基于多模态融合的任意对称翼型结冰预测方法

屈经国; 王强; 彭博; 易贤

doi:10.13224/j.cnki.jasp.20220143

基于多模态融合的任意对称翼型结冰预测方法

doi: 10.13224/j.cnki.jasp.20220143

屈经国^{1, 2,},
王强^{2, 3},
彭博¹,
易贤^{2, 3,*, ,}

1.
西南石油大学计算机科学学院，成都 610500
2.
中国空气动力研究与发展中心结冰与防除冰重点实验室，四川绵阳 621000
3.
中国空气动力研究与发展中心空气动力学国家重点实验室，四川绵阳 621000

基金项目: 国家自然科学基金重点基金（12132019）；国家重大科技专项（J2019-Ⅲ-0010-0054）；国家自然科学面上基金（12172372）

详细信息

作者简介:
屈经国（1996-），男，硕士生，主要从事深度学习与结冰预测方面的研究。E-mail：xvyn@qq.com

通讯作者:
易贤（1977-），男，研究员，博士，主要从事航空宇航科学与技术方面的研究。E-mail：yixian_2000@163.com

中图分类号: V211.41
计量
- 文章访问数: 246
- HTML浏览量: 53
- PDF量: 91
- 被引次数: 0
出版历程
- 收稿日期: 2022-03-21
- 网络出版日期: 2023-09-04

Icing prediction method for arbitrary symmetric airfoil using multimodal fusion

QU Jingguo^{1, 2
,},
WANG Qiang^{2, 3},
PENG Bo¹,
YI Xian^{2, 3,*
, ,}

1.
School of Computer Science，Southwest Petroleum University，Chengdu 610500，China
2.
Key Laboratory of Icing and Anti/De-icing，China Aerodynamics Research and Development Center，Mianyang Sichuan 621000，China
3.
State Key Laboratory of Aerodynamics，China Aerodynamics Research and Development Center，Mianyang Sichuan 621000，China

摘要

摘要:
为解决目前绝大多数神经网络冰形预测方法只能针对特定翼型且不具备面向多翼型特征的普适性的问题，采用基于多模态融合的深度神经网络方法，以翼型截面图像与结冰工况参数作为输入，以二维冰形曲线傅里叶级数拟合参数作为输出，建立深度神经网络预测模型，实现了对任意对称翼型结冰特征的预测能力。结果表明：提出的模型可以准确地预测任意对称翼型几何特征条件下的结冰外形，冰形面积与最大冰厚等冰形主要参数预测误差均保持在10%以下。
- 对称翼型 /
- 结冰预测 /
- 多模态融合 /
- 深度神经网络 /
- 傅里叶级数
Abstract:
A deep neural network method based on multimodal fusion was adopted to solve the problem that most current neural network ice prediction methods can only target specific airfoils and do not have the universality of multi-airfoil features. This method used the airfoil cross-section image and the icing condition parameters as inputs, and the two-dimensional ice curve Fourier series fitting parameters as outputs. This deep neural network prediction model realized the prediction ability of the ice characteristics of any symmetric airfoil. The results showed that the proposed model can accurately predict the ice shape under the geometric characteristics of any symmetrical airfoil. The prediction error of the main parameters of the ice shape, such as the ice area and the maximum ice thickness, was kept below 10%.
- symmetric airfoils /
- icing prediction /
- multimodal fusion /
- deep neural networks /
- Fourier series

HTML

图 1 结冰快速预测输入条件

Figure 1. Ice accumulation prediction input conditions

下载: 全尺寸图片幻灯片

图 2 研究流程框图

Figure 2. Research flow diagram

下载: 全尺寸图片幻灯片

图 3 网络结构

Figure 3. Network structure

下载: 全尺寸图片幻灯片

图 4 典型训练集翼型截面

Figure 4. Typical training set airfoil sections

下载: 全尺寸图片幻灯片

图 5 结冰外形的坐标转换^[19]

Figure 5. Coordinate transformation of ice accretion shape^[19]

下载: 全尺寸图片幻灯片

图 6 测试集翼型截面

Figure 6. Test set airfoil sections

下载: 全尺寸图片幻灯片

图 7 冰形特征参数^[9]

