基于点云分割算法的战术导弹表面压力分布快速预测

蔺佳哲; 何磊; 程明; 周岭; 杨春明

doi:10.13700/j.bh.1001-5965.2024.0172

基于点云分割算法的战术导弹表面压力分布快速预测

doi: 10.13700/j.bh.1001-5965.2024.0172

蔺佳哲¹,
何磊¹,
程明¹,
周岭^1, ,,
杨春明²

1.
中国空气动力研究与发展中心计算空气动力研究所，绵阳 621000
2.
西南科技大学计算机科学与技术学院，绵阳 621000

详细信息

通讯作者:
E-mail：zlkk72@sina.com

中图分类号: V211.24；V211.3
计量
- 文章访问数: 378
- HTML全文浏览量: 187
- PDF下载量: 7
- 被引次数: 0
出版历程
- 收稿日期: 2024-03-26
- 录用日期: 2024-05-24
- 网络出版日期: 2024-08-29
- 整期出版日期: 2026-05-26

Rapid prediction of surface pressure distribution of tactical missile based on point cloud segmentation algorithm

1.
Computational Aerodynamic Institute，China Aerodynamics Research and Development Center，Mianyang 621000，China
2.
School of Computer Science and Technology，Southwest University of Science and Technology，Mianyang 621000，China

More Information

Corresponding author: E-mail：zlkk72@sina.com

摘要

摘要:
如何缩短飞行器的设计周期是我国航空航天领域亟待解决的难点问题，其中，气动设计是飞行器设计的关键环节，快速精确获取飞行器气动特性可有力加速概念设计方案的迭代改进。因此，基于中国空气动力研究与发展中心现有气动数据库，利用“1+N”模式的深度神经网络建模算法，构建来流参数、气动外形到表面压力分布数据的映射关系。为提升预测精度，结合压力分布数据原有的物面网格划分信息，改进点云分割PointNet++算法，准确识别导弹不同部件，自动增加不同部件的标签特征。测试案例表明，采用改进的点云分割算法和“1+N”模式的深度神经网络建模算法，战术导弹全弹表面压力分布预测平均相对误差(MRE)基本控制在10%以内。所提算法建模效率较高，适用于各类复杂外形飞行器的压力分布预测，具有较好的工程应用前景。
- 战术导弹 /
- 压力分布预测 /
- 点云分割 /
- 深度神经网络 /
- PointNet++算法
Abstract:
How to shorten the design period of aircraft is a difficult problem to be solved urgently in the field of aerospace. Aerodynamic design is the key link of aircraft design. Rapid and accurate acquisition of aircraft aerodynamic characteristics can effectively accelerate the iterative improvement of conceptual design schemes. The deep neural network modeling algorithm of “1+N” mode is used to build the mapping relationship between flow parameters, aerodynamic profile, and surface pressure distribution data based on the China Aerodynamics Research and Development Center’s current aerodynamic database. In order to improve the prediction accuracy, the point cloud segmentation PointNet++ algorithm was improved by combining the original surface grid division information of the pressure distribution data to accurately identify different components of the missile and automatically add label features of different components. The test case shows that by using the improved point cloud segmentation algorithm and the “1+N” mode deep neural network algorithm, the mean relative error (MRE) of the pressure distribution prediction of the tactical missile is basically controlled within 10%. The aforementioned approach has a high modeling efficiency, is appropriate for predicting the pressure distribution of a variety of intricate aircraft forms, and has a promising engineering application.
- tactical missiles /
- pressure distribution prediction /
- point cloud segmentation /
- deep neural network /
- PointNet++ algorithm

HTML全文

图 1 不同类型的战术导弹气动外形（举例）

Figure 1. Aerodynamic shapes of different types of tactical missiles (examples)

下载: 全尺寸图片幻灯片

图 2 PointNet++点云分割基本原理

Figure 2. Basic principle of PointNet++ point cloud segmentation

下载: 全尺寸图片幻灯片

图 3 战术导弹点云部件分割示意图

Figure 3. Schematic diagram of tactical missile point cloud component segmentation

