机器学习方法在气动特性建模中的应用

何磊; 钱炜祺; 汪清; 陈海; 杨俊

doi:10.7638/kqdlxxb-2019.0033

欢迎来到能源电力期刊网！

English Version 期刊登录

您当前的位置：

首页 >

文章列表页 >

机器学习方法在气动特性建模中的应用

专栏——智能空气动力学研究探索 | 更新时间：2026-05-28

- 机器学习方法在气动特性建模中的应用
- Applications of machine learning for aerodynamic characteristics modeling
- 空气动力学学报 2019年37卷第3期页码：470-479
- 作者机构：
  
  中国空气动力研究与发展中心,绵阳,621000
- 作者简介：
  
  何磊(1988-), 男, 四川绵竹人, 助理研究员, 研究方向:气动建模与仿真.E-mail:helei_email@163.com
  钱炜祺(1973-), 男, 江苏无锡人, 博士, 研究员, 研究方向:气动参数辨识与数据分析.E-mail:qwqhyy@sina.com
- 基金信息：
  
  国家自然科学基金（11802325）
- DOI：10.7638/kqdlxxb-2019.0033
  中图分类号： TP181;V211
- 纸质出版：2019
- 稿件说明：
移动端阅览
何磊, 钱炜祺, 汪清, 陈海, 杨俊. 机器学习方法在气动特性建模中的应用[J]. 空气动力学学报, 2019, 37(3): 470-479.

HE Lei, QIAN Weiqi, WANG Qing, et al. Applications of machine learning for aerodynamic characteristics modeling[J]. 2019, 37(3): 470-479.
何磊, 钱炜祺, 汪清, 陈海, 杨俊. 机器学习方法在气动特性建模中的应用[J]. 空气动力学学报, 2019, 37(3): 470-479. DOI： 10.7638/kqdlxxb-2019.0033.

HE Lei, QIAN Weiqi, WANG Qing, et al. Applications of machine learning for aerodynamic characteristics modeling[J]. 2019, 37(3): 470-479. DOI： 10.7638/kqdlxxb-2019.0033.

摘要

气动数据建模是飞行性能仿真评估的基础。气动特性建模主要有机理建模方法和"黑箱"建模方法。本文对"黑箱"建模的三类机器学习方法——分类与回归树方法、浅层学习方法和深度学习方法，进行了算法说明与分析应用。将分类与回归树方法、浅层学习方法中的Kriging建模方法、RBF神经网络方法及SVM支持向量机方法分别应用于火箭气动特性建模、三角翼大迎角非定常气动特性建模、气动热试验数据融合，对这几类建模方法的优势和不足进行了比较分析。同时，将流动条件参数组成向量，再映射为图像，与翼型图像构成"合成图像"，建立了基于翼型几何图像、来流马赫数、迎角的翼型气动特性深度神经网络模型，得到了比较好的预测效果，拓展了气动特性深度学习建模方法的使用范围。

Abstract

Mathematical modeling of aerodynamic data plays an important role in the performance evaluation of a flight vehicle. There are two kinds of aerodynamic characteristics modeling methods

i.e.

the rational modeling method based on physical mechanism and the "black-box" modeling methods. In this paper

three types of "black-box" modeling methods including the classification and regression tree method(CART)

the shallow learning method

and the deep learning method are investigated. The CART method and three shallow learning methods including Kriging method

radial basis function(RBF) neural network method

and support vector machines(SVM) method are applied to the aerodynamic data modeling of rocket

the unsteady aerodynamic characteristics modeling for delta wing with large angle of attack

the aerodynamic heat data fusion of wind tunnel test and CFD. The advantage and shortage of these modeling methods are compared. A deep learning neural network model of airfoil's aerodynamic coefficients is built. It considers the influence of airfoil picture and flow parameters such as Mach number

angle of attack on the prediction results at the same time through a "synthetic picture".It could significantly expand the applied range of the deep learning modeling method in aerodynamic research field.

关键词

Keywords

references

GRAUER J A, MORELLI E A. A generic nonlinear aerodynamic model for aircraft[R]. AIAA 2014-0542.

GOMAN M, KHRABROV A. State space representation of aerodynamic characteristics of an aircraft at high angles of attack[R]. AIAA 1992-4651.

WANG Q, QIAN W Q, DING D. A review of unsteady aerodynamic modeling of aircrafts at high angles of attack[J]. Acta Aeronautica et Astronautica Sinica, 2016, 37(8):2331-2347. (in Chinese)

HUANG D, ZHENG W X. Unsteady aerodynamic modeling method using differential equations based on wind tunnel test[J]. Journal of Nanjing University of Aeronautics Astronautics, 2014, 46(4):599-602. (in Chinese)

HE K F, WANG W Z, QIAN W Q. Mathematic modeling for the missile aerodynamics with tail-wing according to wind-tunnel test results[J]. Experiments and Measurements in Fluid Dynamics, 2004, 18(4):62-66. (in Chinese)

HE K F, QIAN W Q, ZHANG Y, et al. Application of partial least squares regression method in modeling of aerodynamic data[J]. Journal of Astronautics, 2014, 35(3):277-282. (in Chinese)

FU J M. Three dimensional aerodynamic mathematical model for tactical missiles with jet steering[J]. Aerospace Shanghai, 2005, (4):13-18. (in Chinese)

SANTOS M C, MATTOS B S, GIRARDI R M. Aerodynamic coefficient prediction of airfoils using neural networks[R]. AIAA 2008-0887, 2008.

