• 更多栏目

    易平

    • 教授     博士生导师   硕士生导师
    • 性别:女
    • 毕业院校:大连理工大学
    • 学位:博士
    • 所在单位:土木工程系
    • 学科:结构工程. 防灾减灾工程及防护工程
    • 办公地点:大连理工大学综合实验三号楼524
    • 联系方式:yiping@dlut.edu.cn
    • 电子邮箱:yiping@dlut.edu.cn

    访问量:

    开通时间:..

    最后更新时间:..

    Discussion of mathematical models of probabilistic constraints calculation in reliability-based design optimization

    点击次数:

    论文类型:会议论文

    第一作者:Yi P.

    发表时间:2011-06-18

    收录刊物:EI、Scopus

    卷号:243-249

    页面范围:5717-5726

    摘要:In a reliability-based design optimization (RBDO) problem, most of the computations are used for probabilistic constraints assessment, i.e., reliability analysis. Therefore, the effectiveness, especially the correctness of the reliability analysis is very important. If the probabilistic constraint is misjudged, the optimization iteration would have convergence problems or arrive at erratic solutions. The probabilistic constraint assessment can be carried out using either the conventional reliability index approach (RIA) or the performance measure approach (PMA). In this paper, the mathematical models to calculate the reliability index in RIA and to calculate the probabilistic performance measure (PPM) in PMA are discussed. In RIA, through estimating whether the mean-value point in safe domain or not, we should use a positive or negative reliability index respectively. In PMA, one should always minimize the performance measure to compute PPM whether the performance measure at the mean-value point is positive or negative, which puts right the wrong mathematical model in some literatures and makes it possible to produce effective and efficient approach for RBDO. ? (2011) Trans Tech Publications.