下载: 全尺寸图片幻灯片

图 4 战术导弹表面压力分布预测流程

Figure 4. Flow chart of surface pressure distribution prediction for tactical missiles

下载: 全尺寸图片幻灯片

图 5 导弹模型3在马赫数1.2、攻角2°、滚转角22.5°条件下弹头表面压力分布CFD计算结果及2种建模算法预测结果

Figure 5. CFD calculation results and two modeling algorithms’ prediction results of warhead surface pressure distribution for missile model 3 under Mach number 1.2, angle of attack 2° and roll angle 22.5°

下载: 全尺寸图片幻灯片

图 6 导弹模型3在马赫数1.2、攻角2°、滚转角22.5°条件下弹翼表面压力分布CFD计算结果及2种建模算法预测结果

Figure 6. CFD calculation results and two modeling algorithms’ prediction results of missile wing surface pressure distribution for missile model 3 under Mach number 1.2, angle of attack 2° and roll angle 22.5°

下载: 全尺寸图片幻灯片

图 7 导弹模型3在马赫数1.2、攻角2°、滚转角22.5°条件下尾翼表面压力分布CFD计算结果及2种建模算法预测结果

Figure 7. CFD calculation results and two modeling algorithms’ prediction results of tail wing surface pressure distribution for missile model 3 under Mach number 1.2, angle of attack 2° and roll angle 22.5°

下载: 全尺寸图片幻灯片

图 8 2种建模算法的预测MRE对比

Figure 8. Comparison of predicted MRE between twomodeling algorithms

下载: 全尺寸图片幻灯片

表 1 战术导弹表面压力分布预测误差

Table 1. Prediction error of tactical missile surfacepressure distribution

对象	建模算法	RMSE	MAE	MRE/%
全弹	算法1	0.0586	0.0269	26.56
全弹	算法2	0.0176	0.00896	8.59
弹头	算法1	0.0377	0.0196	17.52
弹头	算法2	0.0050	0.0037	3.33
弹尾喷口	算法1	0.0292	0.0151	7.61
弹尾喷口	算法2	0.0047	0.0030	1.56
弹翼	算法1	0.1114	0.0548	39.46
弹翼	算法2	0.0338	0.0200	14.67
弹身与弹翼融合段	算法1	0.0429	0.0310	35.03
弹身与弹翼融合段	算法2	0.0181	0.0129	13.38
前弹身	算法1	0.0357	0.0189	21.02
前弹身	算法2	0.0090	0.0052	5.97
后弹身	算法1	0.0293	0.0206	30.66
后弹身	算法2	0.0116	0.0080	11.04
尾翼	算法1	0.0451	0.0240	26.93
尾翼	算法2	0.0123	0.0074	7.79
弹身与尾翼融合段	算法1	0.0238	0.0145	19.98
弹身与尾翼融合段	算法2	0.0058	0.0043	5.37

下载: 导出CSV

参考文献(31)