WANG Z J, LAN C E, BRANDON J M. Fuzzy logic modeling of lateral-directional unsteady aerodynamics[R]. AIAA 99-4012.

SHI Z W, MING X. The application of FNN in unsteady aerodynamics modeling based on fuzzy clustering[J]. Acta Aerodynamica Sinica, 2005, 23(1):21-24. (in Chinese)

LIU X L, ZHANG K L, GUO B. Research on Gaussian process surrogate model and its application to rocket aerodynamic analysis[J]. Journal of Solid Rocket Technology, 2010, 33(5):486-490. (in Chinese)

ZHANG T J, QIAN W Q, ZHOU Y, et al. Preliminary Thoughts on the combination of artificial intelligence and aerodynamics[J]. Advances in Aeronautical Science and Engineering, 2019, 10(1):1-11. (in Chinese)

ZHANG S L. An introduction to the methodology of CART-Classification and regression trees[J]. Volcanology Mineral Resources, 1997, 18(1):67-75. (in Chinese)

CHATI Y S, BALAKRISHNAN H. Statistic modeling of aircraft engine fuel flow rate[R]. ICAS-2016-0619.

YANG J. Preliminary study of methods using in modeling and data mining for aerodynamic data[D]. Mianyang: China Aerodynamics Research and Development Center, 2016. (in Chinese)

AUMAN L. Missile DATCOM user's manual-2008 revision[R]. AFRL-RB-WP-TR-2009-3015, 2008.

QIAN W Q, YANG J, ZHOU Y. A comparative study of mathematical modeling methods for rocket aerodynamic data[C]//7th European Conference For Aeronautics And Space Sciences. Milan, Italy, EUCASS-100, 2017.

SU W, GAO Z H, ZUO Y T. Application of RBF neural network ensemble to aerodynamic optimization[R]. AIAA 2008-0154.

SIMPSON T W, et al. Kriging models for global approximation in simulation-based multidisciplinary design optimization[J]. AIAA Journal, 2001, 39(12):2233-2241.

RASMUSSEN C E, WILLIAMS C K I. Gaussian process for machine learning[M]. Cambridge:The MIT Press, 2006.

ANDRÉS E, SALCEDO-SANZ S, MONGE F, et al. Efficient aerodynamic design through evolutionary programming and support vector[J]. Expert Systems with Applications, 2012, 39(12):10700-10708.

WANG Q, QIAN W Q, HE K F. Unsteady aerodynamic modeling at high angles of attack using support vector machines[J]. Chinese Journal of Aeronautics, 2015, 28(3):659-668.

WANG K Q, YANG S C, DAI T H, et al. Method of optimizing parameter of least squares support vector machines by genetic algorithm[J]. Computer Application and Software, 2009, 26(7):109-111. (in Chinese)

JARRAH M A, ASHLEY H. Impact of flow unsteadiness on maneuvers and loads of agile aircraft[R]. AIAA 89-1282.

HINTON G E, SALAKHUTDINOV R R. Reducing the dimensionality of data with neural networks[J]. Science, 2006, 313(5786):504-507.

YIN B C, WANG W T, WANG L C. Review of deep learning[J]. Journal of Beijing University of Technology, 2015, 41(1):48-59. (in Chinese)

KUTZ J N. Deep learning in fluid dynamics[J]. Journal of Fluid Mechanics, 2017, 814:1-4.

LING J, KURZAWSKI A, TEMPLETON J. Reynolds averaged turbulence modeling using deep neural networks with embedded invariance[J]. Journal of Fluid Mechanics, 2016, 807:155-166.

GUO X X, LI W, LORIO F. Convolutional neural networks for steady flow approximation[C]//ACM SigKDD International Conference. San Francisco, CA, 2016: 481-490.

HUANG J J, DUAN L, WANG J X, et al. High-mach-number turbulence modeling using machine learning and direct numerical simulation database[R]. AIAA 2017-0315.

XIAO H, WU J L, WANG J X, et al. Physics informed machine learning for predictive turbulence modeling: Progress and perspectives[R]. AIAA 2017-1712.

YILMAZ E, GERMAN B. A convolutional neural network approach to training predictors for airfoil performance[R]. AIAA 2017-3660.

CHEN H, QIAN W Q, HE L. Aerodynamic coefficient prediction of airfoils based on deep learning[J]. Acta Aerodynamica Sinica, 2018, 36(2):294-299. (in Chinese)

ZHANG Y, SUNG W J, MAVRIS D N. Application of convolutional neural network to predict airfoil lift coefficient[R]. AIAA 2018-1903.

YANG S, RAMANAN D. Multi-scale recognition with DAG-CNNs[C]//2015 IEEE Intenational Conference on Computer Vision. IEEE, 2015: 1215-1223.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于条件扩散模型的火箭气动外形快速设计方法

基于U-Net的磁浮飞行风洞气动力干扰剥离算法

定常来流下钝体二维组合断面气动特性与流场的智能预测方法

基于物理约束深度学习的大跨柔性光伏阵列绕流场重构

基于机器学习的极端风场短时预测研究进展与思考