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

流场特征识别的无量纲时程深度学习方法

战庆亮 葛耀君 白春锦

战庆亮, 葛耀君, 白春锦. 流场特征识别的无量纲时程深度学习方法[J]. 工程力学, 2023, 40(2): 17-24. doi: 10.6052/j.issn.1000-4750.2021.08.0638
引用本文: 战庆亮, 葛耀君, 白春锦. 流场特征识别的无量纲时程深度学习方法[J]. 工程力学, 2023, 40(2): 17-24. doi: 10.6052/j.issn.1000-4750.2021.08.0638
ZHAN Qing-liang, GE Yao-jun, BAI Chun-jin. DEEP LEARNING METHOD FOR FLOW FEATURE RECOGNITION BASED ON DIMENSIONLESS TIME HISTORY[J]. Engineering Mechanics, 2023, 40(2): 17-24. doi: 10.6052/j.issn.1000-4750.2021.08.0638
Citation: ZHAN Qing-liang, GE Yao-jun, BAI Chun-jin. DEEP LEARNING METHOD FOR FLOW FEATURE RECOGNITION BASED ON DIMENSIONLESS TIME HISTORY[J]. Engineering Mechanics, 2023, 40(2): 17-24. doi: 10.6052/j.issn.1000-4750.2021.08.0638

流场特征识别的无量纲时程深度学习方法

doi: 10.6052/j.issn.1000-4750.2021.08.0638
基金项目: 国家自然科学基金项目(51778495,51978527);桥梁结构抗风技术交通行业重点实验室(上海)开放课题项目(KLWRTBMC21-02);辽宁省教育厅研究计划项目(LJKZ0052);中央高校基本科研业务费专项资金项目(3132022189)
详细信息
    作者简介:

    战庆亮(1987−),男,辽宁人,讲师,博士,主要从事计算流体力学研究(E-mail: zhanqingliang@163.com)

    白春锦(1996−),男,辽宁人,硕士生,主要从事计算流体力学研究(E-mail: baichunjin_2020@163.com)

    通讯作者:

    葛耀君(1958−),男,上海人,教授,博士,博导,主要从事结构风工程研究(E-mail: yaojunge@tongji.edu.cn)

  • 中图分类号: O357

DEEP LEARNING METHOD FOR FLOW FEATURE RECOGNITION BASED ON DIMENSIONLESS TIME HISTORY

  • 摘要: 流场的特征直接影响结构的流致振动状态,对结构绕流流场的特征分析具有重要的研究意义。在中高雷诺数与不同流速的情况下,尾流中隐含的流场特征不同,传统数学物理方法很难对其特征进行提取与识别。该文提出了采用无量纲的物理量时程进行流场特征识别的深度学习方法,消除了不同来流速度的影响,仅通过时程的时变特征进行特征识别,扩大了特征识别方法的应用范围。采用两种不同深度学习模型对三种棱柱的尾流进行了特征提取与识别,通过比较可以发现:归一化的时程中仍包含不同形状物体所引起流场的关键特征,可用于流场的特征提取;使用归一化时程进行流场特征识别可降低模型训练难度,又提高了特征提取的精度,是一种流场特征提取的新方法。
  • 图  1  流场时程数据的FCNN结构示意图

    Figure  1.  Architecture of FCNN for flow time history

    图  2  整体计算域及平面网格划分

    Figure  2.  Computation domain and mesh setup

    图  3  流场监测点的布置位置

    Figure  3.  Position of measuring points at wake

    图  4  三棱柱尾流区域典型测点位置处流场参数时程

    Figure  4.  Time history at typical measuring points of flow field parameters

    图  5  模型的损失函数结果

    Figure  5.  Loss value of different cases

    图  6  模型训练过程中的识别精度

    Figure  6.  Recognition accuracy with respect to epoches

    图  7  MLP模型的变量U精度散点图

    Figure  7.  Scatter accuracy map of MLP

    表  1  样本的标签设置

    Table  1.   Label settings for samples

    形状三棱柱方柱六棱柱
    标签123
    变量P, U, V, WP, U, V, WP, U, V, W
    下载: 导出CSV

    表  2  算例设置

    Table  2.   Setup of cases

    算例模型样本参数
    C1MLP原始时程P, U, V, W
    C2MLP归一化时程P, U, V, W
    C3FCNN原始时程P, U, V, W
    C4FCNN归一化时程P, U, V, W
    下载: 导出CSV

    表  3  各算例的特征提取精度

    Table  3.   Results of feature recognition accuracy

    参数C1C2精度提升率C3C4精度提升率/(%)
    P 0.8598 0.9935 13.37 0.9841 0.9991 0.15
    U 0.7361 0.9936 25.75 0.9926 0.9969 0.43
    V 0.9995 0.9992 −0.03 0.9998 0.9975 −0.23
    W 0.5930 0.6568 6.38 0.9652 0.9875 2.23
    下载: 导出CSV
  • [1] 葛耀君. 大跨度桥梁抗风的技术挑战与精细化研究[J]. 工程力学, 2011(增刊 2): 11 − 23.

