基于稀疏贝叶斯优化的翼型设计可解释性研究

林健; 吕宏强; 黄增辉; 刘子敬; 虞建; 刘学军

doi:10.7638/kqdlxxb-2023.0183

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 >

基于稀疏贝叶斯优化的翼型设计可解释性研究

研究论文 | 更新时间：2026-05-28

- 基于稀疏贝叶斯优化的翼型设计可解释性研究
- Research on interpretability of airfoil design based on sparse Bayesian optimization
- 空气动力学学报 2025年43卷第1期页码：22-33
- 作者机构：
  
  1. 南京航空航天大学计算机科学与技术学院, 模式分析与机器智能工业和信息化部重点实验室,南京,211106
  2. 南京航空航天大学航空学院,南京,210016
  3. 中国商飞上海飞机设计研究院,上海,201210
  4. 中国航空研究院,北京,100012
- 作者简介：
  
  林健（1998—），男，安徽人，硕士研究生，研究方向：机器学习. E-mail：linjian@nuaa.edu.cn
  吕宏强*（1977—），男，教授，研究方向：计算流体力学. E-mail：hongqiang.lu@nuaa.edu.cn
  刘学军*（1976—），女，教授，研究方向：机器学习. E-mail：xuejun.liu@nuaa.edu.cn
- 基金信息：
  
  航空科学基金（2018ZA52002，2019ZA052011）
- DOI：10.7638/kqdlxxb-2023.0183
  中图分类号： TP181;V211.3
- 纸质出版：2025
- 稿件说明：
移动端阅览
林健, 吕宏强, 黄增辉, 等. 基于稀疏贝叶斯优化的翼型设计可解释性研究[J]. 空气动力学学报, 2025, 43(1): 22−33.

LIN Jian, LYU Hongqiang, HUANG Zenghui, et al. Research on interpretability of airfoil design based on sparse Bayesian optimization[J]. 2025, 43(1): 22-33.
林健, 吕宏强, 黄增辉, 等. 基于稀疏贝叶斯优化的翼型设计可解释性研究[J]. 空气动力学学报, 2025, 43(1): 22−33. DOI： 10.7638/kqdlxxb-2023.0183

LIN Jian, LYU Hongqiang, HUANG Zenghui, et al. Research on interpretability of airfoil design based on sparse Bayesian optimization[J]. 2025, 43(1): 22-33. DOI： 10.7638/kqdlxxb-2023.0183.

摘要

贝叶斯优化框架具有优化效率高、效果好等特点，适合解决高维黑盒优化问题，适用于飞机翼型设计领域。然而其优化过程不透明，难以直观理解机器优化结果和翼型典型物理特征之间的联系，如何解释贝叶斯优化进程仍然是一个挑战。针对这一问题，本文提出了一种基于稀疏贝叶斯优化框架的翼型优化可解释性方法，使用具有物理意义的典型几何特征参与优化进程，在贝叶斯优化过程中对翼型特征进行稀疏，同时获得可解释性信息。在以RAE2822为基准翼型的超临界翼型优化算例上验证该方法。实验结果表明，该方法在优化气动性能的同时尽可能地减少了翼型设计维度，使其在保证气动性能良好的情况下具备了一定的可解释性，能直观地了解翼型各参数对优化目标的影响程度，辅助翼型设计人员进行决策和判断。

Abstract

The Bayesian optimization framework has the characteristics of high optimization efficiency and good optimization effect

which is suitable for solving high-dimensional black-box optimization problems

i.e.

in the field of airfoil design. However

due to the opacity of the optimization process

it is difficult to intuitively understand the relationship between the machine optimization results and typical physical characteristics of the airfoil. Therefore

how to interpret the Bayesian optimization process is still a challenge. To solve this problem

an airfoil optimization interpretability method based on the sparse Bayesian optimization framework is proposed

which utilizes typical geometric features of physical significance in the optimization process. During the process of Bayesian optimization

the airfoil features are sparse

and interpretability information is obtained. The proposed method is verified using the example of supercritical airfoil optimization with RAE2822 as the reference airfoil. The experimental results show that the proposed method reduces the airfoil design dimension as much as possible

and makes the data explainable to a certain extent while ensuring good aerodynamic performance

which can help intuitively understand the influence degree of airfoil parameters on the optimization objective and assist designers to make decision and judgment during the airfoil design.

关键词

Keywords

references

PRASAD K, KRISHNA V, KUMAR B B A . Aerofoil profile analysis and design optimisation[J ] . Journal of Aerospace Engineering and Technology, 2013 , 3 ( 2 ): 1 − 10 .

LIU Z J, QU X Y, LIU X J, et al . Correlation-concerned Bayesian optimization for multi-objective airfoil design[J ] . Aerospace Science and Technology, 2022 , 129 : 107867 .

CARVALHO D V, PEREIRA E M, CARDOSO J S . Machine learning interpretability: A survey on methods and metrics[J ] . Electronics, 2019 , 8 ( 8 ): 832 .

ZHOU B L, KHOSLA A, LAPEDRIZA A, et al. Learning deep features for discriminative localization[C]//Proc of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA. IEEE, 2016: 2921-2929.