[1]	BEKEMEYER P, BERTRAM A, HINES-CHAVES D A, et al. Data-driven aerodynamic modeling using the DLR SMARTy toolbox[C]//AIAA AVIATION 2022 Forum. Reston: AIAA, 2022: 3899.
[2]	SABATER C, STÜRMER P, BEKEMEYER P. Fast Predictions of aircraft aerodynamics using deep learning techniques[C]//AIAA AVIATION 2021 Forum. Reston: AIAA, 2021: 2549.
[3]	张伟伟, 寇家庆, 刘溢浪. 智能赋能流体力学展望[J]. 航空学报, 2021, 42(4): 20-65. ZHANG W W, KOU J Q, LIU Y L. Prospect of artificial intelligence empowered fluid mechanics[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4): 20-65(in Chinese).
[4]	曹晓峰, 李鸿岩, 郭承鹏, 等. 基于深度学习的二维翼型流场重构技术研究[J]. 航空科学技术, 2022, 33(7): 106-112. CAO X F, LI H Y, GUO C P, et al. Research on two-dimensional airfoil flow field reconstruction based on deep learning[J]. Aeronautical Science & Technology, 2022, 33(7): 106-112(in Chinese).
[5]	HU J W, DOU Z H, ZHANG W W. Fast fluid-structure interaction simulation method based on deep learning flow field modeling[J]. Physics of Fluids, 2024, 36(4): 045106.
[6]	HU J W, ZHANG W W. Flow field modeling of airfoil based on convolutional neural networks from transform domain perspective[J]. Aerospace Science and Technology, 2023, 136: 108198.
[7]	蔡国飙, 张百一, 贺碧蛟, 等. 真空羽流智能化计算[J]. 航空学报, 2022, 43(10): 102-114. CAI G B, ZHANG B Y, HE B J, et al. Intelligent computation of vacuum plume[J]. Acta Aeronautica et Astronautica Sinica, 2022, 43(10): 102-114(in Chinese).
[8]	何磊, 钱炜祺, 董康生, 等. 基于卷积神经网络的结冰翼型气动特性建模[J]. 航空学报, 2023, 44(5): 59-72. HE L, QIAN W Q, DONG K S, et al. Aerodynamic characteristics modeling of iced airfoil based on convolution neural networks[J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(5): 59-72(in Chinese).
[9]	柴聪聪, 王强, 易贤, 等. 基于卷积神经网络的结冰翼型气动参数预测[J]. 飞行力学, 2021, 39(5): 13-18. CHAI C C, WANG Q, YI X, et al. Aerodynamic parameters prediction of airfoil ice accretion based on convolutional neural network[J]. Flight Dynamics, 2021, 39(5): 13-18(in Chinese).
[10]	叶舒然, 张珍, 王一伟, 等. 基于卷积神经网络的深度学习流场特征识别及应用进展[J]. 航空学报, 2021, 42(4): 179-193. YE S R, ZHANG Z, WANG Y W, et al. Progress in deep convolutional neural network based flow field recognition and its applications[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4): 179-193(in Chinese).
[11]	OBAYASHI W, AONO H, TATSUKAWA T, et al. Feature extraction of fields of fluid dynamics data using sparse convolutional autoencoder[J]. AIP Advances, 2021, 11(10): 105211.
[12]	MURATA T, FUKAMI K, FUKAGATA K. Nonlinear mode decomposition with convolutional neural networks for fluid dynamics[J]. Journal of Fluid Mechanics, 2020, 882: A13.
[13]	吕召阳, 聂雪媛, 赵奥博. 基于CNN机翼气动系数预测[J]. 北京航空航天大学学报, 2023, 49(3): 674-680. LYU Z Y, NIE X Y, ZHAO A B. Prediction of wing aerodynamic coefficient based on CNN[J]. Journal of Beijing University of Aeronautics and Astronautics, 2023, 49(3): 674-680(in Chinese).
[14]	CAO W B, SONG J H, ZHANG W W. A solver for subsonic flow around airfoils based on physics-informed neural networks and mesh transformation[J]. Physics of Fluids, 2024, 36(2): 027134.
[15]	WU L, CUI B, WANG R, et al. Artificial neural network-substituted transition model for crossflow instability: modeling strategy and application prospect[J]. Physics of Fluids, 2024, 36(4): 045110.
[16]	SHEN Y, HUANG W, WANG Z G, et al. A deep learning framework for aerodynamic pressure prediction on general three-dimensional configurations[J]. Physics of Fluids, 2023, 35(10): 107111.
