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ResNet数据驱动的压气机叶型优化

杜周 马玉林 徐全勇 吴锋 冯旭栋

杜周, 马玉林, 徐全勇, 等. 基于ResNet数据驱动的压气机叶型优化[J]. 航空动力学报, 2023, 38(7):1592-1603 doi: 10.13224/j.cnki.jasp.20220611
引用本文: 杜周, 马玉林, 徐全勇, 等. 基于ResNet数据驱动的压气机叶型优化[J]. 航空动力学报, 2023, 38(7):1592-1603 doi: 10.13224/j.cnki.jasp.20220611
DU Zhou, MA Yulin, XU Quanyong, et al. Optimization of compressor blade profile based on ResNet data drive[J]. Journal of Aerospace Power, 2023, 38(7):1592-1603 doi: 10.13224/j.cnki.jasp.20220611
Citation: DU Zhou, MA Yulin, XU Quanyong, et al. Optimization of compressor blade profile based on ResNet data drive[J]. Journal of Aerospace Power, 2023, 38(7):1592-1603 doi: 10.13224/j.cnki.jasp.20220611

基于ResNet数据驱动的压气机叶型优化

doi: 10.13224/j.cnki.jasp.20220611
基金项目: 国家科技重大专项(J2019-Ⅴ-0001-0092,J2019-Ⅴ-0013-0108)
详细信息
    作者简介:

    杜周(1996-),男,博士生,主要从事流体仿真与人工智能相结合的领域研究

    通讯作者:

    徐全勇(1980-),男,副教授,博士,研究领域为航空发动机气动热力学。E-mail:xuquanyong@tsinghua.edu.cn

  • 中图分类号: V231.3

Optimization of compressor blade profile based on ResNet data drive

  • 摘要:

    为提高用于叶型优化设计ResNet深度学习模型的泛化性,对一种适用于亚声速和跨声速的压气机叶型进行了参数化设计。叶型基于Matlab进行几何模型构建,设计变量为叶型最大厚度、最大厚度位置、栅距和叶型尾缘与轴向的夹角,几何模型通过Pointwise软件进行批量网格划分,网格量级为30万,通过OpenFOAM流体仿真软件进行批量计算。最终通过4个设计变量参数化建模后进行仿真得到了叶型流场仿真数据集,该数据集包含22331个叶型仿真算例,可为ResNet深度学习模型提供训练集和测试题,有助于提高模型的泛化性。

     

  • 图 1  ResNet的基本残差单元

    Figure 1.  Basic residual element of ResNet

    图 2  叶型优化设计流程图

    Figure 2.  Flow chart for blade profile optimization design

    图 3  抛物线型中弧线

    Figure 3.  Parabolic-arc mean camberline

    图 4  厚度分布

    Figure 4.  Thickmess distribution

    图 5  3种不同最大厚度位置的叶型厚度分布曲线图

    Figure 5.  Blade profile thickness distribution curves for three different maximum thickness positions

    图 6  3种不同最大厚度位置的叶型轮廓图

    Figure 6.  Blade profile for three different maximum thickness positions

    图 7  3个算例的流体域几何

    Figure 7.  Fluid domain geometry for three examples

    图 8  叶片前缘和尾缘轮廓

    Figure 8.  Profile of leading and trailing edges of blades

    图 9  处理前的非结构网格

    Figure 9.  Unstructured mesh before processing

    图 10  使用T-Rex处理过的流体域网格

    Figure 10.  Fluid domain mesh processed using T-Rex

    图 11  前缘和尾缘的边界层网格

    Figure 11.  Boundary layer mesh of leading and trailing edges

    图 12  OpenFOAM的计算流程

    Figure 12.  Calculation process of OpenFOAM

    图 13  CA-2压气机叶型网格

    Figure 13.  Blade profile mesh of CA-2 compressor

    图 14  CA-2叶型仿真结果

    Figure 14.  Simulation results of CA-2 blade profile

    图 15  CA-2叶型OpenFOAM仿真结果和实验结果的Cp对比

    Figure 15.  Cp comparison of OpenFOAM simulation results and experimental results of CA-2 blade profile

    图 16  3种网格数的叶型Cp分布与网格前缘分布

    Figure 16.  Blade profile Cp distributions and mesh leading edge distributions of three meshes

    图 17  3种网格数的压力云图和马赫数云图

    Figure 17.  Pressure and Mach number clouds of three meshes

    图 18  50万网格每1000步的计算的压力云图

    Figure 18.  Pressure clouds of 500000 meshes per 1000 steps

    图 19  某一案例的压力云图和马赫数云图

    Figure 19.  Pressure and Mach number clouds of a particular case

    表  1  算例变量统计

    Table  1.   Example variable statistics

    变量最小值最大值区间间断点个数
    叶片尾缘角/(°)−1010121
    最大厚度/m0.040.080.0059
    最大厚度位置/m0.20.60.059
    栅距/m0.410.0513
    下载: 导出CSV
  • [1] 秦鹏. 轴流压气机气动设计[M]. 北京: 国防工业出版社, 1975.
    [2] 胡骏. 航空叶片机原理[M]. 北京: 国防工业出版社, 2014.
    [3] 阎军,许琦,张起,等. 人工智能在结构拓扑优化领域的现状与未来趋势[J]. 计算力学学报,2021,38(4): 412-422. doi: 10.7511/jslx20210517401

    YAN Jun,XU Qi,ZHANG Qi,et al. Current and future trends of artificial intelligence in the field of structural topology optimization[J]. Chinese Journal of Computational Mechanics,2021,38(4): 412-422. (in Chinese) doi: 10.7511/jslx20210517401
    [4] 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): 524689.1-524689.46.

