关键词: 多学科设计优化, 协同优化, 系统级一致性约束, 目标变量, 近似技术, 动态松弛, 均匀设计

Abstract: Collaborative optimization (CO) is an effective and non-hierarchical multidisciplinary design architecture that is well-suited to large-scale multidisciplinary design optimization (MDO). CO and approximate technology are combined to carry out a wing structure/aerodynamics design optimization. According to CO, how to solve the system-level optimization with compatibility equation constraints is the key problem. Before convergence, disagreement among disciplinary optimums is dynamic. Therefore, dynamic slack method is introduced. It regards the optimums’ discrepancy (almost 2 norm) among disciplinary as a slack factor, which can translate the system-level compatibility constraints into inequality constraints. By this way, CO is more robust. The CO mathematical model of wing design is established and the model is approximated using quadratic response surface based on uniform design. Then using CO based on dynamic slack method, the reasonable results are acquired. It indicates that CO combined with approximate technology can effectively solve complex engineering design optimization problem.

Key words: MDO, CO, system-level compatibility constraint, target variables, approximate technology, dynamic slave, uniform design