[17]	LE CLAINCHE S, LI J I, THEOFILIS V, et al. Flow around a hemisphere-cylinder at high angle of attack and low Reynolds number. part I: experimental and numerical investigation[J]. Aerospace Science and Technology, 2015, 44: 77-87.
[18]	LE CLAINCHE S, RODRÍGUEZ D, THEOFILIS V, et al. Flow around a hemisphere-cylinder at high angle of attack and low Reynolds number. part II: POD and DMD applied to reduced domains[J]. Aerospace Science and Technology, 2015, 44: 88-100.
[19]	AGOSTINI L. Exploration and prediction of fluid dynamical systems using auto-encoder technology[J]. Physics of Fluids, 2020, 32(6): 067103.
[20]	CAO C Q, NIE C S, PAN S C, et al. A constrained reduced-order method for fast prediction of steady hypersonic flows[J]. Aerospace Science and Technology, 2019, 91: 679-690.
[21]	完颜振海, 徐嘉, 梁磊, 等. 基于特征代理模型的复杂流场预测方法研究[J]. 弹箭与制导学报, 2017, 37(6): 105-108. WANYAN Z H, XU J, LIANG L, et al. Research on method of prediction for complicated flow fields based on characteristic surrogate model[J]. Journal of Projectiles, Rockets, Missiles and Guidance, 2017, 37(6): 105-108(in Chinese).
[22]	侯强, 苏纬仪, 孙斐, 等. 高超声速乘波体扩容设计及流场快速预测[J]. 航空动力学报, 2021, 36(3): 564-574. HOU Q, SU W Y, SUN F, et al. Design of improving volumetric efficiency of hypersonic waverider and rapid prediction of flow field[J]. Journal of Aerospace Power, 2021, 36(3): 564-574(in Chinese).
[23]	聂春生, 黄建栋, 王迅, 等. 基于POD方法的复杂外形飞行器热环境快速预测方法[J]. 空气动力学学报, 2017, 35(6): 760-765. NIE C S, HUANG J D, WANG X, et al. Fast aeroheating prediction method for complex shape vehicles based on proper orthogonal decomposition[J]. Acta Aerodynamica Sinica, 2017, 35(6): 760-765(in Chinese).
[24]	WANG J, HE C, LI R Z, et al. Flow field prediction of supercritical airfoils via variational autoencoder based deep learning framework[J]. Physics of Fluids, 2021, 33(8): 086108.
[25]	WU P, GONG S Q, PAN K K, et al. Reduced order model using convolutional auto-encoder with self-attention[J]. Physics of Fluids, 2021, 33(7): 077107.
[26]	HASEGAWA K, FUKAMI K, MURATA T, et al. Machine-learning-based reduced-order modeling for unsteady flows around bluff bodies of various shapes[J]. Theoretical and Computational Fluid Dynamics, 2020, 34(4): 367-383.
[27]	HASEGAWA K, FUKAMI K, MURATA T, et al. CNN-LSTM based reduced order modeling of two-dimensional unsteady flows around a circular cylinder at different Reynolds numbers[J]. Fluid Dynamics Research, 2020, 52(6): 065501.
[28]	EIVAZI H, VEISI H, NADERI M H, et al. Deep neural networks for nonlinear model order reduction of unsteady flows[J]. Physics of Fluids, 2020, 32(10): 105104.
[29]	张智超, 高太元, 张磊, 等. 基于径向基神经网络的气动热预测代理模型[J]. 航空学报, 2021, 42(4): 297-306. ZHANG Z C, GAO T Y, ZHANG L, et al. Aeroheating agent model based on radial basis function neural network[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4): 297-306(in Chinese).
[30]	朱安迪, 达飞鹏, 盖绍彦. 对融合特征敏感的三维点云识别与分割[J]. 西安交通大学学报, 2024, 58(5): 52-63. ZHU A D, DA F P, GAI S Y. Recognition and segmentation of point clouds sensitive to fusion features[J]. Journal of Xi’an Jiaotong University, 2024, 58(5): 52-63(in Chinese).
[31]	孙端正, 高飞, 叶周润, 等. 改进PointNet++模型在道路杆状物提取中的应用[J]. 测绘通报, 2023(11): 95-99. SUN D Z, GAO F, YE Z R, et al. Application of improved PointNet++ model in extracting road rods[J]. Bulletin of Surveying and Mapping, 2023(11): 95-99(in Chinese).