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

基于超分辨率重构方法的湍流流场重构

江昊 王伯福 庄启亮 卢志明

江昊,王伯福,庄启亮,等. 基于超分辨率重构方法的湍流流场重构[J]. 实验流体力学,2022,36(3):102-109 doi: 10.11729/syltlx20210185
引用本文: 江昊,王伯福,庄启亮,等. 基于超分辨率重构方法的湍流流场重构[J]. 实验流体力学,2022,36(3):102-109 doi: 10.11729/syltlx20210185
JIANG H,WANG B F,CHONG K L,et al. Reconstruction of turbulent fields based on super-resolution reconstruction method[J]. Journal of Experiments in Fluid Mechanics, 2022,36(3):102-109. doi: 10.11729/syltlx20210185
Citation: JIANG H,WANG B F,CHONG K L,et al. Reconstruction of turbulent fields based on super-resolution reconstruction method[J]. Journal of Experiments in Fluid Mechanics, 2022,36(3):102-109. doi: 10.11729/syltlx20210185

基于超分辨率重构方法的湍流流场重构

doi: 10.11729/syltlx20210185
基金项目: 国家自然科学基金(12072185,11732010,11972220)
详细信息
    作者简介:

    江昊:(1997—),男,安徽黄山人,博士研究生。研究方向:机器学习在湍流中的应用,多相流。通信地址:上海市静安区延长路149号上海大学延长校区力学与工程科学学院上海市应用数学和力学研究所(200072)。E-mail:jiangH@shu.edu.cn

    通讯作者:

    E-mail:王伯福,bofuwang@shu.edu.cn

    庄启亮,klchong@shu.edu.cn

  • 中图分类号: O357

Reconstruction of turbulent fields based on super-resolution reconstruction method

  • 摘要: 从低分辨率流场数据中获取精细流场信息具有重要的研究意义。基于卷积神经网络的超分辨率重构方法是近年来发展的一种较为有效的精细流场重构方法。本文采用高效亚像素卷积神经网络(Efficient Sub-Pixel Convolutional Neural Network,ESPCN),对Rayleigh–Bénard(RB)对流的数值模拟数据和湍流边界层(Turbulent Boundary Layer,TBL)的实验测量数据进行了超分辨率重构,并与双三次插值方法(Bicubic Interpolation)的重构结果进行对比。对比结果表明:在较小的下采样比下,ESPCN方法和Bicubic方法的重构精度相当;在较大的下采样比下,ESPCN方法的重构精度明显优于Bicubic方法。此外,ESPCN方法对数据梯度较大区域的超分辨率重构效果优于Bicubic方法。
  • 图  1  ESPCN模型示意图

    Figure  1.  Sketch of ESPCN

    图  2  下采样比为3的RB系统温度场数据的超分辨率重构结果

    Figure  2.  The super-resolution reconstruction result of the temperature field data of the RB system with rus=3

    图  3  下采样比为30的RB系统温度场数据的超分辨率重构结果

    Figure  3.  The super-resolution reconstruction result of the temperature field data of the RB system with rus=30

    图  4  RB系统中心线温度数据剖面的超分辨率重构结果

    Figure  4.  Comparison of the results of the temperature profile of the horizontal centerline of the RB system temperature field data

    图  5  RB系统瞬时温度场数据PDF分布的超分辨率重构结果

    Figure  5.  Comparison of PDF results of instantaneous temperature field data of the RB system temperature field data

    图  6  湍流边界层实验装置图[30]

    Figure  6.  Sketch of turbulent boundary layer experimental device[30]

    图  7  下采样比为4的TBL流向速度数据的超分辨率重构结果

    Figure  7.  The super-resolution reconstruction result of the TBL streamwise velocity with the down-sampling rus =4

    图  8  下采样比为8的TBL流向速度数据的超分辨率重构结果

    Figure  8.  The super-resolution reconstruction result of the TBL streamwise velocity with the down-sampling rus =8

    图  9  TBL流向速度数据的PDF分布结果对比

    Figure  9.  Comparison of PDF results of TBL streamwise velocity

    表  1  不同下采样比的MSE

    Table  1.   MSE with different down-sampling rus

    rus=3rus=10rus=20rus=30
    ESPCN4.02×10–64.58×10–62.41×10–53.08×10–5
    Bicubic6.97×10–61.95×10–47.11×10–41.39×10–3
    下载: 导出CSV

