多目标进化算法和代理模型技术在气动稳健优化设计中的应用

摘要

According to the differences analysis compared with the traditional optimization method, here constructs an aerodynamic robust optimization design model aimed at the uncertainty influence of the flight environment parameter, and also demonstrates how both multi-objective evolutionary algorithm and surrogate model can be used to achieve robust optimal designs. Airfoil robust optimization example using the chosen technique has executed and then obtained the Pareto frontier of the result. The illustrated result shows that the robust optimization gets more stable performance and more advantage in aerodynamic design.%通过与传统优化方法的优缺点对比,提出针对飞行器环境参数存在不确定性影响时的气动稳健优化设计模型,并应用多目标进化算法和代理模型技术对稳健模型进行优化设计研究.提供了翼型稳健性优化实例,通过选择合适的翼型参数化方法和CFD求解程序以及不确定性分析方法,得到了稳健设计结果的Pareto前缘图.从其中选择的设计结果可以看出,稳健设计结果表现出更稳定的特性,证明了稳健设计的优势所在.