BELINKOV Y, GEHRMANN S, PAVLICK E. Interpretability and analysis in neural NLP[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: Tutorial Abstracts, Online. Stroudsburg, PA, USA: Association for Computational Linguistics, 2020: 15.

LIU Z J, AN W, QU X Y, et al . Portfolio-based Bayesian optimization for airfoil design[J ] . AIAA Journal, 2021 , 59 ( 6 ): 1975 − 1989 .

MASTERS D A, TAYLOR N J, RENDALL T, et al. Review of aerofoil parameterisation methods for aerodynamic shape optimisation[C]// 53rd AIAA Aerospace Sciences Meeting, Kissimmee, Florida. Reston, Virginia: AIAA, 2015: 0761.

SCHULZ E, SPEEKENBRINK M, KRAUSE A . A tutorial on Gaussian process regression: Modelling, exploring, and exploiting functions[J ] . Journal of Mathematical Psychology, 2018 , 85 : 1 − 16 .

ABDOLSHAH M, SHILTON A, RANA S, et al. Expected hypervolume improvement with constraints[C]//2018 24th International Conference on Pattern Recognition (ICPR), Beijing, China. IEEE, 2018: 32383243.

ZHU C Y, BYRD R H, LU P H, et al . Algorithm 778: L-BFGS-B: Fortran subroutines for large-scale bound-constrained optimization[J ] . ACM Transactions on mathematical software (TOMS), 1997 , 23 ( 4 ): 550 − 560 .

MARTINEZ-CANTIN R, DE FREITAS N, BROCHU E, et al . A Bayesian exploration-exploitation approach for optimal online sensing and planning with a visually guided mobile robot[J ] . Autonomous Robots, 2009 , 27(2) : 93 − 103 .

BROCHU E, BROCHU T, DE FREITAS N. A Bayesian interactive optimization approach to procedural animation design[J]. Computer Animation 2010 ACM SIGGRAPH / Eurographics Symposium Proceedings, SCA 2010, 2010: 103112.

SEEGER M . Gaussian processes for machine learning[J ] . International Journal of Neural Systems, 2004 , 14 ( 02 ): 69 − 106 .

MARCO A, BERKENKAMP F, HENNIG P, et al. Virtual vs. real: Trading off simulations and physical experiments in reinforcement learning with Bayesian optimization[C]//2017 IEEE International Conference on Robotics and Automation (ICRA), Singapore. IEEE, 2017: 15571563.

DE ATH G, EVERSON R M, RAHAT A A M, et al . Greed is good: Exploration and exploitation trade-offs in Bayesian optimisation[J ] . ACM Transa ctions on Evolutionary Learning and Optimization, 2021 , 1 ( 1 ): 1 − 22 .

VAN VELDHUIZEN D A, LAMONT G B. Evolutionary computation and convergence to a Pareto front[C]//Late Breaking Papers at the Genetic Programming 1998 Conference, 1998.

SONG C, LI W B, ZHOU Z, et al. Multi-objective aerodynamic optimization algorithm based on manifold reconstruction[J]. Acta Aeronautica et Astronautica Sinica, 2020, 41(5): 623687(in Chinese).

LIU Z J, LIU X J, CAI X Y . A new hybrid aerodynamic optimization framework based on differential evolution and invasive weed optimization[J ] . Chinese Journal of Aeronautics, 2018 , 31 ( 7 ): 1437 − 1448 .

LIU Z J. Research on key technologies of optimal design of supercritical airfoil[D].Nanjing: Nanjing University of Aeronautics and Astronautics, 2018(in Chinese).

BIRAJDAR M R, KALE S A . Effect of leading edge radius and blending distance from leading edge on the aerodynamic performance of small wind turbine blade airfoils[J ] . International Journal of Energy and Power Engineering, 2015 , 4 ( 5-1 ): 54 − 58 .

HENDERSON W P. Studies of various factors affecting drag due to lift at subsonic speeds: NASA-TN-D-3584[R]. 1966.

ABDULKAREEM O A, KHUDHEYER A F, ABBAS A S. Numerical investigation of the effect of changing the thickness of airfoils used in wind turbines on the lift to drag ratio[J]. IOP Conference Series: Materials Science and Engineering, 2021, 1094(1): 012078.

SETHUNATHAN P, NIVENTHRAN M, SIVA V, et al . Analysis of aerodynamic characteristics of a supercritical airfoil for low speed aircraft[J ] . IJRET: International Journal of Research in Engineering and Technology, 2014 , 3 ( 6 ): 179 − 183 .

SOBIECZKY H, SEEBASS A R . Supercritical airfoil and wing design[J ] . Annual Review of Fluid Mechanics, 1984 , 16 ( 1 ): 337 − 363 .

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

基于一种贝叶斯优化框架的高空螺旋桨气动优化设计

空天飞行器机翼/翼型的需求分析及应用

旋翼翼型气动设计与评估软件HRADesign

中空长航时无人机两段翼型设计研究

风扇叶片颗粒侵蚀特性的多因素分析与模拟