    表  2  不同下采样比的MSE

    Table  2.   MSE with different down-sampling rus

    rus =2rus =4rus =8
    ESPCN1.63×10–44.32×10–47.00×10–4
    Bicubic1.60×10–47.16×10–43.62×10–3
    下载: 导出CSV
  • [1] ADRIAN R J. Twenty years of particle image velocimetry[J]. Experiments in Fluids,2005,39(2):159-169. doi: 10.1007/s00348-005-0991-7
    [2] WANG Z J,FIDKOWSKI K,ABGRALL R,et al. High-order CFD methods: current status and perspective[J]. International Journal for Numerical Methods in Fluids,2013,72(8):811-845. doi: 10.1002/fld.3767
    [3] 郭中州,何志强,赵文文,等. 高效非结构网格变形与流场插值方法[J]. 航空学报,2018,39(12):126-137. doi: 10.7527/S1000-6893.2018.22411

    GUO Z Z,HE Z Q,ZHAO W W,et al. Efficient mesh deformation and flowfield interpolation method for unstruc-tured mesh[J]. Acta Aeronautica Et Astronautica Sinica,2018,39(12):126-137. doi: 10.7527/S1000-6893.2018.22411
    [4] TÖLKE J,KRAFCZYK M. Second order interpolation of the flow field in the lattice Boltzmann method[J]. Computers & Mathematics With Applications,2009,58(5):898-902. doi: 10.1016/j.camwa.2009.02.012
    [5] DRUAULT P,GUIBERT P,ALIZON F. Use of proper orthogonal decomposition for time interpolation from PIV data[J]. Experiments in Fluids,2005,39(6):1009-1023. doi: 10.1007/s00348-005-0035-3
    [6] GUNES H,RIST U. Spatial resolution enhancement/smoothing of stereo-particle-image-velocimetry data using proper-orthogonal-decomposition-based and Kriging inter-polation methods[J]. Physics of Fluids,2007,19(6):064101. doi: 10.1063/1.2740710
    [7] ROESGEN T. Optimal subpixel interpolation in particle image velocimetry[J]. Experiments in Fluids,2003,35(3):252-256. doi: 10.1007/s00348-003-0627-8
    [8] DUNLOP G R. A rapid computational method for improve-ments to nearest neighbour interpolation[J]. Computers & Mathematics With Applications,1980,6(3):349-353. doi: 10.1016/0898-1221(80)90042-5
    [9] BLU T,THÉVENAZ P,UNSER M. Linear interpolation revitalized[J]. IEEE Transactions on Image Processing,2004,13(5):710-719. doi: 10.1109/tip.2004.826093
    [10] CARLSON R E,FRITSCH F N. Monotone piecewise bicubic interpolation[J]. SIAM Journal on Numerical Analysis,1985,22(2):386-400. doi: 10.1137/0722023
    [11] DONG C, LOY C C, HE K M, et al. Learning a deep convolutional network for image super-resolution[C]//Computer Vision – ECCV 2014, 2014: 184-199. doi: 10.1007/978-3-319-10593-2_13
    [12] DONG C,LOY C C,HE K M,et al. Image super-resolution using deep convolutional networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,38(2):295-307. doi: 10.1109/TPAMI.2015.2439281
    [13] DONG C, LOY C C, TANG X O. Accelerating the super-resolution convolutional neural network[C]//Computer Vision – ECCV 2016, 2016: 391-407. doi: 10.1007/978-3-319-46475-6_25
    [14] SHI W Z, CABALLERO J, HUSZÁR F, et al. Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network[C]// Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition. 2016: 1874-1883. doi: 10.1109/CVPR.2016.207
    [15] LEDIG C, THEIS L, HUSZÁR F, et al. Photo-realistic single image super-resolution using a generative adversarial network[C]// Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. 2017: 105-114. doi: 10.1109/CVPR.2017.19
    [16] HE C X,WANG P,LIU Y Z. Sequential data assimilation of turbulent flow and pressure fields over aerofoil[J]. AIAA Journal,2021,60(2):1091-1103. doi: 10.2514/1.J060697
    [17] DENG Z W,CHEN Y J,LIU Y Z,et al. Time-resolved turbulent velocity field reconstruction using a long short-term memory(LSTM)-based artificial intelligence framework[J]. Physics of Fluids,2019,31(7):075108. doi: 10.1063/1.5111558
    [18] 谢晨月,袁泽龙,王建春,等. 基于人工神经网络的湍流大涡模拟方法[J]. 力学学报,2021,53(1):1-16. doi: 10.6052/0459-1879-20-420

