基于统计分析的损伤识别方法及程序设计

Structure Damage Identification Method and Program Designing Based on Statistical Analysis

为实现大型复杂结构实时在线的健康监测,及时发现结构状态变化,文中基于主成分分析方法,以随机荷载作用下动态响应拟合的AR模型系数为分析对象,得出的前两阶主成分数据有效保留了结构状态的重要信息。应用椭圆控制图的基本原理,提出新的损伤特征指标ω,并将该损伤识别方法与LabVIEW相结合进行程序设计,在一钢框架模型上针对不同工况进行了试验验证,结果表明,该方法能够正确识别结构状态变化,所编软件运行可靠,并在损伤状态下进行了声光报警。

In order to realize the monitoring of the big and complex structure and to find the change of its condition, based on the PCA （principle component analysis） theory,taking AR （auto-regressive） model coefficient by the constant load as the object, the principle component data of the first two order effectively retain the important information of structure state. Utilizing the theory of ellipse control figure, a new damage feature factor ω is put forward. Then a program is written combined the damage identification theory and Lab VIEW. Finally, an experiment was conducted on a steel frame model under different conditions, the result shows that the method can iden tify the changes of the structure condition correctly, also the program works well and gives right sound and light alarm under the damage condition.