摘要
为使用随机子空间方法对悬臂结构进行损伤判断和定位,对振动测量信号,利用数据驱动随机子空间识别方法得到随机状态空间模型,通过变换矩阵T、T_1将离散状态空间矩阵A转化为特定形式,克服A矩阵的多样性问题。T矩阵采用A矩阵特征值分解后的特征向量矩阵构造,变换后的A矩阵用于判断结构是否存在损伤及相对损伤程度。T_1矩阵采用A矩阵和离散输出矩阵C构造,变换后的A矩阵包含测点信息,用于结构损伤定位。损伤指标采用统计方法确定,由损伤前后样本马氏距离计算得到。通过一三自由度悬臂结构仿真算例和一座模型斜拉桥桥塔损伤试验,验证了该方法的有效性。
In order to judge and locate the damage of cantilever structures by using the stochastic subspace identification（ SSI） method, a data-driven SSI algorithm was proposed to identify the state matrix of vibration measurement signals. In order to consider the variousness of the state matrix A,two transformation matrixes T and T1 were built,so that the matrix A in different states could be converted to a specific state by using matrix T or T1. The matrix T was constructed by the feature vector matrix of eigenvalue decomposition of matrix A. The transformed matrix A was used to judge whether the structure was damaged or not and how the relative damage degree was. The matrix T1 was constructed by utilizing the matrix A and discrete output matrix C. The transformed matrix A contains measuring point informations,which can be used for structural damage localization. The damage index was determined by the statistical method,which was calculated according to the Mahalanobis distance of specimens before and after damage. The efficiency of the method was verified by a simulated cantilever structure with three degrees of freedom and the tower damage experiment of a model cable-stayed bridge.
作者
唐盛华
方志
张国刚
TANG Shenghua;FANG Zhi;ZHANG Guogang(College of Civil Engineering and Mechanics,Xiangtan University,Xiangtan 411105,China;College of Civil Engineering,Hunan University,Changsha410082,China;Hunan Provincial Conmmnications Planning,Survey ＆ Design Institute,Changsha 410008,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2018年第14期141-148,共8页
Journal of Vibration and Shock
基金
国家自然科学基金(51508488)
"工程结构动力学与可靠性分析湖南省高等学校重点实验室(湘潭大学)"开放课题资助(KLDRES002)
关键词
随机子空间
损伤
状态空间模型
悬臂结构
stochastic subspace
damage
state space model
cantilever structure
