二维气动问题中Kriging代理模型精度影响因素

马洋; 张青斌; 韩启龙; 华明军; 李红霞

doi:10.13224/j.cnki.jasp.2016.11.014

二维气动问题中Kriging代理模型精度影响因素

doi: 10.13224/j.cnki.jasp.2016.11.014

1. 火箭军工程大学动力工程系, 西安 710025;
2. 国防科学技术大学航天科学与工程学院, 长沙 410073

基金项目:

第二炮兵工程大学科研基金青年项目（2015QNJJ034）；国防科学技术大学科学研究项目（JC13-01-04）

详细信息

作者简介:
马洋(1982-),男,湖南常德人,讲师,博士,主要从事飞行器外形优化设计研究.

中图分类号: V411.3
计量
- 文章访问数: 846
- HTML浏览量: 6
- PDF量: 553
- 被引次数: 0
出版历程
- 收稿日期: 2015-02-02
- 刊出日期: 2016-11-28

Factors influencing the accuracy of Kriging surrogate model in two-dimensional aerodynamic problem

1. Department of Power Engineering, Rocket Force University of Engineering, Xi'an 710025, China;
2. College of Aerospace Science and Engineering, National University of Defense Technology, Changsha 410073, China

摘要

摘要: 以二维跨声速临界翼型的阻力特性为对象，探讨样本点数目、Kriging代理模型参数及其类型等对模型精度的影响.阻力系数采用计算流体力学（CFD）方法得到.模型精度的验证采用交叉验证方法，采用平均误差、最大误差和标准交叉验证残差来衡量Kriging代理模型的精度.研究结果表明：①Kriging代理模型预测气动阻力效果较好.②模型精度随样本点的增多而提高，剔除与样本点响应趋势不相符的“奇异点”后，模型精度显著提高，平均误差减小5%~38%，最大误差减小13%~77%.③核函数类型对模型精度的影响最大，相关参数次之，回归模型的影响最小.采用高斯相关函数、2阶多项式回归模型，以及合适的相关参数值时，Kriging代理模型的精度最高.
- Kriging代理模型 /
- 精度验证 /
- 气动力 /
- 跨声速临界翼型 /
- 气动外形优化
Abstract: The factors influencing the accuracy of Kriging surrogate model including the number of sample points, the parameters of model and their types, were researched. The drag property of two-dimensional transonic airfoil was used to construct the surrogate model. The computational fluid dynamics (CFD) was employed to compute the drag coefficient. Three kinds of errors,i.e. average error, maximal error and standardized cross-validated residual were employed to measure the accuracy of the Kriging surrogate model while the cross validation was applied as the accuracy validation method. The results obtained are summarized as follows. First, the Kriging surrogate model performs well when predicting the aerodynamic drag of the two-dimensional transonic airfoil. Second, the accuracy of model improves with the increase of sample number, and when the ‘bizarre airfoil’ whose responses based on Kriging surrogate model are opposite with the normal ones are deleted, the accuracy of the model is improved obviously, and the average error and maximal error decrease 5%-38% and 13%-77% respectively. Third, the model accuracy is mainly affected by type of kernal function, followed by the correlation parameter, while the regression model has little influences. The Kriging surrogate model with Gauss correlation function, second order regression model and optimal correlation parameter has the best accuracy.
- Kriging surrogate model /
- accuracy validation /
- aerodynamic force /
- transonic airfoil /
- aerodynamic configuration optimization

HTML

参考文献(26)

