Ice cloud parameter identification method in icing wind tunnel based on multimodal fusion
-
摘要: 结冰风洞云雾场校测通常存在仪器依赖度高的问题。针对该问题,提出了一种基于多模态融合的结冰风洞云雾参数辨识方法,该方法以试验模型结冰图像及迎角、来流速度、来流温度、结冰时长等参数作为输入,提取并融合两者特征参数,以液态水含量和水滴平均体积直径作为输出训练神经网络模型,进而实现对结冰风洞云雾参数的快速辨识。为验证该方法的有效性和可行性,以NACA0012标准翼型结冰为研究对象,开发了结冰风洞云雾参数辨识程序,分析了融合比例的影响,获得了适用于结冰风洞云雾参数辨识的最佳网络模型。在此基础上,开展了仿真和试验评估,结果表明:所提出的方法对液态水含量和水滴平均体积直径的辨识满度误差均小于12%,具有较高的辨识精度与良好的泛化性能,可为结冰风洞云雾参数辨识提供补充。Abstract: The cloud field calibration of icing wind tunnels usually has the disadvantage of high instrument dependence. To solve this problem, this paper proposes a method for identifying the parameters of cloud fields in icing wind tunnels based on multi-modal fusion. This method takes the icing image of the test model together with the parameters such as the angle of attack, air velocity, air temperature, and icing duration of the model as input, extracts and fuses the two characteristic parameters, and takes the liquid water content (LWC) and the average volume diameter of water droplets (MVD) as the output to train the neural network model. And then the rapid identification of icing cloud parameters is realized. In order to verify the effectiveness and feasibility of the proposed method, the paper takes NACA0012 airfoil icing as the research object, develops the cloud field identification program of the icing wind tunnel, analyzes the influence of the fusion proportion, and obtains the best network model suitable for ice parameter identification. On this basis, simulation and experimental evaluation are carried out. The full scale error of the proposed method for LWC and MVD is less than 12%, which has high identification accuracy and good generalization performance, and can provide an important supplement for the identification of cloud fields in the icing wind tunnel.
-
表 1 结冰工况参数设置
Table 1. Icing parameter setting
参数 取值 迎角α/(°) 0,3 来流速度v/(m·s−1) 80,100 来流温度T/(℃) −30,−20,−10 液态水含量LWC/(g·m−3) 0.1~1.0 水滴平均体积直径MVD/μm 20~100 结冰时长t/s 1350 表 2 测试集1的结冰工况
Table 2. Test set 1 icing conditions
算例 迎角
/(°)来流速度
/(m·s−1)来流
温度/(℃)结冰
时长/s水滴平均体积
直径/μm液态水含量
/(g·m−3)a 0 80 −20 1350 20 0.25 b 3 80 −30 1350 26 0.61 c 0 100 −20 1350 26 0.76 d 3 80 −20 1350 38 0.91 e 0 80 −10 1350 47 0.76 f 0 80 −20 1350 56 0.94 g 3 100 −10 1350 65 0.25 h 0 80 −30 1350 74 0.40 i 3 80 −30 1350 89 0.22 表 3 测试集2的结冰工况
Table 3. Test set 2 icing conditions
算例 迎角
/(°)来流速度
/(m·s−1)来流
温度/(℃)结冰
时长/s水滴平均
体积直径/μm液态水含量
/(g·m−3)a 0 67.1 −23 1500 20 0.50 b 3 67.1 −23 1500 20 0.50 c 3 67.1 −7 1500 20 0.50 表 4 测试集1的MVD与LWC辨识结果及满度误差
Table 4. Test set 1 MVD and LWC identification results and errors
算例 MVDT/μm MVDP/μm MVDe LWCT/(g·m−3) LWCP/(g·m−3) LWCe a 20 20.410 0.51% 0.25 0.251 0.11% b 26 27.891 2.36% 0.61 0.612 0.22% c 26 27.699 2.12% 0.76 0.756 0.44% d 38 38.634 0.79% 0.91 0.898 1.33% e 47 46.579 0.53% 0.76 0.745 1.67% f 56 56.677 0.85% 0.94 0.921 2.11% g 65 65.396 0.50% 0.25 0.251 0.11% h 74 74.582 0.73% 0.40 0.392 0.89% i 89 89.623 0.78% 0.22 0.224 0.44% 表 5 测试集2的MVD与LWC辨识结果及满度误差
Table 5. Test set 2 MVD and LWC identification results and errors
算例 MVDT/μm MVDP/μm MVDe LWCT/(g·m−3) LWCP/(g·m−3) LWCe a 20 22.713 3.39% 0.50 0.528 3.11% b 20 23.504 4.38% 0.50 0.562 6.89% c 20 23.269 4.08% 0.50 0.541 4.56% 表 6 测试集3的试验工况
Table 6. Test set 3 icing conditions
工况 迎角/(°) 来流速度/(m·s−1) 来流温度/(℃) 结冰时长/s 水滴平均体积直径/μm 液态水含量/(g·m−3) a 0 67.1 −23 1500 20 0.50 b 3 67.1 −23 1500 20 0.50 c 3 67.1 −7 1500 20 0.50 表 7 测试集3的MVD与LWC辨识结果及满度误差
Table 7. Test set 3 MVD and LWC identification results and errors
算例 MVDT/μm MVDP/μm MVDe LWCT/(g·m−3) LWCP/(g·m−3) LWCe a 20 24.142 5.17% 0.50 0.552 5.78% b 20 25.129 6.41% 0.50 0.601 11.20% c 20 23.864 4.83% 0.50 0.578 8.67% -
