中文摘要：

在分析协同优化系统级优化的基础上，从保证其优化过程可行有解的角度，提出了自适应罚函数协同优化算法。采用罚函数法将学科间一致性约束条件下的系统级优化问题转化为无约束优化问题，利用学科间的不一致信息，构造动态罚因子的表达式。根据学科间的不一致性情况，对学科间一致性约束赋予相应的权值，从而在保证学科间一致性要求的前提下，使系统级目标函数达到最优。最后，利用典型算例对该方法进行了验证，结果表明该方法优化效率较高，且具有一定的收敛性和鲁棒性。

英文摘要：

In order to ensure the solvability of system-level optimization, a new adaptive penalty scheme for collaborative optimization was proposed based on the analysis of system-level optimization in collaborative optimization. The penalty function approach was applied to turn the system-level compatibility constrained optimization problem to a no-constrained optimization problem and the expression of dynamic penalty factor was determined by interdisciplinary discrepancy information. The interdisciplinary compatibility constrains were assigned weights according to the interdisciplinary discrepancy information. The optimal solutions of system-level objective function were obtained while ensuring the requirement of interdisciplinary consistency. Finally, two typical examples were adopted to test this optimization algorithm. The results show that the presented approach is efficient and has some convergence and robustness.