工程结构模态参数辨识与损伤识别方法研究

发布时间：2018-05-31 20:31

本文选题：工程结构 + 模态参数辨识　；参考：《重庆大学》2013年博士论文

【摘要】：工程结构模态参数辨识与损伤识别技术作为近20年来适应工程实际需要而发展起来的一门新学科，有很强的工程背景，具有重要的实用价值。近些年来，工程结构的健康监测、损伤评估越来越受到人们的关注，，而模态参数辨识和损伤识别作为其核心技术和理论基础已日益成为土木工程领域的研究热点。由于工程结构体积庞大、约束条件复杂、材料混杂等原因，对其进行人为激励以及对激励信号进行有效测量是相当困难的，因此，基于输入输出信号的传统模态参数识别理论和方法在工程结构中难以适用。而环境激励下的结构模态参数识别方法，具有无需施加人为激励、费用低廉、不影响结构的正常工作、无需测量激励信号、更加符合实际情况等优点，在工程界得到了广泛的应用。但现有的模态参数辨识方法和损伤识别方法在精度、鲁棒性、效率以及经济性能指标方面仍存在很多不足，在实际工程中的应用尚处于发展阶段，仍需进一步研究和完善。针对目前研究中存在的问题，本文围绕环境激励下工程结构模态参数辨识方法和基于模态参数灵敏度的损伤识别方法开展研究，归结起来主要内容如下： ①计算三层钢筋砼框架模型在环境激励下的位移响应，利用两种数据预处理方法（随机减量法和NExT法）和ITD法、STD法、复指数法、ERA法和ARMA法等5种时域模态参数识别方法，开展了结构模态参数辨识的比较研究。结果表明：预处理方法中的NExT法在精度、抗噪性上均优于随机减量法；五种模态参数识别方法中STD法和ARMA法的对频率识别精度比其它三种方法稍高，在抗噪方面，STD、ERA法、ARMA法的抗噪能力比其它两种方法稍强，所有方法对频率的识别精度均远大于对阻尼比的识别精度。 ②结合随机子空间法提出了环境激励下结构模态参数识别的改进ITD法、改进STD法与改进复指数法。随机子空间法的识别精度高，其中数据的协方差计算（矩阵正交投影计算）可以保留原始数据中的所有信息，同时去除了噪声，将得到的Toeplitz矩阵（P矩阵）中的数据作为ITD法、STD法与复指数法的输入数据，这三种方法就不再需要采用随机减量法或者自然激励技术法进行前处理，从而避免了这两种前处理方法的不准确性带来的误差。对两跨三层框架模型及一自锚式悬索桥模型的模态参数进行了识别，结果表明：基于协方差驱动SSI与数据驱动SSI的改进方法对比对应随机子空间法，在精度未减小的前提下提高了计算效率，仅用较少的数据就可较准确地识别出系统的模态参数，且识别精度较高、抗噪性较好；改进方法与ITD、STD、复指数法相比有精度上的优势。 ③开展了在环境激励下的框架结构模态辨识实验，同时针对十二层钢筋混凝土框架结构的振动台试验模型进行了模态参数识别。结果进一步证明了所提出的改进方法的正确性、可行性；基于SSI法的改进辨识方法计算时间约为SSI法计算时间的50%，当输出信号较多时，这种优势更明显。从而可见，基于SSI法对ITD、STD、复指数法进行改进后，精度没有降低，同时缩短了计算时间，这将为改进方法应用到结构的实时监测提供了可能。 ④在随机子空间辨识法与特征系统实现算法（ERA）的识别结构参数过程中，确定系统阶次是关键。研究了基于奇异值差分谱的去噪原理以及基于奇异值差分谱的分量分离原理。提出了基于奇异值差分谱的随机子空间和ERA模型定阶方法，通过该法来确定模型阶次所产生的虚假模态是最少的，且包含信号中所有模态，同时识别精度不受影响，并且计算量最小的阶次。通过试验和数值分析进行结构的模态识别，结果表明该方法是有效的。 ⑤采用相关系数法进行真实模态分量的挑选，剔除低频虚假模态分量。采用数字滤波器对EMD分解进行改进，从而克服EMD分解时出现模态混叠的情况。针对经过EMD分解或者数字滤波后的单频率分量的信号，提出了基于奇异值差分谱的去噪的方法，仿真信号和实验数据分析表明该法去噪效果相当明显。提出了基于奇异值差分谱的模态分离HHT法，通过对仿真信号和实验数据的模态辨识，表明该方法是可行的。 ⑥提出了基于数字滤波器的STD法、复指数法、ARMA法等三种识别方法，通过仿真信号、实验数据处理证明了方法的正确性、可行性，同时表明该方法能够有效分离密集模态，且在识别过程中无需考虑定阶问题。提出了基于EMD分解的STD法、复指数法、ARMA法等三种识别方法，通过仿真信号、实验数据证明了其正确性、可行性，同时表明该法能够处理非平稳信号，在识别过程中无需考虑定阶问题。 ⑦在现有的直接解析法基础上，本文从三个方面对其进行了改进（包括直接解析法的模型缩聚改进，方程迭代解法的改进，模态截尾误差的改进），提出了框架结构损伤识别的改进直接解析法。针对五层两跨的框架结构，开展了基于改进模态参数灵敏度法的结构损伤识别方法研究，结果表明：在无噪声情况下识别结果非常正确，而在有噪声的情况下识别结果受到显著的影响，但在0.1%的噪声水平下，对于可接受识别结果的正确保证率可以达到80%以上。成果可为确立工程结构的科学鉴定系统，制定新的检测规范提供参考。
[Abstract]:In recent years , the traditional modal parameter identification method based on input - output signal has been widely used in engineering structure . However , the traditional modal parameter identification method and the damage identification method have many advantages in engineering structure . However , the existing modal parameter identification method and damage identification method still have many shortcomings in the aspects of accuracy , robustness , efficiency and economic performance .

