基于移动荷载作用下结构响应及小波分析的桥梁损伤诊断研究

发布时间：2018-03-15 00:10

本文选题：桥梁损伤识别　切入点：小波分析　出处：《北京交通大学》2014年博士论文　论文类型：学位论文

【摘要】：本文是基于国家重点基础研究发展计划973项目(项目编号：2013CB036203)和国家自然科学基金项目《基于车桥耦合振动分析的桥梁损伤诊断方法研究》(项目编号：51078029)的研究成果之一。桥梁结构在服役中会出现损伤病害从而影响其使用寿命和行车安全,这就使得研究损伤桥梁的动力特性以及对桥梁结构及时进行损伤诊断尤为必要。在传统的各种基于振动理论的无损检测技术中,存在着各种缺陷,如较难获得精准的测试数据,或是需要暂时中断交通等。如何对桥梁结构进行快速有效的损伤识别成为目前工程界关注的课题之一。本文研究了损伤桥梁的动力特性,用基于小波分析的方法识别出结构损伤,并进行了试验验证,主要研究内容和创新成果如下： (1)在比较了几种常用的损伤模型的基础上,以扭转弹簧模型模拟桥梁的裂缝损伤,用结构动力学运动方程对裂缝桥梁进行模态分析,研究了损伤桥梁的模态特性,如自振频率、振型等。理论推导了裂缝损伤梁在移动力和移动质量作用下的动力响应,通过数值算例,研究了在不同荷载大小和荷载速度等因素下损伤梁的动力特性。推导了移动质量和轮对-簧上质量两种模型下简支梁有载频率的数值求解方法。 (2)在介绍了小波理论的基础上,探讨了基于结构模态和连续小波变换的桥梁损伤识别方法。以损伤结构的位移模态为输入,对其进行连续小波变换,用小波系数灰度图和小波系数模极大值轨迹图识别出损伤位置,并利用可以评价信号奇异性特征的Lipschitz指数来评价损伤程度。考察了不同损伤程度、多损伤位置的识别、不同模态阶数、稀疏布置测点以及噪声的影响,比较了Euler梁和Timoshenko梁的损伤识别结果的异同。 (3)用位移互等定理和影响线理论证明了移动荷载作用下结构的位移时程在荷载通过损伤位置时存在奇异性。提出了基于移动荷载作用下结构响应的损伤识别方法。以移动荷载通过桥梁时梁上某一测点的位移响应时程为输入,对其进行连续小波变换,通过小波系数灰度图和小波系数模极大值轨迹图识别损伤的位置,并用Lipschitz指数评价损伤程度。通过数值算例研究了该损伤识别方法的有效性,考察了不同程度损伤、多位置损伤下的损伤识别,研究了不同测点位置、荷载速度以及荷载大小等因素对识别结果的影响。此外,用一个模型试验对该损伤识别方法进行验证,试验结果表明该方法具有很好的实用性。 (4)将曲率模态差法应用到识别空间结构的损伤中,提出了基于曲率模态差的系杆拱桥吊杆损伤识别方法。考察两个吊杆面内各个吊杆与拱桥主梁结合点,对其正对称和反对称竖向弯曲的曲率模态差进行分析,通过综合比对,识别出损伤吊杆,并考虑了一定程度噪声的影响。 (5)以曲率的概念和小波分析理论为基础,提出了小波包能量曲率差的损伤定位指标。以桥梁损伤前后各个节点在冲击振动下的加速度响应为输入,对其进行小波包分解,构造出小波包能量曲率差来进行损伤定位并研究了稀疏测点布置的影响。此外,利用一个实际的简支梁进行了试验验证,考察了小波函数和小波分解层数对损伤识别结果的影响。
[Abstract]:This paper is based on the national key basic research development plan 973 Project (project number: 2013CB036203) and the National Natural Science Foundation of China, which is one of the research results of bridge damage diagnosis method based on vehicle bridge coupled vibration analysis (item number: 51078029).
The bridge structure will be damaged and affect its service life and traffic safety in service, which makes the research on the dynamic characteristics of the bridge and the damage of the bridge structure damage diagnosis timely is necessary. In the traditional kinds of nondestructive testing based on the technology of vibration theory, the existence of various defects, such as it is difficult to obtain accurate test data, or need to temporarily interrupt traffic. How to bridge damage identification quickly and effectively has become one of the subject in engineering circles at present.
In this paper, the dynamic characteristics of damaged bridges are studied, and structural damage is identified based on wavelet analysis.
(1) based on comparing several damage models in common use, the torsion spring model to simulate the crack damage of the bridge, the modal analysis of the cracks of the bridge structure dynamic equation, the modal characteristics of the bridge damage, such as natural frequency, vibration mode and so on. The theory derived the power beam with crack damage force and moving mass response, through numerical example, study the dynamic characteristics of the beam damage factors of different loads and load speed. Deduced the moving mass and the wheel - sprung mass two models with numerical methods for solving the beam carrier rate.
(2) based on the theory of wavelet transform, discusses the structure mode and the bridge damage identification method based on continuous wavelet transform. The displacement modal for input, continuous wavelet transform of the wavelet coefficients and gray image wavelet coefficients modulus maxima trajectory identify damage location, and can use evaluation of singular signal feature of Lipschitz index to evaluate the degree of injury. The effects of different degree of injury, damage location identification, different modes, sparse arrangement of measuring points and the influence of the noise, and compares the identification results of Euler beam and Timoshenko beam.
(3) with reciprocal displacement theorem and the influence of line theory proves the existence of singularity under moving load structure displacement through the damage location in the load. The damage identification method of structure response under moving load. Based on the moving load through the bridge to a point on the beam displacement response time history as input, continuous wavelet transform is applied to the maximum trajectory identification of damage location by wavelet coefficients and wavelet coefficients of gray image, and use the Lipschitz index to evaluate the degree of injury. Through the numerical example to study the damage identification method is effective, the effects of different degree of damage, damage identification of damage location study on the different location, speed and load, impact load and other factors on the recognition result. In addition, to verify the method of damage identification using a model test, the test results show the party The law has good practicality.
(4) the modal curvature difference method is used to identify the spatial structure damage, put forward the tied arch bridge damage identification method based on curvature mode difference. The influences of two boom boom and girder arch bridge surface in each combination, on the symmetric and antisymmetric vertical bending curvature mode difference are analyzed, through the comprehensive comparison, identify the damage suspender, and considering the influence of noise.
(5) the concept and the theory of wavelet analysis of curvature as the foundation, proposed a damage orientation index of wavelet packet energy curvature difference. The bridge before and after the damage of each node under impact vibration acceleration as input, the wavelet packet decomposition, construct the difference to the damage location and influence on the sparse arrangement of measuring points the wavelet packet energy curvature. In addition, experiments are carried out using a simply supported beam on the wavelet function and wavelet decomposition effect of layers on the damage identification results.

【学位授予单位】：北京交通大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：U441.4;O174.2

【参考文献】