改进EGO算法在跨声速翼型气动优化设计中的应用  

Application of the Improved EGO Algorithm in Transonic Airfoil Aerodynamic Optimization Design

作　　者：王红涛[1] 竺晓程[1] 杜朝辉[1] 

机构地区：[1]上海交通大学机械与动力工程学院,上海200240

出　　处：《上海交通大学学报》2009年第11期1832-1836,共5页Journal of Shanghai Jiaotong University

基　　金：国家自然科学基金(50576052);博士点基金(20060248036)资助项目

摘　　要：为了解决复杂工程优化问题计算量大的问题,提出了基于Kriging代理模型的改进EGO(Efficient Global Optimization)算法.采用小生境微种群遗传算法求解Kriging模型的相关向量,避免了模式搜索算法求解相关向量时对初始值的敏感性问题.采用小生境微种群遗传算法,结合无惩罚因子的惩罚函数法对EI(Expected Improvement)函数寻优,解决了惩罚因子难以选择的问题,增强了算法的鲁棒性.采用2个数值算例和1个工程算例对算法进行测试的结果表明,改进后的EGO算法收敛精度更高,比较适合在工程中应用.In order to decrease the computational expenses of complicated engineering optimization problem, the improved EGO（Efficient Global Optimization） algorithm based on Kriging surrogate model was proposed. Niching micro genetic algorithm instead of the pattern search algorithm was used to get the correlation vector of Kriging mode, which eliminates the dependence of correlation vector starting search points. Niching micro genetic algorithm coupled penalty funciton approach which does not require any penalty parameter was introduced to optimize the El（Expected Improvement） function. This method reduces the difficulty of finding appropriate penalty parameters and increases the robustness of the algorithm. The improved EGO algorithm was discussed and applied in two numerical and one engineering example. The test results show that the improved EGO algorithm is more efficient and accurate. It can be used in the complicated engineering optimization problems conveniently.

关 键 词：翼型 KRIGING模型 改进EGO算法 全局优化 

分 类 号：V211.41[航空宇航科学与技术—航空宇航推进理论与工程]