EI、Scopus 收录
中文核心期刊

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

高超声速风洞短时气动力智能辨识算法研究

王钦超 李世超 高宏力 马贵林 伍广 段志琴

王钦超, 李世超, 高宏力, 马贵林, 伍广, 段志琴. 高超声速风洞短时气动力智能辨识算法研究. 力学学报, 待出版 doi: 10.6052/0459-1879-21-484
引用本文: 王钦超, 李世超, 高宏力, 马贵林, 伍广, 段志琴. 高超声速风洞短时气动力智能辨识算法研究. 力学学报, 待出版 doi: 10.6052/0459-1879-21-484
Wang Qinchao, Li Shichao, Gao Hongli, Ma Guilin, Wu Guang, Duan Zhiqin. Research on intelligent identification algorithms for short-term aerodynamics of hypersonic wind tunnels. Chinese Journal of Theoretical and Applied Mechanics, in press doi: 10.6052/0459-1879-21-484
Citation: Wang Qinchao, Li Shichao, Gao Hongli, Ma Guilin, Wu Guang, Duan Zhiqin. Research on intelligent identification algorithms for short-term aerodynamics of hypersonic wind tunnels. Chinese Journal of Theoretical and Applied Mechanics, in press doi: 10.6052/0459-1879-21-484

高超声速风洞短时气动力智能辨识算法研究

doi: 10.6052/0459-1879-21-484
基金项目: 国家自然科学基金项目(52105562)资助
详细信息
    作者简介:

    李世超, 助理研究员, 主要研究方向: 高速结构动力学 . E-mail: Docterlsc9077@swjtu.edu.cn

  • 中图分类号: V211.751

Research on Intelligent Identification Algorithms for Short-term Aerodynamics of Hypersonic Wind Tunnels

  • 摘要: 风洞测力试验是高超声速飞行器研发的重要环节, 随着这项技术的发展, 试验模型的大尺度化成为高超声速风洞试验的趋势. 在几百毫秒的有效测试时间内, 大尺度测力系统刚度减弱等问题会严重导致气动力辨识精度变差, 试验模型大尺度化对短时脉冲燃烧风洞精确气动力辨识带来了挑战. 对此本文提出了一种新的基于传统信号处理结合深度学习的智能气动力辨识算法, 该框架分解两个主要阶段: (1)信号分解(2)数据训练. 其中信号分解阶段通过变分模态分解将原始数据分解为不同模态子信号, 随后通过Pearson相关性分析筛除干扰子信号; 在训练阶段通过深度学习模型提取训练数据集中含有有效特征的子信号, 最终得到真实气动力信号. 此外, 为增强算法的鲁棒性, 在算法框架不同阶段通过不同方法对算法中的超参数进行优化得出最优参数组合. 此算法在气动力辨识精度以及抗干扰等方面都得到了比较理想的结果. 通过悬挂测力实验台进行验证, 结果表明该算法可以有效滤除由大尺度模型带来的传统方法难以消除的干扰分量. 最后应用于脉冲燃烧风洞的大尺度模型测力系统, 气动力辨识精度得到有效提高.

     

  • 图  1  LSTM和GRU网络模块结构

    Figure  1.  LSTM and GRU network module structure

    图  2  基于VMD-CNN-GRU的气动力信号提取算法流程

    Figure  2.  Aerodynamic signal extraction algorithm flow based on VMD-CNN-GRU

    图  3  阶跃载荷训练样本采集系统

    Figure  3.  Step load training sample acquisition system

    图  4  天平以及总压输出信号

    Figure  4.  Total pressure and balance output signals

    表  1  CNN-GRU深度学习模型参数

    Table  1.   Neural network structure and parameters

    NumberNetwork structureparameter settingsOutput size
    1DenseUnit Nodes:128/64(2000, 128)
    2Conv1 DNumber of kernel: 128(2000, 128)
    3MaxpoolingPool size: 2
    Padding: ‘same’
    (1000, 128)
    4GRUUnit Nodes: 128/64(64)
    5DenseUnit Nodes: 128(64)
    6DropoutRate: 0.2(64)
    下载: 导出CSV

    表  2  悬挂测力实验台与风洞悬挂测力系统对比

    Table  2.   Comparison of Suspension Force Measuring Test bench with Suspension Force Measuring System

