基于迁移学习的点云机翼气动性能预测

冯怡忆; 刘珊珊; 董守鸿; 徐淑颖; 王亮权; 陈德华; 张书俊

doi:10.7638/kqdlxxb-2024.0221

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 >

基于迁移学习的点云机翼气动性能预测

研究论文 | 更新时间：2026-05-28

- 基于迁移学习的点云机翼气动性能预测
- Research on aerodynamic performance prediction of point cloud wings based on transfer learning
- 空气动力学学报 2026年44卷第2期页码：50-58
- 作者机构：
  
  1. 电子科技大学航空航天学院,成都,611731
  2. 中国空气动力研究与发展中心旋翼空气动力学重点实验室,绵阳,621000
- 作者简介：
  
  冯怡忆（2001—），湖北荆门人，硕士研究生，研究方向：航空智能计算. E-mail：1246093277@qq.com
  张书俊*，河南内乡人，副研究员，研究方向：流体力学、飞行器多学科设计优化. E-mail：zhangsj@uestc.edu.cn
- 基金信息：
  
  旋翼空气动力学重点实验室研究开放课题（2104RAL202102-1）
- DOI：10.7638/kqdlxxb-2024.0221
  中图分类号： TP181;V211.3
- 纸质出版：2026
- 稿件说明：
移动端阅览
冯怡忆, 刘珊珊, 董守鸿, 等. 基于迁移学习的点云机翼气动性能预测[J]. 空气动力学学报, 2026, 44(2): 50−58.

Feng Yiyi, Liu Shanshan, Dong Shouhong, et al. Research on aerodynamic performance prediction of point cloud wings based on transfer learning[J]. 2026, 44(2): 50-58.
冯怡忆, 刘珊珊, 董守鸿, 等. 基于迁移学习的点云机翼气动性能预测[J]. 空气动力学学报, 2026, 44(2): 50−58. DOI： 10.7638/kqdlxxb-2024.0221

Feng Yiyi, Liu Shanshan, Dong Shouhong, et al. Research on aerodynamic performance prediction of point cloud wings based on transfer learning[J]. 2026, 44(2): 50-58. DOI： 10.7638/kqdlxxb-2024.0221.

摘要

当前机翼设计阶段获取大量机翼气动性能需要巨大成本，因此亟需发展一种可以高效且较为准确地获取机翼气动性能数据的方法。以低速双梯形机翼为研究对象，通过OpenVSP和CFD构建低亚声速飞行条件下的机翼数据集，基于此数据集设计了点云机翼气动性能耦合预测网络（PointNet-AP），并利用高低精度数据进行迁移学习。PointNet-AP以机翼点云数据、传统平面特征以及飞行条件为输入，预测机翼对应飞行条件下的升力系数（$C_L$）、阻力系数（$C_D$）和俯仰力矩系数（$C_m$）。实验结果表明，迁移学习后的PointNet-AP网络模型对机翼气动性能有较好的预测能力，$C_L$、$C_D$和$C_m$的平均相对预测误差分别为2.889%、4.728%和6.743%。本文所提出的预测网络可为后续气动性能预测大模型的开发提供数据特征提取和模型迁移的技术基础。

Abstract

Acquiring extensive aerodynamic performance data in the current wing design phase incurs substantial costs

necessitating the development of efficient and accurate predictive methods. This study focused on a low-speed double-trapezoid wing. A dataset under subsonic flight conditions was constructed using OpenVSP and CFD simulations. Based on this dataset

a PointNet-based aerodynamic performance prediction network (PointNet-AP) was designed

which incorporated transfer learning utilizing multi-fidelity data. The PointNet-AP network took wing point cloud data

conventional planar geometric features

and flight conditions as inputs to predict the corresponding lift coefficient ($C_L$)

drag coefficient ($C_D$)

and pitching moment coefficient ($C_m$). Experimental results demonstrated that the PointNet-AP model with transfer learning achieved satisfactory predictive performance

with mean relative prediction errors of 2.88%

4.75%

and 6.74% for $C_L$

$C_D$

and $C_m$

respectively.The proposed network provides a technical foundation for developing large-scale aerodynamic performance prediction models by offering solutions for data feature extraction and model transfer.

