基于代理模型的翼型多目标集成优化设计  

Multi-Objective Integrated Optimization of Airfoil Based on Surrogate Model

作　　者：龙腾[1] 刘莉[1] 孟令涛[1] 朱华光[1] 

机构地区：[1]北京理工大学宇航学院,北京100081

出　　处：《系统仿真学报》2010年第7期1604-1608,共5页Journal of System Simulation

摘　　要：针对翼型优化中直接使用计算流体力学方法计算量大的问题,提出了一种基于代理模型的翼型多目标优化设计方法。使用形函数线性扰动法描述翼型外形,求解N-S方程获得翼型气动参数。采用Pareto多目标遗传算法进行优化,使用代理模型降低计算量,通过优化集成,实现优化过程自动化以提高设计效率。算例结果表明,本方法能有效降低计算成本,缩短设计周期,提高设计质量,具有一定实用价值。To reduce the computation burden of airfoil optimization with computational fluid dynamics method and numerical optimization technology,a multi-objective optimization approach for the airfoil was proposed.The airfoil geometry was formulated by linear perturbation method with shape functions,then,by solving N-S formulation the aerodynamic performance was obtained.Pareto Multi-objective genetic algorithm was employed in the optimization,and surrogate model was adopted to reduce computational cost.Due to optimization integration,automation of entire optimization procedure was achieved,which improved the design efficiency.The results of case studies demonstrate that this approach has reduced computation burden and shortened the design circle,which is also proved to be practicable in engineering.

关 键 词：翼型优化设计 多目标Pareto优化 代理模型 CFD MODEL CENTER 

分 类 号：V221[航空宇航科学与技术—飞行器设计]