Figure 7. Ice shape characteristic parameters^[9]

下载: 全尺寸图片幻灯片

图 8 预测霜冰冰形与数值计算冰形的比较

Figure 8. Comparison of prediction rime ice shape and numerical calculation ice shape

下载: 全尺寸图片幻灯片

图 9 预测明冰冰形与数值计算冰形的比较

Figure 9. Comparison of prediction glaze ice shape and numerical calculation ice shape

下载: 全尺寸图片幻灯片

表 1 结冰工况参数取值

Table 1. Parameter values for icing conditions

参数	取值
飞行攻角α/（°）	3
来流速度v/（m/s）	90
来流温度t/℃	−35, −30, −25, −20, −15, −10
液态水含量C_lw/（g/m³）	0.1, 0.2, 0.3, 0.4, 0.5, 0.7, 0.9, 1.1, 1.3, 1.5
水滴平均体积直径D_mv/μm	15, 20, 25, 30, 40, 50, 60
结冰时长τ/min	22.5

下载: 导出CSV

表 2 4组算例工况参数

Table 2. Parameters of 4 cases

算例	翼型	α/（°）	v/（m/s）	t/℃	C_lw/（g/m³）	D_mv/μm	τ/min
1	EPPLER 479	3	90	−30	0.4	25	22.5
2	STCYR172	3	90	−35	0.5	50	22.5
3	RAF 30	3	90	−30	1.5	40	22.5
4	Ryan BQM-34	3	90	−10	0.9	50	22.5

下载: 导出CSV

表 3 冰形特征参数相对误差

Table 3. Relative error of ice shape characteristic parameters

算例	损失函数	相对误差/%
算例	损失函数	S_u	S_l	S_ice	h_sp	h_u	h_l	θ_u	θ_l
1	MSE	2.40	1.53	6.28	1.61
1	Huber	7.09	7.85	17.55	0.49
2	MSE	13.64	15.95	1.55	0.97
2	Huber	4.57	9.78	4.88	8.90
3	MSE	3.20	18.89	0.47	7.19	4.76	2.10	0.22	0.27
3	Huber	16.63	16.12	14.27	0.65	7.55	1.54	0.24	0.15
4	MSE	0.35	24.86	4.25	7.60	5.10	7.04	0.25	0.32
4	Huber	5.69	21.49	6.00	8.01	6.01	8.48	0.22	0.20

下载: 导出CSV

参考文献(23)

