节点文献

钢框架结构抗震可靠度的概率重要性分析

Probability importance analysis of seismic reliability for steel frame structures

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 吕大刚李晓鹏

【Author】 L Dagang,LI Xiaopeng(School of Civil Engineering,Harbin Institute of Technology,Harbin 150090,China)

【机构】 哈尔滨工业大学土木工程学院

【摘要】 概率重要度是一类特殊的参数灵敏度,对于结构可靠度设计、优化和评定等具有重要的价值。本文引进四类重要性测度,即重要性向量α、γ、δ和η。前两个测度分别描述标准正态空间内和原始空间内随机变量的本质特征和相对重要性程度以及可靠指标对设计点变化的灵敏性;后两个测度分别描述可靠指标对随机变量均值和标准差变化的灵敏性。采用基于FORM的有限元可靠度方法对钢框架结构进行抗震可靠度分析和概率重要性分析,以一个实际工程结构为例,分析了承载能力和变形能力极限状态抗震可靠度的概率重要性,结果表明:四类重要性测度的变化规律基本一致,将概率重要度很小的随机变量作为确定性变量处理,可以显著地提高大型复杂结构可靠度分析的计算效率。

【Abstract】 As a special type of parametric sensitivity,probability importance is very valuable for reliability design,optimization and evaluation for large complex structures.In this paper,four measures of probability importance are introduced,which include importance vectors α,γ,δ and η.Importance vectors α and γ character the degree of relative importance in standard normal space and original space respectively,while importance vectors δ and η are the sensitivities of reliability index with respect to the variations of mean values and standard deviations of random variables respectively.The seismic reliability and the corresponding probability importance of plane steel frames are analyzed by using the finite element reliability method based on FORM.By taking a practical engineering structure as an example,the probability importance of seismic reliability both in ultimate limit state and deformation limit state is computed.The analysis results demonstrate that the changing rule of the four measures of probability importance is consistent,and that replacing the unimportant random variables by deterministic values can greatly improve the computation efficiency of reliability analysis for large complex structures.

【基金】 国家自然科学基金重大研究计划(90715021);国家自然科学基金(50678057,50108005);地震行业科研专项(2008419073);教育部留学回国人员科研启动基金;黑龙江省留学归国人员基金(LC05C01);哈尔滨工业大学优秀青年教师基金
  • 【文献出处】 建筑结构学报 ,Journal of Building Structures , 编辑部邮箱 ,2007年S1期
  • 【分类号】TU391
  • 【被引频次】6
  • 【下载频次】441
节点文献中: 

本文链接的文献网络图示:

本文的引文网络