静力试验表明,考虑初始模型的不确定性以及静力响应测量误差,本文方法相较已有方法可以更有效地识别梁式结构的损伤。     

Dynamic stability of a pipe conveying fluid with an uncertain computational model

This paper deals with the problem of a pipe conveying fluid of interest in several engineering applications, such as micro-systems or drill-string dynamics. The deterministic stability analysis developed by Paidoussis and Issid (1974) is extended to the case for which there are model uncertainties induced by modeling errors in the computational model. The aim of this work is twofold: (1) to propose a probabilistic model for the fluid–structure interaction considering modeling errors and (2) to analyze the stability and reliability of the stochastic system. The Euler–Bernoulli beam model is used to model the pipe and the plug flow model is used to take into account the internal flow in the pipe. The resulting differential equation is discretized by means of the finite element method and a reduced-order model is constructed from some eigenmodes of the beam. A probabilistic approach is used to model uncertainties in the fluid–structure interaction. The proposed strategy takes into account global uncertainties related to the noninertial coupled fluid forces (related to damping and stiffness). The resulting random eigenvalue problem is used to analyze flutter and divergence unstable modes of the system for different values of the dimensionless flow speed. The numerical results show the random response of the system for different levels of uncertainty, and the reliability of the system for different dimensionless speeds and levels of uncertainty.     

The effect of stacking sequence on the impact-induced damage in cross-ply E-glass/epoxy composite plates

The aim of this study is to determine the damage mechanisms of unidirectional E-glass/epoxy laminated composites under localized impacts. The projectile velocities at low ranges (0.54–3.10 m/s) have been considered to establish a parametric analysis of clamped laminated composite responses. The used circular plates are symmetrical laminates which are composed of ten plies and have three different cross-ply stacking sequences: [0₂/90₆/0₂], [0₃/90₄/0₃] and [0₄/90₂/0₄]. They are subjected at their centers to an impact of an aluminum projectile which is applied in the transverse direction by using a drop weight machine. The time histories of the impactor acceleration, the projectile displacement and the plate deflection were measured. The analysis of the nature of the damage mechanisms and its relation to the structural responses has been performed. The identification of the matrix cracks and the delamination at the interfaces has allowed to determine the initiation and extension criteria of the damage.