    Suspension force measurement systemSuspension test bench
    Structure compositionModels, tie rods and fixtures, support frames, sensorsModels, tie rods and fixtures, bases, sensors
    Material00Ni19Co8Mo5TiAl00Ni19Co8Mo5TiAl
    Load channel$ \begin{array}{l}Fx、Fy、Fz、\\ Mx、My、Mz\end{array} $$ \begin{array}{l}Fx、Fy、Fz、\\ Mx、My、Mz\end{array} $
    Loading methodForm of a step loadRelease the weight to apply a step load impact
    下载: 导出CSV

    表  3  不同算法的数据误差验证

    Table  3.   validation data errors of VCG m, MEAN and Fourier method

    AlgorithmMAEMSERMSE
    Filter-mean0.210.090.28
    FFT0.390.170.41
    Transfer Function0.230.090.28
    CNN0.310.110.37
    VMD-CNN-GRU0.210.060.25
    下载: 导出CSV
  • [1] Squire LC. A REVIEW OF THE ROLE OF SOME SMALL HIGH-SPEED WIND TUNNELS IN AERONAUTICAL RESEARCH. Progress in Aerospace Sciences, 1998, 34(3-4): 107-166 doi: 10.1016/S0376-0421(98)00001-3
    [2] Wiriadidjaja S, F Hasim, Mansor S, et al. Subsonic wind tunnels in Malaysia. 2012
    [3] 乐嘉陵, 刘伟雄, 贺伟等. 脉冲燃烧风洞及其在火箭和超燃发动机研究中的应用. 实验流体力学, 2005(1): 1-10 (Le Jialing, Liu Weixiong, He Wei, et al. Impulse combustion wind tunnel and its application in rocket and scramjet research. , 2005(1): 1-10 (in Chinese)
    [4] 刘伟雄, 谭宇, 毛雄兵等. 一种新运行方式脉冲燃烧风洞研制及初步应用. 实验流体力学, 2007, 21(4): 59-64 (Liu Weixiong, Tan Yu, Mao Xiongbin, et al. The development and preliminary application of a pulse combustion wind tunnel with new running way. , 2007, 21(4): 59-64 (in Chinese)
    [5] 程忠宇, 陈宏, 张琦. 多加速度计振动分离惯性补偿测力技术. 实验流体力学, 1999, 13(4): 57-61 (Cheng Zhongyu, Cheng Hong, Zhang Qi. Inertia compensation technology based on multi-accelerometer vibration separating. Experiments and Measurements In Fluid Mechanics, 1999, 13(4): 57-61 (in Chinese)
    [6] 刘伟雄, 谭宇, 毛雄兵, 等. 一种新运行方式脉冲燃烧风洞研制及初步应用. 实验流体力学, 2007, 21(4): 59-64 (Liu Weixiong, Tan Yu, Mao Xiongbin, et al. The development and preliminary application of a pulse combustion wind tunnel with new running way. Journal of Experiments in Fluid Mechanics, 2007, 21(4): 59-64 (in Chinese)
    [7] Joshi MV, Reddy NM. Aerodynamic force measurements over missile configurations in IISc shock tunnel at M∞ = 5. 5[J]. Experiments in Fluids, 1986, 4(6): 338-340 doi: 10.1007/BF00266300
    [8] Menezes V, Saravanan S, Reddy K. Shock tunnel study of spiked aerodynamic bodies flying at hypersonic Mach numbers. Shock Waves, 2002, 12(3): 197-204 doi: 10.1007/s00193-002-0160-3
    [9] Sahoo N, Mahapatra DR, Jagadeesh G, et al. An accelerometer balance system for measurement of aerodynamic force coefficients over blunt bodies in a hypersonic shock tunnel. Measurement Science & Technology, 2003, 14(3): 260
    [10] Abdel-Jawad MM, Mee DJ, Morgan RG. New calibration technique for multiple-component stress wave force balances. Review of Scientific Instruments, 2007, 78(6): 670
    [11] Smith AL, Mee DJ, Daniel W. Design, modeling and analysis of a six-component force balance for hypervelocity wind tunnel testing. Computers & Structures, 2001, 79(11): 1077-1088
    [12] Wang Y, Liu Y, Luo C, et al. Force measurement using the strain-gauge balance in a shock tunnel with long test duration. Review of Scientific Instruments, 2016, 87(5): 1068
    [13] Wang Y, Liu Y, Jiang Z. Design of a pulse-type strain gauge balance for a long-test-duration hypersonic shock tunnel. Shock Waves, 2016, 26(6): 1-10
    [14] Li S, Li K, Liu B, et al. A new dynamic modeling methodology of a force measuring system for hypersonic impulse wind tunnel. Measurement, 2020, 164: 108012 doi: 10.1016/j.measurement.2020.108012
