一种运用少量传递率函数值与稀疏正则化的结构损伤识别方法  

Structural damage detection using few transmissibility functions and sparse regularization

作　　者：刘焕林 骆紫薇 余岭 LIU Huan-lin;LUO Zi-wei;YU Ling(School of Mechanics and Construction Engineering,Jinan University,Guangzhou 510632,China;MOE Key Lab of Disaster Forecast and Control in Engineering,Jinan University,Guangzhou 510632,China)

机构地区：[1]暨南大学力学与建筑工程学院,广东广州510632 [2]暨南大学重大工程灾害与控制教育部重点实验室,广东广州510632

出　　处：《振动工程学报》2020年第6期1181-1188,共8页Journal of Vibration Engineering

基　　金：国家自然科学基金资助项目(51678278,51278226)。

摘　　要：基于传递率函数的结构损伤识别方法因其无需结构荷载仅需结构响应信息的特点而具有良好的工程应用前景。但是,利用传递率函数进行损伤量化的大多数该类既有方法只能实现结构损伤定位,而不能有效估算结构损伤程度。针对该问题,从频响函数和传递率函数的定义出发,推导传递率函数与结构单元刚度折减系数的直接关系式,考虑结构损伤的稀疏特性并基于稀疏正则化技术,提出了一种运用少量传递率函数值与稀疏正则化的结构损伤识别新方法。数值仿真和实验结果表明:该方法仅需少量频率点对应的传递率函数值即可有效进行结构损伤识别,既能定位结构损伤又能估算损伤程度,具有识别精度高与强鲁棒性的特点。The transmissibility functions(TFs)-based SDD algorithms can avoid measuring input loads,employ the structural responses only and have great application prospects.However,most of the existing TFs-based SDD algorithms can only effectively locate structural damages but are lack of the ability of structural damage quantification.For this problem,a direct relationship between TFs and elemental stiffness damage indexes is established based on definitions of both frequency response functions(FRFs)and TFs.After the structural damage sparsity and the sparse regularization are introduced,a new SDD method is proposed by using both few values of transmissibility functions in frequency domain and sparse regularization.The numerical simulation and experimental results show that the novel SDD method can be effectively used for the SDD problems through adopting few values of TFs.It can not only locate structural damages but also quantify structural damage severity with a higher SDD accuracy and strong robustness to measurement noises.

关 键 词：结构健康监测 结构损伤识别 传递率函数 稀疏正则化 未知激励 

分 类 号：O327[理学—一般力学与力学基础] TU311[理学—力学]