环境振动下工程结构模态辨识方法改进与软件系统研发

发布时间：2018-01-16 05:35

本文关键词：环境振动下工程结构模态辨识方法改进与软件系统研发　出处：《重庆大学》2014年硕士论文　论文类型：学位论文

【摘要】：工程结构模态辨识和损伤识别作为健康监测与安全性评估核心技术和理论基础，已经成为了土木工程领域关注的焦点和研究的热点。鉴于工程结构具有结构尺寸大、约束条件复杂、造型复杂等特点，对工程结构进行人为激励以及对激励信号进行有效的测量变得非常困难，从而传统的基于输入输出信号模态参数辨识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辨识方法具有其自身突出的优点：不需要施加人为激励、不需要测量激励信号、测量费用较低、测量时不影响结构的正常工作、更加符合实际情况等，在工程界得到了广泛的应用。但现有的模态参数辨识方法在精度、鲁棒性、效率以及经济性能指标等方面仍然存在许多缺点，，在实际工程中的应用尚处于发展阶段，仍需进一地步深入研究和不断完善。本文针对环境激励下工程结构模态参数辨识方法进行了研究，并针对这些方法存在的不足进行了改进，进一步研发了模态参数辨识的软件系统。归结起来主要内容如下： ①在简要介绍工程结构健康监测、安全性评估的重要意义的基础上，详细论述了环境激励下工程结构模态参数识别的研究背景和意义。对环境激励下模态参数识别的主要方法进行了阐述，最后介绍了本文的主要工作内容。 ②针对EMD分解出现的端点效应问题和模态混叠问题，通过实例分别采用了基于支持向量回归机抑制端点效应的EMD方法和基于Hilbert特性的EMD模态解混叠方法。实例结果表明：本文所采用的改进方法能够有效抑制端点效应和消除模态混叠现象。进一步运用改进的EMD方法对四层钢框架试验进行了模态参数的辨识，证明了改进EMD方法能够正确识别结构模态参数。 ③针对信号预处理方法NExT法进行了改进，并将改进的NExT法和五种模态参数辨识方法（STD法、ITD法、复指数法、ARMA法和ERA法）相结合来识别结构的模态参数。通过同济大学十二层钢筋混凝土模型振动台试验数据，作者利用改进的NExT法和未改进的NExT法，开展了小震作用下和大震作用下的结构模态参数识别工作。结果表明：针对地震作用下结构模态参数识别的实例，改进的NExT法比未改进的NExT法具有更高的精度。 ④利用MATLAB下的GUI平台，研发了环境振动下结构模态参数辨识的可视化软件系统。包括文件读取模块，几何数据模块，振动数据模块（包含数据处理，例如滤波，奇异熵去噪等），处理分析模块（包含各种参数识别方法，例如随机子空间法、HHT法、STD法等），结果处理模块。基于所研发的软件系统，对数值算例钢桁架悬索桥在环境激励下模态参数问题进行了辨识，结果表明：整个辨识过程既清晰明了又准确方便。
[Abstract]:Modal identification and damage identification of engineering structures are the core technology and theoretical basis of health monitoring and safety assessment. It has become the focus and research focus in the field of civil engineering. In view of the large structure size, complex constraints, complex modeling and other characteristics. It is very difficult to carry out artificial excitation and effective measurement of the excitation signal of engineering structure, so the traditional modal parameter identification theory and method based on input and output signal is difficult to apply in engineering structure. The modal parameter identification method under environmental excitation has its own outstanding advantages: no artificial excitation, no need to measure the excitation signal, the measurement cost is low, the measurement does not affect the normal operation of the structure. More in line with the actual situation, has been widely used in engineering, but the existing modal parameter identification methods in the accuracy, robustness, efficiency and economic performance indicators still have many shortcomings. The application in practical engineering is still in the developing stage, which needs further research and improvement. In this paper, the identification methods of modal parameters of engineering structures under environmental excitation are studied, and the shortcomings of these methods are improved. The software system of modal parameter identification is further developed. 1 on the basis of brief introduction of the importance of engineering structure health monitoring and safety assessment. The research background and significance of modal parameter identification of engineering structures under environmental excitation are discussed in detail. The main methods of modal parameter identification under environmental excitation are described. Finally, the main work contents of this paper are introduced. 2 for the end point effect problem and modal aliasing problem of EMD decomposition. EMD method based on support vector regression machine to suppress endpoint effect and EMD mode de-aliasing method based on Hilbert characteristic are used by examples. The results show that:. The improved method in this paper can effectively suppress the end-point effect and eliminate the phenomenon of modal aliasing. Further, the improved EMD method is used to identify the modal parameters of the four-story steel frame test. It is proved that the improved EMD method can correctly identify the structural modal parameters. (3) the signal preprocessing method (NExT) is improved, and the improved NExT method and five modal parameter identification methods (NExT method and complex exponent method) are introduced. ARMA method and ERA method are combined to identify the modal parameters of the structure. The shaking table test data of 12 story reinforced concrete model of Tongji University are used. The author uses the improved NExT method and the unimproved NExT method. The identification of structural modal parameters under the action of small earthquakes and large earthquakes is carried out. The results show that the identification of structural modal parameters under earthquake action is an example. The improved NExT method has higher accuracy than the unimproved NExT method. 4 using the GUI platform under MATLAB, a visual software system for structural modal parameter identification under ambient vibration is developed, including file reading module and geometric data module. Vibration data module (including data processing, such as filtering, singular entropy denoising, etc.) processing and analysis module (including a variety of parameter identification methods, such as random subspace method / HHT / STD method). Based on the developed software system, the modal parameters of steel truss suspension bridge under ambient excitation are identified. The results show that the identification process is clear, accurate and convenient.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP311.52;TU317

【参考文献】