【作者】 何旭辉；

【导师】 陈政清； 黄方林；

【作者基本信息】 中南大学 ， 桥梁与隧道工程， 2004， 博士

【摘要】 大型桥梁的结构健康监测是目前桥梁工程界比较活跃的研究领域，同时也是一难度很大的课题，涉及不同的学科，国内外在这方面也都还处于探索阶段，没有更多的经验可供借鉴。本文首先对国内外桥梁结构健康监测研究现状进行了全面、系统的评述，然后以南京长江大桥为工程背景，对涉及既有铁路桥梁结构健康监测的一些关键问题进行了较为完整的研究。具体研究内容和取得的成果为： (1)首次在我国大型既有公铁两用桥上成功建立了监测内容全面、监测手段和方法可靠、比较先进的结构健康监测系统。解决了系统建设过程中许多具体的难题，并基于Visual Basic和MATLAB平台，自主开发了功能比较完善的主控软件和实时数据分析处理专用软件。目前该系统稳定、运行正常。 (2)监测信号的分析处理是桥梁结构健康监测的关键问题之一。利用小波分析理论，对大型桥梁结构健康监测信号进行多尺度分解，从Lipschitz指数出发，分析了小波去噪的原理，实现了实时监测信号的噪声剔除。同时利用小波包对监测信号进行分解，得到了监测信号能量分布特征向量，为我们同时在时域和频域分析结构状态提供了科学的依据。 (3)提出对于公铁两用桥梁，将结构健康监测振动信号通通视为非平稳信号来分析处理。并回顾和比较了几种常用的非平稳信号处理方法，指出希尔伯特—黄变换(HHT)方法用于非平稳信号处理的优越性。结合HHT与随机减量技术(RDT)，提出了一种新的模态参数识别方法，并应用于南京长江大桥结构模态参数的识别。 (4)根据灵敏度的物理意义，从简单实用的角度出发，充分利用现有的大型有限元分析程序，提出了以结构设计参数为修改参数，以结构自振特性为目标函数，以结构应力时程响应为验证参数的有限元模型修正方法。利用南京长江大桥结构健康监测信号，对该桥ANSYS有限元模型进行了修正，得到了能够表征大桥目前状态的结构有限元“基准”模型。中南大学博士学位论文摘要 (5)对铁路铆接钢桥的疲劳性能进行了较为详细的回顾和探讨，依据可靠度理论，推导了基于s一N曲线和线弹性断裂力学(LEFM)方法的结构疲劳破坏极限状态方程，对方程中各随机变量进行了分析。采用改进的一次二阶矩(ASM)法和Monte Carfo方法，对南京长江大桥各主要杆件的疲劳损伤及其可靠度指标和疲劳失效概率进行了分析。结果表明南京桥主要杆件的疲劳损伤度非常低，可能是由于杆件活载应力水平比较低的缘故。基于N口JLAB平台，编制了实时雨流计数和结构疲劳损伤可靠性分析专用程序。关键词:结构健康监测，小波分析，参数识别，模型修正，疲劳可 靠性分析更多还原

【Abstract】 The structural health monitoring (SHM) of a large bridge is an active research field of the bridge engineering. This is a very difficult task, which involves different subjects, and there is no more experience for reference because the study on this field in the word is still at the exploring phase. Firstly, state-of-arts of structural health monitoring is comprehensively reviewed. Then, taking the Nanjing Yangtze river bridge (NYRB) as engineering background, some key techniques about SHM for a railway bridge are investigated in this dissertation. The main contents are as following:(1) As an important part of this dissertation, the first SHM system for existing both highway and railway bridge has been built successfully in our country. The built system, which has the comprehensive monitoring contents and reliable monitoring means, is an advanced SHM system. Many difficult problems about the establishment of SHM system have been solved. Main software based on the Visual Basic platform and the real-time signal analysis program based on MATLAB platform are independently designed. At present, this system is working normally.(2) Signal analysis is a key technique for SHM. By using wavelet transform, the monitoring signals of bridge SHM are described on different scales. According to Lipschitz index, the principle of wavelet anti-noise is introduced, and noise elimination of the real-time monitoring signal is realized. At the same time, the eigenvector of monitoring signal energy obtained by wavelet packets can offer us the scientific basis for evaluating the structural state in both time domain and frequency domain.(3) Taking all the vibration monitoring signals of the bridge carrying both highway and railway traffic as non-stationary signals is firstly proposed. Combined the Hilbert-Huang transform (HHT)with random decrement technique (RDT), a new method of structural modal parameter identification in which the structural modal parameters of NYRB are identified is presented.(4) Based on the physical meaning of sensitivity and the existing finite element (FE) program, a simple and applied new method of model updating is proposed. In this method, the structural model is updated by modifying the parameters of design, and validated by structural natural vibration characteristics as well as stress-time response of the main members of NYRB. The FE model of NYRB built by ANSYS is updated by using the SHM signals, and the FE "baseline" model reflecting the structural current state is obtained.(5) The fatigue performance of railway riveted steel bridge is discussed in detail. According to the reliability theory, a limit-state function of fatigue failure for the S-N curve and Linear-elastic Fracture Mechanics (LEFM) methods are deduced, and the different random variables involved in the limit-state equation are discussed respectively. Using the advanced first-order second-moment (ASM) and Monte Carlo methods, the fatigue damage reliability and failure probability of the main members of NYRB are calculated. The results indicate that the fatigue failure probability of the NYRB main members is very small for its low live-load stress level.更多还原