基于协同优化方法的多学科非概率可靠性优化设计

Multidisciplinary Non-probabilistic Reliability Optimization Based on Collaborative Optimization Method

针对传统多学科优化是基于确定性的优化设计方法,为了获得更为可靠的结构设计,在设计中考虑了不确定因素对结构性能的影响。本文将非概率可靠性优化方法与多学科优化设计方法相结合,引入了非概率可靠性指标,提出了一种新的基于协同优化方法的非概率可靠性多学科优化设计方法。同时采用了基于均匀试验设计的响应面方法来近似学科优化,提高了遗传算法的收敛速度,降低了计算量。最后通过对跨声速机翼气动结构多学科可靠性优化设计算例的计算,验证了本文提出的方法能够保证设计解的可靠性,对工程实践具有一定的指导意义。

The traditional multidisciplinary optimization design is based on deterministic optimization method. The influence of the uncertain factors on the structural performance is considered to obtain more reliable structural designs. A multidisciplinary non-probabilistic reliability optimization method based on collaborative optimization is presented by combining non-probabilistic reliability optimal design and multidisciplinary collaborative optimization with introducing non-probabilistic the reliability index. Meanwhile, the response surface approximation based on the design of experiment （DOE） is adopted in disciplinary level optimization. Thus, the efficiency of the genetic algorithm is improved. A numerical example for aerodynamic-structure multidisciplinary reliability optimization design of the transonic wing is used. Results illustrate that the method can ensure the design solution reliability .