    [15] S. Li, Z. You, H. Gao, Q. Wang, G. Wu and G. Ma, "Force measurement and support integrated device in hypersonic wind tunnel, " in IEEE Transactions on Instrumentation and Measurement,doi: 10.1109/TIM.2021.3129495.
    [16] Mizuno T, Ishino Y, Takasaki M. Fabrication of a Three-dimensional Force Measurement System Us-ing Double Series Magnetic Suspension. Ifac Paper-sonline, 2016, 49(21): 536-540 doi: 10.1016/j.ifacol.2016.10.657
    [17] Laurence S, Schramm J M, Hannemann K. Force and Moment Measurements on a Free-Flying Capsule Model in a High-Enthalpy Shock Tunnel[C]// 28th Aerodynamic Measurement Technology, Ground Test-ing, and Flight Testing Conference. DLR, 2012
    [18] 王锋, 任虎, 周正, 等. 载荷辨识方法用于脉冲风洞模型阻力测量研究. 振动与冲击, 2015, 000(24): 202-208 (Wang Feng, Ren Hu, Zhou Zheng, et al, et al. Drag force measurement in impulse facilities by using load identification method. Journal of Vibration and Shock, 2015, 000(24): 202-208 (in Chinese)
    [19] 王锋, 武龙, 张小庆等. 动态载荷辨识问题的正则化求解新方法. 振动工程学报, 2016, 29(1): 31-37 (Wang Feng, Wu Long, Zhang Xiaoqing, et al. A technique for solving dynamical force identification problems by Tikhonov regularization method. Journal of Vibration Engineering, 2016, 29(1): 31-37 (in Chinese)
    [20] 汪运鹏, 杨瑞鑫, 聂少军等. 基于深度学习技术的激波风洞智能测力系统研究. 力学学报, 2020(5) (Wang Yunpeng, Yang Ruixin, Nie Shaojun, et al. Deep-learning-based intelligent force measurement system using in a shock tunnel. Chinese Journal of Theoretical and Applied Mechanics, 2020(5) (in Chinese)
    [21] Luo, Changtong, Wang, et al. Wave system fitting: A new method for force measurements in shock tunnels with long test duration[J]. Mechanical Systems & Signal Processing, 2015
    [22] 吴子牛, 白晨媛, 李娟等. 高超声速飞行器流动特征分析. 航空学报, 2015, 36(1): 58-85 (Wu Ziniu, Bai Chenjuan, Li Juan, et al. Analysis of flow characteristics for hypersonic vehicle. Acta Aeronautica ET Astronautica Sinica, 2015, 36(1): 58-85 (in Chinese)
    [23] Mee DJ. Dynamic Calibration of Force Balances. Shock Waves, 2003, 12(6): 443-455 doi: 10.1007/s00193-003-0181-6
    [24] Dragomiretskiy K, Zosso D. Variational Mode De-composition. IEEE Transactions on Signal Pro-cessing, 2014, 62(3): 531-544 doi: 10.1109/TSP.2013.2288675
    [25] Yandong LI, Hao Z, Lei H. Survey of convolutional neural network[J]. Journal of Computer Applications, 2016
    [26] Hochreiter S, Schmidhuber J. Long Short-Term Memory. Neural Computation, 1997, 9(8): 1735-1780 doi: 10.1162/neco.1997.9.8.1735
    [27] Buda A, Jarynowski A. Life time of correlations and its applications. [Vol. 1] [M]. ABRASCO-Associ-ação Brasileira de Saúde Coletiva, 2010
    [28] Afrasiabi M, Mohammadi M, Rastegar M, et al. Advanced Deep Learning Approach for Probabilistic Wind Speed Forecasting. IEEE Transactions on Industrial Informatics, 2020, PP(99): 1-1
    [29] Lecun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015, 521(7553): 436 doi: 10.1038/nature14539
    [30] Mao Z, Jung TP, Lin CT, et al. Predicting EEG Sample Size Required for Classification Calibration[C]// International Conference on Augmented Cognition. Springer International Publishing, 2016.
    [31] Cho J, Lee K, Shin E, et al. How much data is needed to train a medical image deep learning system to achieve necessary high accuracy?[J]. Computer Science, 2016
  • 加载中
图(4) / 表(3)
计量
  • 文章访问数:  37
  • HTML全文浏览量:  10
  • PDF下载量:  10
  • 被引次数: 0
出版历程
  • 录用日期:  2022-01-07
  • 网络出版日期:  2022-01-07

目录

    /

    返回文章
    返回