基于小波的结构损伤检测.doc

西安建筑科技大学硕士论文 PAGE  PAGE I 基于小波分析的结构损伤检测 专 业：防灾减灾工程及防护工程 硕 士 生：张 宇 指导教师：胡卫兵 教授 摘 要 大型结构在长期使用中会因各种因素而产生损伤和退化，如果没有被及时发现和采取必要的措施，将可能发生严重的事故，造成生命和财产的巨大损失。因此，对结构的健康状况进行适时的评估变得尤为重要，而损伤识别则是结构健康检测中的一项关键技术。在土木、机械以及航空工程等领域，利用结构动力特性的改变对损伤进行识别，已经成为一个受到广泛关注的研究方向。 本文对多种基于结构动力特性的损伤识别方法进行了分析和评述，分别用模态分析和瞬态分析对结构的损伤问题进行了探讨，随后选择用小波变换分解结构动力响应信号得到的小波包能量谱来研究结构的损伤检测问题。 首先，建立结构的有限元模型，利用ANSYS软件研究了因损伤而引起的结构的模态参数的变化，并求出其动力响应。 其次，对结构模型进行瞬态分析。假定给结构突然施加脉冲荷载，得到结构的动力学响应信号，然后利用MATLAB中小波分析将动力响应信号分解得到子信号，求出各阶子信号的能量谱，再得到结构损伤前后的能量谱的变化量。 最后，通过比较损伤前后小波分解子信号的差异和能量谱的变化，列出由于损伤程度和位置不同，结构模态参数和小波子信号能量谱所产生的变化，以用于结构的在线损伤检测。 本文通过对一大跨度桥梁结构的损伤进行了分析，结果表明基于小波分析的结构损伤检测是可行的。 关键词：结构，小波分析，损伤检测 PAGE 61 Damage Detection for Structure Based on Wavelet Analysis Specialty：Disaster Prevention and Reduction Engineering, Protective Engineering Name：Zhang Yu Adviser：Prof. Hu Weibing ABSTRACT Large structural performance degradation will inevitably occur because of various factors as well as structural damage, which will result in structural collapse and great loss once not found in time, appears in the structure. Consequently, the structure damage identification using the changes of their vibration characteristics has become a hot research topic throughout the civil, mechanical and aerospace engineering communities. Various damage identification methods based on changes of vibration characteristics are summarized. In the dissertation, crack damage detection for a bridge structure on the basis of structural vibration characteristics is studied by using modal analysis and transient dynamic analysis, then, the wavelet packet energy spectrum of decomposed dynamic response signal by wavelet transform is chosen so as to study the structure. Firstly, the finite element dynamic model of structure is established. Variations of modal parameters due to structural cracks are analyzed according to ANSYS. Meanwhile, the structural dynamic response signals are calculated. Secondly, transient dynamic analysis is used for structure m