改进Kriging模型在翼型气动优化设计中的应用研究  

Application of Improved Kriging-Model-Based Optimization Method in Airfoil Aerodynamic Design

作　　者：许瑞飞[1] 宋文萍[1] 韩忠华[1] 

机构地区：[1]西北工业大学翼型叶栅空气动力学国防科技重点实验室,陕西西安710072

出　　处：《西北工业大学学报》2010年第4期503-510,共8页Journal of Northwestern Polytechnical University

基　　金：国家自然科学基金(90605004)资助

摘　　要：文章将拉丁超立方抽样试验设计方法,Kriging模型与优化方法结合发展了一套基于改进Kriging模型的优化设计算法。首先对Kriging模型的拟合精度进行了研究,针对一维函数以及气动问题进行了测试,验证了文中发展的Kriging模型能够较好地对函数进行拟合以及能够运用到气动问题中。为了提高Kriging模型的拟合精度,文章对原有的Kriging模型优化设计算法进行了改进,在引入EI方法的同时将优化搜索算法得到的最优点加入到原有的样本集中,每次增加2个样本点,然后再重新生成Kriging模型进行优化设计。为了验证该方法的正确性和有效性,使用该算法进行了RAE2822翼型减阻优化设计,优化后的RAE2822翼型的阻力系数降低了33.6%,算例表明采用该方法能够逐步提高Kriging模型的拟合精度,最终实现高效的翼型气动性能优化设计。An improved Kriging-model-based optimization design method is presented.Such a method is based on design of experiment of Latin hypercube sampling,Kriging model and a genetic optimization algorithm.By simultaneously adding the sample point with maximum expected improvement（EI） and the optimal point from optimization of the initial samples,a new Kriging model of higher accuracy can be formed gradually.The fitting accuracy of Kriging model based on EI method is investigated and validated through the tests of a one-dimensional function and an aerodynamic problem.Test results show that the developed Kriging model is effective for optimization problems.Finally,a drag reduction optimization design of RAE2822 airfoil is carried out for examining the validity and efficiency of our method.The drag coefficient of RAE2822 airfoil is reduced by 33.6%.It shows that this method can gradually improve the fitting accuracy of the Kriging model,thus greatly improve the aerodynamic performance of the airfoil.

关 键 词：拉丁超立方抽样 KRIGING模型 EI方法 优化设计 

分 类 号：V211[航空宇航科学与技术—航空宇航推进理论与工程]