[1]	Erguven V.Hybridizing genetic algorithm with artificial neural network in the aerodynamic optimization of the forward swept wing[R].AIAA-2010-2915,2010.
[2]	Griffin L W,Dorney D J,Huber F W,et al.Detailed aerodynamic design optimization of an RLV turbine[R].AIAA-2001-3397,2001.
[3]	Kirn Y,Kim J,Jeon Y,et al.Multidisciplinary aerodynamic-structural design optimization of supersonic fighter wing using response surface methodology[R].AIAA-2002-0322,2002.
[4]	Trizila P,Kang C,Visbal M,et al.A surrogate model approach in 2D versus 3D flapping wing aerodynamic analysis[R].AIAA-2008-5914,2008.
[5]	Tyan M,Park J,Kim S,et al.Design optimization of subsonic airfoil and slotted flap shape using multi-fidelity aerodynamic analysis[R].AIAA-2013-2704,2013.
[6]	张健.飞机多学科设计优化中的近似方法研究[D].西安:西北工业大学,2006. ZHANG Jian.Research of approximation methods in multidisciplinary design optimization of aircraft[D].Xi'an:Northwestern Polytechnical University,2006.(in Chinese)
[7]	Paiva R M,Carvalho A R D,Crawford C.Comparison of surrogate models in a multidisciplinary optimization framework for wing design[J].AIAA Journal,2010,48(5):995-1006.
[8]	Wang L,Beeson D,Wiggs G,et al.A comparison of meta-modeling methods using practical industry requirements[R].AIAA-2006-1811,2006.
[9]	Stander N,Roux W,Giger M,et al.A comparison of metamodeling techniques for crashworthiness optimization[R].AIAA-2004-4489,2004.
[10]	朱玉强.基于Kriging插值的反射面天线机电集成优化设计[D].西安:西安电子科技大学,2012. ZHU Yuqiang.Integrated structural-electromagnetic optimization of reflector antennas based on Kriging interpolation[D].Xi'an:Xi'an University of Electronic Science and Technology,2012.(in Chinese)
[11]	郭炳岐.基于Kriging方法的GPS高程拟合模型及其应用研究[D].西安:西安电子科技大学,2008. GUO Bingqi.Research on models of GPS height fitting based on Kriging method and their applications[D].Xi'an:Xi'an University of Electronic Science and Technology,2008.(in Chinese)
[12]	陈志民.基于Kriging模型的汽轮机基础可靠性动力优化方法[D].大连:大连理工大学,2007. CHEN Zhimin.Dynamic optimization of reliability for the turbine engine frame foundation based on Kriging model[D].Dalian:Dalian University of Technology,2007.(in Chinese)
[13]	罗霞.基于Kriging法的单索面斜拉桥静力可靠度分析[D].广州:华南理工大学,2012. LUO Xia.Static reliability analysis of single cable plane cable stayed bridge based on Kriging method[D].Guangzhou:South China University of Technology,2012.(in Chinese)
[14]	马洋,王丹丹,杨涛,等.基于代理模型的运载器头罩外形优化设计[J].航空动力学报,2014,29(1):192-198. MA Yang,WANG Dandan,YANG Tao,et al.Research on shroud configuration optimization design of transporter based on surrogate model[J].Journal of Aerospace Power,2014,29(1):192-198.(in Chinese)
[15]	Crowley D,Douglas R,Mavris D N,et al.Reducing the computational expense to train Euler-level aerodynamic surrogate models for a family of vehicle configurations[R].AIAA-2012-5282,2012.
[16]	Mehmani A,Zhang J,Chowdhury S,et al.Surrogate-based design optimization with adaptive sequential sampling[R].AIAA-2012-1527,2012.
[17]	Mehmani A,Chowdhury S,Messac A.A novel approach to simultaneous selection of surrogate models,constitutive kernels,and hyper-parameter values[R].AIAA-2014-1487,2014.
[18]	陈鹏,李剑,管涛.基于PSO的Kriging相关模型参数优化[J].微电子学与计算机,2009,26(4):178-181. CHEN Peng,LI Jian,GUAN Tao.The optimization of parameters of Kriging correlation model based on particle swarm optimization[J].Microelectronics and Computer,2009,26(4):178-181.(in Chinese)
[19]	Ahmed M Y M,Qin N.Metamodels for aerothermodynamic design optimization of hypersonic spiked blunt bodies[J].Aerospace Science and Technology,2010,14(1):364-376.
[20]	谢延敏.基于Kriging模型和灰色关联分析的板料成形工艺稳健优化设计研究[D].上海:上海交通大学,2007. XIE Yanmin.Research on robust optimization of sheet metal forming based on Kriging and grey relational analysis[D].Shanghai:Shanghai Jiao Tong University,2007.(in Chinese)
[21]	马洋.基于代理模型和MOEA/D的飞行器气动外形优化设计研究[D].长沙:国防科学技术大学,2015. MA Yang.Research on the aerodynamic shape optimization of vehicle based on surrogate model and MOEA/D[D].Changsha:National University of Defense Technology,2015.(in Chinese)
[22]	Jones D R,Schonlau M,Welch W J.Efficient global optimization of expensive black-box functions[J].Journal of Global Optimization,1998,13(1):455-492.
[23]	LIANG Haoquan,ZHU Ming,WU Zhe.Using cross-validation to design trend function in Kriging surrogate modeling[J].AIAA Journal,2014,52(10):2313-2327.
[24]	Kulfan B M.Universal parametric geometry representation method[J].Journal of Aircraft,2008,45(1):142-158.
[25]	李端琪,周华.基于CFD的斜置翼翼型选型研究[J].力学季刊,2013,34(3):361-366. LI Duanqi,ZHOU Hua.CFD-based studies on airfoil selection of oblique wing[J].Chinese Quarterly of Mechanics,2013,34(3):361-366.(in Chinese)
[26]	Keane A J.Wing optimization using design of experiment,response surface,and data fusion methods[J].Journal of Aircraft,2003,40(4):741-750.