[1]	LYNCH F T,KHODADOUST A. Effects of ice accretions on aircraft aerodynamics[J]. Progress in Aerospace Sciences,2001,37(8): 669-767. doi: 10.1016/S0376-0421(01)00018-5
[2]	BRAGG M. Aircraft aerodynamic effects due to large droplet ice accretions[R]. AIAA-1996-932, 1996.
[3]	易贤,朱国林,王开春,等. 翼型积冰的数值模拟[J]. 空气动力学学报,2002,20(4): 428-433. YI Xian,ZHU Guolin,WANG Kaichun,et al. Numerically simulating of ice accretion on airfoil[J]. Acta Aerodynamica Sinica,2002,20(4): 428-433. (in Chinese)
[4]	李小龙, 洪冠新. 一种基于神经网络的机翼结冰冰型预测方法[C]//飞行力学与飞行试验学术交流年会论文集. 北京: 国防工业出版社, 2006: 98-103. LI Xiaolong, HONG Guanxin. A neural network-based prediction method for wing icing shape[C]// Proceedings of the Annual Conference of Academic Exchanges on Flight Mechanics and Flight Tests (2006). Beijing: National Defense Industry Press, 2006: 98-103. (in Chinese)
[5]	OGRETIM E,HUEBSCH W,SHINN A. Aircraft ice accretion prediction based on neural networks[J]. Journal of Aircraft,2006,43(1): 233-240. doi: 10.2514/1.16241
[6]	CHANG Shinan,LENG Mengyao,WU Hongwei,et al. Aircraft ice accretion prediction using neural network and wavelet packet transform[J]. Aircraft Engineering and Aerospace Technology,2016,88(1): 128-136. doi: 10.1108/AEAT-05-2014-0057
[7]	何磊,钱炜祺,易贤,等. 基于转置卷积神经网络的翼型结冰冰形图像化预测方法[J]. 国防科技大学学报,2021,43(3): 98-106. HE Lei,QIAN Weiqi,YI Xian,et al. Graphical prediction method of airfoil ice shape based on transposed convolution neural networks[J]. Journal of National University of Defense Technology,2021,43(3): 98-106. (in Chinese)
[8]	YI Xian, WANG Qiang, CHAI Congcong, et al. Prediction model of aircraft icing based on deep neural network[J]. Transactions of Nanjing University of Aeronautics and Astronautics, 2021, 38(4): 535-544.
[9]	柴聪聪,易贤,郭磊,等. 基于BP神经网络的冰形特征参数预测[J]. 实验流体力学,2021,35(3): 16-21. CHAI Congcong,YI Xian,GUO Lei,et al. Prediction of ice shape characteristic parameters based on BP nerual network[J]. Journal of Experiments in Fluid Mechanics,2021,35(3): 16-21. (in Chinese)
[10]	柴聪聪. 基于深度学习的翼型结冰及其气动特性预测[D]. 成都: 电子科技大学, 2021. CHAI Congcong. Prediction of airfoil ice accretion and aerodynamic characteristics based on deep learning[D]. Chengdu: University of Electronic Science and Technology of China, 2021. (in Chinese)
[11]	DONG Yiqun. An application of deep neural networks to the in-flight parameter identification for detection and characterization of aircraft icing[J]. Aerospace Science and Technology,2018,77: 34-49. doi: 10.1016/j.ast.2018.02.026
[12]	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 Congcong,WANG Qiang,YI Xian,et al. Aerodynamic parameters prediction of airfoil ice accretion based on convolutional neural network[J]. Flight Dynamics,2021,39(5): 13-18. (in Chinese)
[13]	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 Lei,QIAN Weiqi,DONG Kangsheng,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)
[14]	ADDY H E. Ice accretions and icing effects for modern airfoils[M]. Cleveland, US: National Aeronautics and Space Administration, Glenn Research Center, 2000.
[15]	何俊,张彩庆,李小珍,等. 面向深度学习的多模态融合技术研究综述[J]. 计算机工程,2020,46(5): 1-11. HE Jun,ZHANG Caiqing,LI Xiaozhen,et al. Survey of research on multimodal fusion technology for deep learning[J]. Computer Engineering,2020,46(5): 1-11. (in Chinese)
[16]	GU Jiuxiang,WANG Zhenhua,KUEN J,et al. Recent advances in convolutional neural networks[J]. Pattern Recognition,2018,77: 354-377. doi: 10.1016/j.patcog.2017.10.013
[17]	REGULATIONS F A. Part 25-Airworthiness standards: transport category airplanes Appendix C[R]. 4080-29-FR-18291, 1970.
[18]	陈海,钱炜祺,何磊. 基于深度学习的翼型气动系数预测[J]. 空气动力学学报,2018,36(2): 294-299. CHEN Hai,QIAN Weiqi,HE Lei. Aerodynamic coefficient prediction of airfoils based on deep learning[J]. Acta Aerodynamica Sinica,2018,36(2): 294-299. (in Chinese)
[19]	张强,高正红. 基于神经网络的翼型积冰预测[J]. 飞行力学,2011,29(2): 6-9. ZHANG Qiang,GAO Zhenghong. Prediction of ice accretions based on the neural net[J]. Flight Dynamics,2011,29(2): 6-9. (in Chinese)
[20]	KINGMA D P, BA J. Adam: a method for stochastic optimization[EB/OL]. [2014-12-22]. https: //arxiv.org/abs/1412.6980
[21]	HUBER P J. Robust estimation of a location parameter[M]//KOTZ S, JOHNSON N L. Breakthroughs in Statistics. New York: Springer, 1992: 492-518.
[22]	JAMES G, WITTEN D, HASTIE T, et al. An Introduction to Statistical Learning: vol 103[M]. New York: Springer, 2013.
[23]	AC-9C Aircraft Icing Technology Committee. Icing wind tunnel interfacility comparison tests[EB/OL]. [2018-10-16]. https: //www.sae.org/standards/content/air5666a/