    ZHANG Weiwei,KOU Jiaqing,LIU Yilang. Prospect of artificial intelligence empowering fluid mechanics[J]. Acta Aeronautica et Astronautica Sinica,2021,42(4): 524689.1-524689.46. (in Chinese)
    [5] 尹宇辉, 杨普, 张宇飞. 基于机器学习的湍流建模研究[R]. 杭州: 中国力学大会, 2019.
    [6] 张伟伟,朱林阳,刘溢浪,等. 机器学习在湍流模型构建中的应用进展[J]. 空气动力学学报,2019,37(3): 444-454.

    ZHANG Weiwei,ZHU Linyang,LIU Yilang,et al. Progresses in the application of machine learning in turbulence modeling[J]. Acta Aeronamica Sinica,2019,37(3): 444-454. (in Chinese)
    [7] 徐启华,师军. 基于支持向量机的航空发动机故障诊断[J]. 航空动力学报,2005,20(2): 298-302. doi: 10.3969/j.issn.1000-8055.2005.02.025

    XU Qihua,SHI Jun. Aero-engine fault diagnosis based on support vector machine[J]. Journal of Aerospace Power,2005,20(2): 298-302. (in Chinese) doi: 10.3969/j.issn.1000-8055.2005.02.025
    [8] 金东海,梁栋,刘晓恒,等. 航空发动机整机周向平均稳态仿真方法[J]. 航空动力学报,2022,37(11): 2598-2616. doi: 10.13224/j.cnki.jasp.20220279

    JIN Donghai,LIANG Dong,LIU Xiaoheng,et al. Steady state simulation method of whole aero-engine based on circumferentially averaged method[J]. Journal of Aerospace Power,2022,37(11): 2598-2616. (in Chinese) doi: 10.13224/j.cnki.jasp.20220279
    [9] 谢寿生,樊思齐. 自适应变结构神经网络在航空发动机故障诊断上的应用[J]. 航空动力学报,1997,12(4): 367-370. doi: 10.13224/j.cnki.jasp.1997.04.008

    XIE Shousheng,FAN Siqi. Application of adaptive variable neural network to aeroengine failure diagnosis[J]. Journal of Aerospace Power,1997,12(4): 367-370. (in Chinese) doi: 10.13224/j.cnki.jasp.1997.04.008
    [10] FONT B,WEYMOUTH G D,NGUYEN V T,et al. Deep learning of the spanwise-averaged Navier-Stokes equations[J]. Journal of Computational Physics,2021,434(10): 110199.1-110199.25.
    [11] 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)
    [12] 姚琳. 机器学习在翼型优化问题中的应用[J]. 航天标准化,2020,179(1): 43-47. doi: 10.3969/j.issn.1009-234X.2020.01.013

    YAO Lin. Application of machine learning in airfoil optimization problem[J]. Aerospace Standardization,2020,179(1): 43-47. (in Chinese) doi: 10.3969/j.issn.1009-234X.2020.01.013
    [13] 廖鹏,姚磊江,白国栋,等. 基于深度学习的混合翼型前缘压力分布预测[J]. 航空动力学报,2019,34(8): 1751-1758. doi: 10.13224/j.cnki.jasp.2019.08.014

    LIAO Peng,YAO Leijiang,BAI Guodong,et al. Prediction of hybrid airfoil leading edge pressure distribution based on deep learning[J]. Journal of Aerospace Power,2019,34(8): 1751-1758. (in Chinese) doi: 10.13224/j.cnki.jasp.2019.08.014
    [14] 陈海昕,邓凯文,李润泽. 机器学习技术在气动优化中的应用[J]. 航空学报,2019,40(1): 52-68.

    CHEN Haixin,DENG Kaiwen,LI Runze. Utilization of machine learning technology in aerodynamic optimization[J]. Acta Aeronautica et Astronautica Sinica,2019,40(1): 52-68. (in Chinese)
    [15] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[R]. Las Vegas, US: IEEE Conference on Computer Vision and Pattern Recognition, 2016.
    [16] 杜周, 徐全勇, 宋振寿, 等. 基于深度学习的压气机叶型气动特性预测[EB/OL]. [2022-09-23].https: //kns.cnki.net/kcms2/article/abstract?v=3uoqIhG8C45S0n9fL2suRadTyEVl2pW9UrhTDCdPD656v2WvdV0bJ853b6ZyauyhzTAJGro7r6jl6zgN3sn11UPTzegKJzp9&uniplatform=NZKPT.
    [17] AUNGIER R. Axial-flow compressors: a strategy for aerodynamic design and analysis[M]. New York: The American Society of Mechanical Engineers, 2003.
    [18] WENNERSTROM A J. Design of highly loaded axial-flow fans and compressors[M]. Wilder, US: Concepts ETI Incorporation, 2000.
    [19] KULFAN B M. Universal parametric geometry representation method[J]. Journal of Aircraft,2008,45(1): 142-158. doi: 10.2514/1.29958
    [20] FOTTNER L. Test cases for computation of internal flows in aero engine components[M]. Paris, France: Advisory Group for Aerospace Research and Development, 1990.
    [21] STARK U,HOHEISEL H. The combined effect of axial velocity density ratio and aspect ratio on compressor cascade performance[J]. Journal of Engineering for Gas Turbines and Power,1981,103(1): 247-255.
  • 加载中
图(19) / 表(1)
计量
  • 文章访问数:  1219
  • HTML浏览量:  365
  • PDF量:  160
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-08-23
  • 网络出版日期:  2023-04-19

目录

    /

    返回文章
    返回