Aiming at the problems existing in the current research , this paper studies the modal parameter identification method and the damage identification method based on modal parameter sensitivity under the environment excitation , which is summarized as follows :

The results show that the NExT method is superior to the random decrement method in the preprocessing method .
In the five modal parameter identification methods , the frequency recognition accuracy of the STD method is slightly higher than that of the other three methods . In the aspect of anti - noise , the anti - noise ability of the STD , ERA method and the auto - detection method is slightly stronger than that of the other two methods , and the recognition accuracy of all the methods to the frequency is much greater than the recognition accuracy of the damping ratio .

In this paper , the method of improving the structure modal parameters based on the random subspace method is proposed . The method improves the STD method and the improved complex index method . The recognition accuracy of the random subspace method is high , and the data in the data covariance matrix ( matrix orthogonal projection calculation ) can retain all the information in the original data . The results show that the method of the two - span three - layer frame model and the self - anchored suspension bridge model can be used to improve the calculation efficiency . The modal parameters of the system can be accurately identified with less data , and the recognition accuracy is high and the noise - noise is better .
The improved method has the advantage of precision compared with itd , STD and complex exponential method .

( 3 ) The modal identification experiment of frame structure under the environment excitation is carried out , and the modal parameter identification is carried out for the vibration table test model of the twelve - layer reinforced concrete frame structure .
The improved identification method based on the SSI method is about 50 % of the calculation time of the SSI method . This advantage is more obvious when the output signal is much more obvious . It can be seen that the accuracy is not reduced after the improvement of the itd , STD and complex index method based on the SSI method , and the calculation time is shortened , which provides the possibility for the improvement method to be applied to the real - time monitoring of the structure .

In the process of identifying structure parameters of stochastic subspace identification and characteristic system , it is crucial to determine the order of the system . The principle of noise removal based on singular value difference spectrum and component separation principle based on singular value difference spectrum are studied . A stochastic subspace and ERA model order method based on singular value difference spectrum is studied .

The method , simulation signal and experimental data analysis based on singular value difference spectrum are proposed for the signal of single frequency component subjected to EMD decomposition or digital filtering . The method of modal separation based on singular value difference spectrum is proposed , and the modal identification of the simulation signal and experimental data is proposed .

This paper presents three recognition methods based on the digital filter , such as the STD method , the complex exponential method and the auto - index method . The correctness and feasibility of the method are proved through the simulation signal and the experimental data processing .

On the basis of the existing direct analysis method , this paper improves the structure damage identification method based on the improved modal parameter sensitivity method .
【学位授予单位】：重庆大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：TU317

【参考文献】