[1] 郭向东,张平涛,赵照,等. 大型结冰风洞云雾场适航应用符合性验证[J]. 航空学报,2020,41(10):123879. doi: 10.7527/S1000-6893.2020.23879GUO X D,ZHANG P T,ZHAO Z,et al. Airworthiness application compliance verification of cloud flowfield in large icing wind tunnel[J]. Acta Aeronautica et Astronautica Sinica,2020,41(10):123879. doi: 10.7527/S1000-6893.2020.23879 [2] 战培国. 结冰风洞校准标准研究[J]. 标准科学,2020(5):87-90. doi: 10.3969/j.issn.1674-5698.2020.05.014ZHAN P G. Research on standards of icing wind tunnel calibration[J]. Standard Science,2020(5):87-90. doi: 10.3969/j.issn.1674-5698.2020.05.014 [3] 周靓,战培国,王梓旭. 结冰风洞云雾场校准标准初探[J]. 标准科学,2020(9):94-98. doi: 10.3969/j.issn.1674-5698.2020.09.019ZHOU L,ZHAN P G,WANG Z X. Preliminary study on calibration standard of cloud field in icing wind tunnel[J]. Standard Science,2020(9):94-98. doi: 10.3969/j.issn.1674-5698.2020.09.019 [4] 易贤,桂业伟,肖春华,等. 结冰风洞液态水含量测量方法研究[J]. 科技导报,2009,27(21):86-90. doi: 10.3321/j.issn:1000-7857.2009.21.018YI X,GUI Y W,XIAO C H,et al. A method of liquid water content measurement in icing wind tunnel[J]. Science & Technology Review,2009,27(21):86-90. doi: 10.3321/j.issn:1000-7857.2009.21.018 [5] 赖庆仁,郭龙,李明,等. 结冰风洞液态水含量测量装置设计与实现[J]. 空气动力学学报,2016,34(6):750-755. doi: 10.7638/kqdlxxb-2015.0137LAI Q R,GUO L,LI M,et al. Design and implementation of the device for liquid water content measurement in icing wind tunnel[J]. Acta Aerodynamica Sinica,2016,34(6):750-755. doi: 10.7638/kqdlxxb-2015.0137 [6] 战培国. 结冰云小水滴粒径测量设备综述[J]. 测控技术,2020,39(6):1-7. doi: 10.19708/j.ckjs.2020.04.217ZHAN P G. Review of measuring instruments for water droplet sizing in icing clouds[J]. Measurement & Control Technology,2020,39(6):1-7. doi: 10.19708/j.ckjs.2020.04.217 [7] HOVENAC E, IDE R. Performance of the forward scattering spectrometer probe in NASA's Icing Research Tunnel[C]// Proc of the 27th Aerospace Sciences Meeting. 1989: 769. doi: 10.2514/6.1989-769 [8] VECCHIONE L, DE MATTEIS P. An overview of the CIRA icing wind tunnel[C]//Proc of the 41st Aerospace Sciences Meeting and Exhibit. 2003: 900. doi: 10.2514/6.2003-900 [9] 王梓旭,沈浩,郭龙,等. 3 m×2 m结冰风洞云雾参数校测方法[J]. 实验流体力学,2018,32(2):61-67. doi: 10.11729/syltlx20170163WANG Z X,SHEN H,GUO L,et al. Cloud calibration method of 3 m×2 m icing wind tunnel[J]. Journal of Experiments in Fluid Mechanics,2018,32(2):61-67. doi: 10.11729/syltlx20170163 [10] AC-9C Aircraft Icing Technology Committee. Calibration and acceptance of icing wind tunnels[R/OL]. SAE International[2022-06-24]. https://www.sae.org/content/arp5905. doi: 10.4271/ARP5905 [11] 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. doi: 10.19678/j.issn.1000-3428.0057370HE J,ZHANG C Q,LI X Z,et al. Survey of research on multimodal fusion technology for deep learning[J]. Computer Engineering,2020,46(5):1-11. doi: 10.19678/j.issn.1000-3428.0057370 [12] 易贤,王强,柴聪聪,等. 基于深度神经网络的飞机结冰冰形预测模型[J]. 南京航空航天大学学报(英文版),2021,38(4):535-544. doi: 10.16356/j.1005-1120.2021.04.001YI X,WANG Q,CHAI C C,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. doi: 10.16356/j.1005-1120.2021.04.001 [13] 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. doi: 10.3969/j.issn.0258-1825.2002.04.009YI X,ZHU G L,WANG K C,et al. Numerically simulating of ice accretion on airfoil[J]. Acta Aerodynamica Sinica,2002,20(4):428-433. doi: 10.3969/j.issn.0258-1825.2002.04.009 [14] 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. doi: 10.13645/j.cnki.f.d.20210811.001CHAI 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. doi: 10.13645/j.cnki.f.d.20210811.001 [15] 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. doi: 10.11729/syltlx20200016CHAI C C,YI X,GUO L,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. doi: 10.11729/syltlx20200016 [16] HUBER P J. Robust estimation of a location parameter[J]. The Annals of Mathematical Statistics,1964,35(1):73-101. doi: 10.1214/aoms/1177703732 [17] JAMES G, WITTEN D, HASTIE T, et al. An introduction to statistical learning: with Applications in R[M]. New York: Springer, 2013. [18] 易贤,郭龙,符澄,等. 结冰风洞试验段水滴分布特性分析[J]. 实验流体力学,2016,30(3):2-7,1. doi: 10.11729/syltlx20160034YI X,GUO L,FU C,et al. Analysis of water droplets distribution in the test section of an icing wind tunnel[J]. Journal of Experiments in Fluid Mechanics,2016,30(3):2-7,1. doi: 10.11729/syltlx20160034 -