Abstract: To improve the efficiency of multidisciplinary design optimization with mixed random-interval uncertainty,a Bi-Level Integrated System Synthesis (BLISS)-based Reliability Based Multidisciplinary Design Optimization (RBMDO) method was proposed.An efficient strategy by integrating single-stage of reliability design optimization and response surface-based BLISS was used in this method,which decoupled nested loop RBMDO into two sequential modules include deterministic MDO and multidisciplinary reliability analysis.It could reduce the frequency of multidisciplinary reliability analysis and improve the efficiency of RBMDO.In MRA module,a serialized multidisciplinary reliability analysis method was proposed.An example was provided to verify the validity the method.

Key words: mixed uncertainty quantification, bi-level integrated system synthesis, multidisciplinary analysis, reliability analysis

摘要： 为了提高多源不确定下多学科设计优化的效率,提出一种基于两级一体化策略的多学科可靠性设计优化方法。该方法采用一种单级概率可靠性设计优化与基于响应面的BLISS进行集成的策略,将传统嵌套循环的确定性多学科设计优化和多学科可靠性分析进行解耦、形成顺序执行的两个模块,减少了多学科可靠性分析次数和多学科分析调用频率,提高了多学科可靠性设计优化的效率。在多学科可靠性分析模块,采用一种序列化方法进行概率和非概率多学科可靠性分析。最后通过一个实例验证了方法的有效性。

关键词: 混合不确定性量化, 两级一体化策略, 多学科优化设计, 可靠性分析