    GE Yaojun. Technical challenges and refinement research on wind resistance of long-span bridges [J]. Engineering Mechanics, 2011(Suppl 2): 11 − 23. (in Chinese)
    [2] 李永乐, 陈星宇, 汪斌, 等. 扁平箱梁涡激共振阻塞效应及振幅修正[J]. 工程力学, 2018, 35(11): 45 − 52. doi: 10.6052/j.issn.1000-4750.2017.07.0576

    LI Yongle, CHEN Xingyu, WANG Bin, et al. Blockage-effects and amplitude conversion of vortex-induced vibration for flat-box girder [J]. Engineering Mechanics, 2018, 35(11): 45 − 52. (in Chinese) doi: 10.6052/j.issn.1000-4750.2017.07.0576
    [3] 刘剑寒, 马文勇. 旋转圆柱气动力特性风洞试验研究[J]. 工程力学, 2021, 38(增刊 1): 89 − 92. doi: 10.6052/j.issn.1000-4750.2020.05.S016

    LIU Jianhan, MA Wenyong. Wind tunnel test on aerodynamic characteristics of a rotating cylinder [J]. Engineering Mechanics, 2021, 38(Suppl 1): 89 − 92. (in Chinese) doi: 10.6052/j.issn.1000-4750.2020.05.S016
    [4] 刘庆宽, 孙一飞, 张磊杰, 等. 凹痕对斜拉桥斜拉索气动性能影响研究[J]. 工程力学, 2019, 36(增刊 1): 272 − 277. doi: 10.6052/j.issn.1000-4750.2018.05.S053

    LIU Qingkuan, SUN Yifei, ZHANG Leijie, et al. Study on the influence of dent on aerodynamic performance of stay cables of cable-stayed bridge [J]. Engineering Mechanics, 2019, 36(Suppl 1): 272 − 277. (in Chinese) doi: 10.6052/j.issn.1000-4750.2018.05.S053
    [5] 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): 185 − 199.

    YE Shuran, ZHANG Zhen, WANG Yiwei, et al. Progress in deep convolutional neural network based flow field recognition and its applications [J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4): 185 − 199. (in Chinese)
    [6] 王义乾, 桂南. 第三代涡识别方法及其应用综述[J]. 水动力学研究与进展, 2019, 34(4): 413 − 429.

    WANG Yiqian, GUI Nan. A review of the third-generation vortex identification method and its applications [J]. Chinese Journal of Hydrodynamics, 2019, 34(4): 413 − 429. (in Chinese)
    [7] 刘超群. Liutex-涡定义和第三代涡识别方法[J]. 空气动力学学报, 2020, 38(3): 413 − 431. doi: 10.7638/kqdlxxb-2020.0015

    LIU Chaoqun. Liutex-third generation of vortex definition and identification methods [J]. Acta Aerodynamica Sinica, 2020, 38(3): 413 − 431. (in Chinese) doi: 10.7638/kqdlxxb-2020.0015
    [8] 王慧, 郭晨林, 王乐, 等. 基于内积矩阵及深度学习的结构健康监测研究[J]. 工程力学, 2022, 39(2): 14 − 22, 75. doi: 10.6052/j.issn.1000-4750.2020.12.0935

    WANG Hui, GUO Chenlin, WANG Le, et al. Structural health monitoring based inner product matrix and deep learning [J]. Engineering Mechanics, 2022, 39(2): 14 − 22, 75. (in Chinese) doi: 10.6052/j.issn.1000-4750.2020.12.0935
    [9] 叶继红, 杨振宇. 基于生成式对抗网络的风场生成研究[J]. 工程力学, 2021, 38(10): 1 − 11. doi: 10.6052/j.issn.1000-4750.2020.12.0885

