• CN:11-2187/TH
  • ISSN:0577-6686

›› 1998, Vol. 34 ›› Issue (5): 7-12.

• 论文 • 上一篇    下一篇

概率优化设计的一种新计算方法

陈立周;翁海珊   

  1. 北京科技大学机械工程学院
  • 发布日期:1998-09-01

NEW METHOD FOR PROBABILISTIC OPTIMAL DESIGN

Chen Lizhou;Weng Haishan   

  1. University of Science and Technology Beijing
  • Published:1998-09-01

摘要: 分析了概率约束的优化设计方法。提出了一种基于概率密度函数矩不等式的契贝雪夫点法。计算结果表明,该方法既保证了概率计算的精度,同时又克服了矩匹配法的不精确性和广义模拟优化方法计算量过大的缺点。效果是十分满意的。

关键词: 概率优化设计, 矩匹配法, 契贝雪夫点法, 随机约束

Abstract: The optimal design methods of probabilistic constraints is discussed. New method based on Chebyshev moment unequality of probability density function, so call Chebyshev point method, is given. The calculating result show that this approach improves both the accuracy of probability calculation and the robustness of optimum and may overcome the inaccuracy of moment matching method and significant processing time of generalized simulation optimization. The effect is quite satisfaction.

Key words: Moment matching method, Probabilistic optimal design, Stochastic constraint Chebyshev point method

中图分类号: