A Unified Framework for Integrated Optimization Under Uncertainty
A Unified Framework for Integrated Optimization Under Uncertainty
复制标题
DOI：
10.1115/1.4001526
复制
发表时间：
2010-05
影响因子：
3.3
通讯作者：
Zhonglai Wang;Hongzhong Huang;Yu Liu
中科院分区：
文献类型：
--
作者：
Zhonglai Wang;Hongzhong Huang;Yu Liu
可靠性和稳健性是不确定性环境下设计的两个主要属性。因此，有必要在设计阶段将基于可靠性的设计与稳健设计相结合。提出了基于可靠性的设计与稳健设计相结合的统一框架。在该框架中，考虑到目标函数的概率特性，通过泰勒级数展开或逆可靠性策略将概率目标函数转换为确定性目标函数。因此，有了这个统一的框架，就不再需要处理基于可靠性的设计和稳健设计相结合的多目标优化问题。同时，采用逆可靠性策略将概率约束转化为确定性约束。为了解决这一统一框架，提出了一种改进的序贯优化可靠性评估方法。文中给出了三个例子来说明所提方法的优点。
Reliability and robustness are two main attributes of design under uncertainty. Hence, it is necessary to combine reliability-based design and robust design at the design stage. In this paper, a unified framework for integrating reliability-based design and robust design is proposed. In the proposed framework, the probabilistic objective function is converted to a deterministic objective function by the Taylor series expansion or inverse reliability strategy with accounting for the probabilistic characteristic of the objective function. Therefore, with this unified framework, there is no need to deal with a multiobjective optimization problem to integrate reliability-based design and robust design any more. The probabilistic constraints are converted to deterministic constraints with inverse reliability strategy at the same time. In order to solve the unified framework, an improved sequential optimization and reliability assessment method is proposed. Three examples are given to illustrate the benefits of the proposed methods.