    YE Jihong, YANG Zhenyu. Research on generation of wind fields based on generative countermeasure network [J]. Engineering Mechanics, 2021, 38(10): 1 − 11. (in Chinese) doi: 10.6052/j.issn.1000-4750.2020.12.0885
    [10] LING J, KURZAWSKI A, TEMPLETON J. Reynolds averaged turbulence modelling using deep neural networks with embedded invariance [J]. Journal of Fluid Mechanics, 2016, 807: 155 − 166. doi: 10.1017/jfm.2016.615
    [11] MAULIK R, SAN O, JACOB J D, et al. Sub-grid scale model classification and blending through deep learning [J]. Journal of Fluid Mechanics, 2019, 870: 784 − 812. doi: 10.1017/jfm.2019.254
    [12] REN F, WANG C, TANG H. Active control of vortex-induced vibration of a circular cylinder using machine learning [J]. Physics of Fluids, 2019, 31(9): 93601. doi: 10.1063/1.5115258
    [13] HUANG J, LIU H, CAI W. Online in situ prediction of 3-D flame evolution from its history 2-D projections via deep learning [J]. Journal of Fluid Mechanics, 2019, 870: 784 − 812. doi: 10.1017/jfm.2019.545
    [14] ZHANG Y, AZMAN A N, XU K W, et al. Two-phase flow regime identification based on the liquid-phase velocity information and machine learning [J]. Experiments in Fluids, 2020, 61(10): 1 − 16.
    [15] HAN J, TAO J, WANG C. FlowNet: A Deep learning framework for clustering and selection of streamlines and stream surfaces [J]. IEEE Transactions On Visualization and Computer Graphics, 2018, 26(4): 1732 − 1744.
    [16] LIU Y, LU Y, WANG Y, et al. A CNN-based shock detection method in flow visualization [J]. Computers & Fluids, 2019, 184: 1 − 9.
    [17] OMATA N, SHIRAYAMA S. A novel method of low-dimensional representation for temporal behavior of flow fields using deep autoencoder [J]. Aip Advances, 2019, 9(1): 015006.
    [18] STRÖFER C, WU J, XIAO H, et al. Data-driven, physics-based feature extraction from fluid flow fields [J]. Communications in Computational Physics, 2018, 25(3): 625 − 650.
    [19] MURATA T, FUKAMI K, FUKAGATA K. Nonlinear mode decomposition with convolutional neural networks for fluid dynamics [J]. Journal of Fluid Mechanics, 2020: 882.
    [20] 战庆亮, 葛耀君, 白春锦. 基于尾流时程目标识别的流场参数选择研究[J]. 力学学报, 2021, 10(53): 2692 − 2702.

    ZHAN Qingliang, GE Yaojun, BAI Chunjin. Study on flow field parameters of wake time history target recognition [J]. Chinese Journal of Theoretical and Applied Mechanics, 2021, 10(53): 2692 − 2702. (in Chinese)
    [21] 战庆亮,葛耀君,白春锦. 基于深度学习的流场时程特征提取模型[J]. 物理学报, 2022, 71(7): 225 − 234.

    ZHAN Qingliang, GE Yaojun, BAI Chunjin. Flow feature extraction models based on deep learning [J]. Acta Physica Sinica, 2022, 71(7): 225 − 234. (in Chinese)
    [22] KRIZHEVSKY A, SUTSKEVER I, HINTON G E . ImageNet Classification with Deep Convolutional Neural Networks [J]. Communications of the ACM, 2017, 60(6): 84 − 90. doi: 10.1145/3065386
    [23] WANG Z, YAN W, OATES T. Time series classification from scratch with deep neural networks: A strong baseline [C]// International Joint Conference on Neural Networks. Anchorage USA: IEEE, 2017.
    [24] 战庆亮, 周志勇, 葛耀君. Re=3900圆柱绕流的三维大涡模拟[J]. 哈尔滨工业大学学报, 2015, 47(12): 75 − 79.

    ZHAN Qingliang, ZHOU Zhiyong, GE Yaojun. 3-Dimensional large eddy simulation of circular cylinder at Re=3900 [J]. Journal of Harbin Institute of Technology, 2015, 47(12): 75 − 79. (in Chinese)
    [25] KINGMA D, BA J. Adam: A method for stochastic optimization [C]. International Conference on Learning Representations. San Diego CA, ICLR, 2015.
  • 加载中
图(7) / 表(3)
计量
  • 文章访问数:  233
  • HTML全文浏览量:  94
  • PDF下载量:  78
  • 被引次数: 0
出版历程
  • 收稿日期:  2021-08-16
  • 录用日期:  2021-11-18
  • 修回日期:  2021-11-08
  • 网络出版日期:  2021-11-18
  • 刊出日期:  2023-02-01

目录

    /

    返回文章
    返回