结构可靠性优化设计的证据理论和微分演化方法  

A Method of Reliability Design Optimization Using Evidence Theory and Differential Evolution

作　　者：唐和生[1,2] 苏瑜[1] 薛松涛[1] 邓立新[1] 

机构地区：[1]同济大学结构工程与防灾研究所,上海200092 [2]同济大学土木工程防灾国家重点实验室,上海200092

出　　处：《湖南大学学报（自然科学版）》2014年第4期33-38,共6页Journal of Hunan University:Natural Sciences

基　　金：国家自然科学基金资助项目(51178337;50708076);科技部国家重点实验室基础研究项目(SLDRCE11-B-01);同济大学土木工程学院光华基金资助项目

摘　　要：采用证据理论作为传统概率的替代方法处理不精确的数据信息,提出了基于证据理论的可靠性优化设计方法.该方法针对给定的失效概率许用值Pf,通过计算证据理论的不确定测度Pl(F),以Pl(F)<Pf作为可靠性约束条件的替代模型进行可靠性优化设计.为了降低基于证据理论不确定量化分析的计算成本,引入微分演化优化算法计算区间边界值.以典型桁架结构形状优化问题为例,验证了所提出方法的准确性和有效性.A new method of reliability design optimization using the evidence theory was proposed. Ev- idence theory was presented as an alternative to the classical probability theory to handle the imprecise data situation. The plausibility measure PI（F） based on evidence theory, with PI（F）〈Pf for a preconcerted failure probability Pf, was used as a surrogate modal of reliability constraint in the reliability design opti- mization. In order to alleviate computational difficulties in the evidence theory based uncertainty quantifi- cation analysis, a differential evolution based interval optimization for computing bounds method was de- veloped. A typical truss structure shape optimization with epistemic uncertainties was investigated to dem- onstrate the accuracy and efficiency of the proposed method.

关 键 词：证据理论 微分演化 可靠性优化设计 形状优化 

分 类 号：TU318[建筑科学—结构工程] TU323.4