摘要
在结构损伤的系统识别中 ,损伤是用结构参数的减小来反映的。由于模型误差和试验测试误差 ,并不是所有的结构参数减小反映损伤的存在。为此 ,提出的损伤区域的判别分析法将每一个可能的损伤状态定义为一个随机总体 ,通过比较样本到每一个总体的距离来确定样品与哪个总体最靠近 ,进而对损伤区域作出判断。这一方法可以避免确定损伤单元数量时的主观判断 ,减小误判的风险。
In the system identification of a structural damage,the damage is simulated by a reduction in the structural parameter.Because of the errors of a model and the test measurements,not all the reductions in the structural parameters reflect damage.In the distinguishing analysis method for deducing the location of the damage developed in this paper,every possible damage state is defined as a random collectivity,a sample is most near to which collectivity is determined by comparing the distance of the sample to every collectivity and the damage region is deduced.The subjective judgement can be avoided in determining the number of the damaged elements and the risk of miscarriage of justice can be reduced by use of this method.
出处
《工业建筑》
CSCD
北大核心
2000年第4期61-63,30,共4页
Industrial Construction
基金
江苏省自然科学基金
关键词
判别分析
损伤位置
系统识别
结构损伤
损伤区域
distinguishing analysis\ location of damage\ random collectivity\ system identification\ system response
