基于BLISS和PMA的多学科可靠性设计优化
收稿日期: 2012-10-29
修回日期: 2012-12-25
网络出版日期: 2012-12-29
基金资助
国家自然科学基金 (51175019,51205238);江苏省自然科学基金(BK20130999);江苏省普通高校自然科学研究资助项目(13KJB460012);中国博士后科学基金(230210235)
Multidisciplinary Reliability Design and Optimization Based on BLISS and PMA
Received date: 2012-10-29
Revised date: 2012-12-25
Online published: 2012-12-29
Supported by
National Natural Science Foundation of China (51175019,51205238);Natural Science Foundation of Jiangsu Province (BK20130999);Natural Science Foundation of Colleges and Universities in Jiangsu Province (13KJB460012);Postdoctoral Science Foundation of China (230210235)
针对多学科设计优化(MDO)中考虑不确定性而带来的计算复杂性问题,采用序列优化与可靠性评估(SORA)法的解耦思想,提出了基于两级集成系统综合(BLISS)策略和性能测量法(PMA)的多学科可靠性设计优化(RBMDO)方法。该方法将传统的3层嵌套循环的RBMDO方法过程解耦为顺序执行的确定性多学科设计优化(DMDO)和多学科可靠性分析(MRA),DMDO和MRA均采用BLISS策略进行计算,避免了每次多学科设计优化对完整可靠性分析模型的反复调用。飞机起落架缓冲器算例结果表明,文中方法的计算效率较其他两种方法分别提高了52.01%和26.51%,验证了方法的有效性。
关键词: 两级集成系统综合; 性能测量法; 多学科可靠性设计优化; 可靠性; 缓冲器
刘云平 , 张俊 , 张冰 , 孙瑾 . 基于BLISS和PMA的多学科可靠性设计优化[J]. 航空学报, 2013 , 34(10) : 2349 -2356 . DOI: 10.7527/S1000-6893.2013.0377
In order to solve the computational complexity caused by the consideration of uncertainties in multidisciplinary design optimization (MDO), an efficient method for reliability-based MDO (RBMDO) based on the bi-level integrated system synthesis (BLISS) and performance measure approach (PMA) is proposed. With the decoupling idea of sequential optimization and reliability assessment (SORA) method, the conventional triple nested loop of RBMDO is decoupled into a series of sequential execution of deterministic multidisciplinary design optimization (DMDO) and multidisciplinary reliability analysis (MRA). Both the DMDO and MRA are implemented by BLISS, which avoids the total computation of the whole multidisciplinary reliability analysis model iteratively in each RBMDO cycle. Finally, a shock absorber of landing gear example has been demonstrated to verify the efficiency of the proposed method, which shows that the efficiency of the proposed method has been improved by 52.01% and 26.51% respectively compared to the other two methods.
/
〈 | 〉 |