    XIE C Y,YUAN Z L,WANG J C,et al. Artificial neural network-based subgrid-scale models for large-eddy simula-tion of turbulence[J]. Chinese Journal of Theoretical and Applied Mechanics,2021,53(1):1-16. doi: 10.6052/0459-1879-20-420
    [19] XIE C Y,WANG J C,LI H,et al. Artificial neural network mixed model for large eddy simulation of compressible isotropic turbulence[J]. Physics of Fluids,2019,31(8):085112. doi: 10.1063/1.5110788
    [20] LI K,KOU J Q,ZHANG W W. Unsteady aerodynamic reduced-order modeling based on machine learning across multiple airfoils[J]. Aerospace Science and Technology,2021,119:107173. doi: 10.1016/j.ast.2021.107173
    [21] 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):524689-524689. doi: 10.7527/S1000-6893.2020.24689

    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):524689-524689. doi: 10.7527/S1000-6893.2020.24689
    [22] WERHAHN M,XIE Y,CHU M Y,et al. A multi-pass GAN for fluid flow super-resolution[J]. Proceedings of the ACM on Computer Graphics and Interactive Techniques,2019,2(2):1-21. doi: 10.1145/3340251
    [23] FUKAMI K,FUKAGATA K,TAIRA K. Super-resolution reconstruction of turbulent flows with machine learning[J]. Journal of Fluid Mechanics,2019,870:106-120. doi: 10.1017/jfm.2019.238
    [24] FUKAMI K,FUKAGATA K,TAIRA K. Machine-learning-based spatio-temporal super resolution reconstruction of turbulent flows[J]. Journal of Fluid Mechanics,2021,909:A9. doi: 10.1017/jfm.2020.948
    [25] LIU B,TANG J P,HUANG H B,et al. Deep learning methods for super-resolution reconstruction of turbulent flows[J]. Physics of Fluids,2020,32(2):025105. doi: 10.1063/1.5140772
    [26] DENG Z W,HE C X,LIU Y Z,et al. Super-resolution reconstruction of turbulent velocity fields using a generative adversarial network-based artificial intelligence framework[J]. Physics of Fluids,2019,31(12):125111. doi: 10.1063/1.5127031
    [27] BAI K,LI W,DESBRUN M,et al. Dynamic upsampling of smoke through dictionary-based learning[J]. ACM Transactions on Graphics,2021,40(1):1-19. doi: 10.1145/3412360
    [28] GAO H,SUN L N,WANG J X. Super-resolution and denoising of fluid flow using physics-informed convolutional neural networks without high-resolution labels[J]. Physics of Fluids,2021,33(7):073603. doi: 10.1063/5.0054312
    [29] KIM H,KIM J,WON S,et al. Unsupervised deep learning for super-resolution reconstruction of turbulence[J]. Journal of Fluid Mechanics,2021,910:A29. doi: 10.1017/jfm.2020.1028
    [30] SCHRÖDER A,GEISLER R,STAACK K,et al. Eulerian and Lagrangian views of a turbulent boundary layer flow using time-resolved tomographic PIV[J]. Experiments in Fluids,2011,50(4):1071-1091. doi: 10.1007/s00348-010-1014-x
  • 加载中
图(9) / 表(2)
计量
  • 文章访问数:  1828
  • HTML全文浏览量:  239
  • PDF下载量:  172
  • 被引次数: 0
出版历程
  • 收稿日期:  2021-11-29
  • 修回日期:  2022-03-04
  • 录用日期:  2022-03-08
  • 网络出版日期:  2022-04-21
  • 刊出日期:  2022-07-04

目录

    /

    返回文章
    返回

    重要公告

    www.syltlx.com是《实验流体力学》期刊唯一官方网站,其他皆为仿冒。请注意识别。

    《实验流体力学》期刊不收取任何费用。如有组织或个人以我刊名义向作者、读者收取费用,皆为假冒。

    相关真实信息均印刷于《实验流体力学》纸刊。如有任何疑问,请先行致电编辑部咨询并确认,以避免损失。编辑部电话0816-2463376,2463374,2463373。

    请广大读者、作者相互转告,广为宣传!

    感谢大家对《实验流体力学》的支持与厚爱,欢迎继续关注我刊!


    《实验流体力学》编辑部

    2021年8月13日