关键词

Keywords

references

Li Junfu. Study on two-level optimization method for aerodynamic shape optimization of wings[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2008 (in Chinese).

Wu Tingfan. CFD grid adaptation based on machine learning[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2023 (in Chinese). doi: 10.27239/d.cnki.gnhhu.2021.001963

Dos Santos M C, De Mattos B S, Girardi R D M. Aerodynamic coefficient prediction of airfoils using neural networks[C]//46th AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada, 2008. doi: 10.2514/6.2008-887

Jihong H, Su H, Zhao X. Aerodynamic coefficient prediction of airfoils using BP neural network[J]. Advances in Aeronautical Science and Engineering, 2010, 1(1): 36−39.

Chen H, He L, Qian W, et al. Multiple aerodynamic coefficient prediction of airfoils using a convolutional neural network[J]. Symmetry, 2020, 12(4), 544. doi: 10.3390/sym12040544

Dong Baoyu. Aerodynamic performance study of airfoils based on machinelearning[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2021 (in Chinese).

Mosbah A B, Botez R M, Dao T M. A hybrid original approach for prediction of the aerodynamic coefficients of an ATR-42 scaled wing model[J]. Chinese Journal of Aeronautics, 2016, 29(1): 41−52. doi: 10.1016/j.cja.2015.12.022

Liu Xin. Simulation study on aerodynamic parameter prediction of wings based on neural networks[J]. Computer Simulation, 2015, 32(12): 67−71 (in Chinese). doi: 10.3969/j.issn.1006-9348.2015.12.014

Wu Yongchuan, Sun Gang, Tao Jun. Optimization design of general aircraft wings based on deep belief network and multi-objective particle swarm algorithm[J]. Acta Aerodynamica Sinica, 2023, 41(12): 16−27 (in Chinese). doi: 10.7638/kqdlxxb-2023.0005

Zhang Junjie, Huang Jun, Liu Zhiqin, et al. Variable credibility aerodynamic modeling method based on transfer learning[J]. Journal of Experiments in Fluid Mechanics, 2025, 39(4): 93−103 (in Chinese). doi: 10.11729/syltlx20220058

Runze L, Zhang Y, Haixin C. Transfer learning from two-dimensional supercritical airfoils to three-dimensional transonic swept wings[J]. Chinese Journal of Aeronautics, 2023, 36(9): 96−110. doi: 10.1016/j.cja.2023.04.008

Chen Shusheng, Jia Muliang, Lin Jiahao, et al. Empowering aircraft technology applications with generative models: research progress and prospects[J]. Acta Aeronautica et Astronautica Sinica, 2025, 46(10): 631194 (in Chinese). doi: 10.7527/S1000-6893.2024.31194

Tang Zhigong, Qian Weiqi, He Lei, et al. Thoughts and prospects on large model research in aerodynamics[J]. Acta Aerodynamica Sinica, 2024, 42(12): 1−11 (in Chinese). doi: 10.7638/kqdlxxb-2024.0056

Zhang Jiawei. Effect of airfoil thickness on aerodynamic performance[J]. Power System Engineering, 2021, 37(02): 5−7 (in Chinese).

Shao Chengshi. Aerodynamic characteristics analysis of tilt-rotor flying cars[D]. Jilin: Jilin University, 2023 (in Chinese).

He W, Liu X. Improved aerofoil parameterisation based on class/shape function transformation[J]. Aeronautical Journal, 2019, 123(1261): 310−339 (in Chinese). doi: 10.1017/aer.2018.165

Charles R Q, Su H, Kaichun M, et al. PointNet: Deep learning on point sets for 3d classification and segmentation[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, HI: IEEE, 2017: 77−85. doi: 10.1109/CVPR.2017.16

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

高原山谷对直升机悬停旋翼的气动干扰

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

基于多精度深度神经网络的汽车气动外形优化设计方法

基于自主软件LSWS的典型地形大气CFD模式研究与应用

风能模型的发展及CFD在风资源开发利用中的应用