基于小波法的独塔钢箱斜拉桥健康状态监测研究

发布时间：2018-03-09 11:15

本文选题：健康监测　切入点：小波变换　出处：《兰州交通大学》2015年硕士论文　论文类型：学位论文

【摘要】：小波分析被誉为“信号显微镜”,基于小波变换的多尺度分析和奇异点检测以及小波包节点能量为处理实时监测信号提供了方便又快捷的工具。本文从国内外研究现状入手,分析了桥梁健康监测中存在的问题,并对研究背景的长期健康监测系统做了简单介绍;之后对小波分析理论进行介绍,并在工程背景和数值模拟相结合的基础上,系统地研究了小波法在桥梁长期健康监测中的应用,重点论述了基于小波变换的实测信号的多尺度分析和阈值去噪分析,以及小波包能量理论和小波奇异性检测理论在独塔钢箱斜拉桥损伤识别和定位中的应用;最后对理论研究成果应用到工程实践中存在的问题给出改进意见。主要的研究内容如下:1)将小波多尺度分析和阈值去噪相结合,利用真实信号和噪声信号在小波变换不同尺度下的差异性表现,实现监测信号的阈值去噪。同时分别进行了实测信号与去噪信号的幅值对比以及采用快速傅立叶变换后实测信号与去噪信号的幅值对比,并且对实测信号采用不同阈值规则时的表现效果进行对比,讨论了阈值规则选取的基本条件。利用小波理论进行多尺度分析时,只需要对采集的加速度信号导入MATLAB软件中,并编制相应程序即能表达损伤特征,此时该位置在某个尺度时刻会出现尖点。2)将小波分解与小波包能量理论相结合,用来实现桥梁结构损伤识别。通过Midas Civil软件建立通海路斜拉桥有限元模型并进行时程分析,分别提取单一损伤和多损伤时主梁节点的加速度信号,对该信号进行小波包能量分析,得到未损和索力变化各种工况下小波包节点能量变化值。通过对比小波包频段能量百分比可得出小波包能量变化对结构微小索力变化有很高的灵敏度,由于只需要求损伤工况的时程加速度信号,故而十分有利于桥梁结构的实时在线监测,并且能够对部分受损桥梁的及时修复提供重要的参考建议。3)将小波变换与曲率模态理论相结合,用于实现桥梁结构损伤定位。通过Midas Civil软件建立通海路斜拉桥有限元模型并进行振型分析,提取主梁第一阶和第二阶位移模态,然后利用中心差分法得到其曲率模态,再对不同曲率模态曲线进行小波变换。通过主梁刚度折减各工况的第一阶和第二阶曲率模态图形可以很容易看出损伤发生位置,不但对主梁单元的细微损伤位置能够得到很好的识别定位效果,而且对斜拉索的损伤也能够通过主梁单元的曲率模态信号展现出来,这对于将小波奇异性理论应用于各种不同类型桥梁的长期健康监测中有极大地推动作用。
[Abstract]:Wavelet analysis is praised as "signal microscope". Multi-scale analysis based on wavelet transform, singular point detection and wavelet packet node energy provide a convenient and fast tool for processing real-time monitoring signal. The existing problems in bridge health monitoring are analyzed, and the long term health monitoring system based on the research background is briefly introduced. Then, the wavelet analysis theory is introduced, and based on the combination of engineering background and numerical simulation, The application of wavelet method in bridge long-term health monitoring is systematically studied, and the multi-scale analysis and threshold de-noising analysis of measured signals based on wavelet transform are emphasized. And the application of wavelet packet energy theory and wavelet singularity detection theory in the damage identification and location of single-tower cable-stayed bridge with steel box; Finally, the problems existing in the application of theoretical research results to engineering practice are improved. The main research contents are as follows: 1) combining wavelet multiscale analysis with threshold denoising. Using the difference of real signal and noise signal in different scales of wavelet transform, At the same time, the amplitude comparison between the measured signal and the de-noised signal and the amplitude comparison between the measured signal and the de-noised signal after fast Fourier transform are carried out. The performance of measured signals using different threshold rules is compared, and the basic conditions of threshold rule selection are discussed. When wavelet theory is used for multi-scale analysis, only the collected acceleration signals need to be imported into MATLAB software. The damage characteristics can be expressed by making a corresponding program. At this time, the cusp will appear at a certain scale. 2) the wavelet decomposition will be combined with the wavelet packet energy theory. The finite element model of the cable-stayed bridge along Tonghai Road is established by Midas Civil software and the time history analysis is carried out to extract the acceleration signal of the main beam node in the case of single damage and multiple damage respectively. The wavelet packet energy of the signal is analyzed. By comparing the energy percentage of wavelet packet frequency band, it can be concluded that the variation of wavelet packet energy has a high sensitivity to the change of minimal cable force of the structure. Because only the time-history acceleration signal of damage condition is needed, it is very helpful for the real-time on-line monitoring of bridge structure. And it can provide important reference suggestion for repairing some damaged bridges in time. 3) combine wavelet transform with curvature mode theory. The finite element model of the cable-stayed bridge along Tonghai Road is established by Midas Civil software, and the vibration mode of the cable-stayed bridge is analyzed. The first and second order displacement modes of the main beam are extracted, and the curvature mode of the bridge is obtained by using the central difference method. Then wavelet transform is carried out on different curvature mode curves. The damage location can be easily seen by the first and second order curvature mode patterns of the main beam stiffness reduction. Not only can the location of the minor damage of the main beam element be well identified, but also the damage to the stay cable can be displayed by the curvature mode signal of the main beam element. This will greatly promote the application of wavelet singularity theory to the long term health monitoring of different types of bridges.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U446

【参考文献】