机器学习辅助湍流建模在分离流预测中的应用

尹宇辉; 李浩然; 张宇飞; 陈海昕

doi:10.7638/kqdlxxb-2020.0155

欢迎来到能源电力期刊网！

English Version 期刊登录

您当前的位置：

首页 >

文章列表页 >

机器学习辅助湍流建模在分离流预测中的应用

研究论文 | 更新时间：2026-05-28

- 机器学习辅助湍流建模在分离流预测中的应用
- Application of machine learning assisted turbulence modeling in flow separation prediction
- 空气动力学学报 2021年39卷第2期页码：23-32
- 作者机构：
  
  清华大学航天航空学院,北京,100084
- 作者简介：
  
  尹宇辉（1996-），男，陕西宝鸡人，博士研究生，研究方向：计算流体力学，湍流模拟. E-mail：yinyh17@mails.tsinghua.edu.cn
  张宇飞*（1983-），副教授，研究方向：计算流体力学，湍流模拟. E-mail：zhangyufei@tsinghua.edu.cn
- 基金信息：
  
  国家自然科学基金（91852108，11872230，92052203）
- DOI：10.7638/kqdlxxb-2020.0155
  中图分类号： TP181
- 纸质出版：2021
- 稿件说明：
移动端阅览
尹宇辉, 李浩然, 张宇飞, 等. 机器学习辅助湍流建模在分离流预测中的应用[J]. 空气动力学学报, 2021, 39(2): 23−32.

YIN Yuhui, LI Haoran, ZHANG Yufei, et al. Application of machine learning assisted turbulence modeling in flow separation prediction[J]. 2021, 39(2): 23-32.
尹宇辉, 李浩然, 张宇飞, 等. 机器学习辅助湍流建模在分离流预测中的应用[J]. 空气动力学学报, 2021, 39(2): 23−32. DOI： 10.7638/kqdlxxb-2020.0155

YIN Yuhui, LI Haoran, ZHANG Yufei, et al. Application of machine learning assisted turbulence modeling in flow separation prediction[J]. 2021, 39(2): 23-32. DOI： 10.7638/kqdlxxb-2020.0155.

摘要

数据驱动湍流建模是近年来发展的提高雷诺平均N-S方程预测精度的有效手段，通过机器学习算法能够从高置信度数据中自动提取特征，建立准确的从平均流动特征到雷诺应力的预测模型。针对高雷诺数下积冰翼型绕流这一类典型的复杂流动分离问题，基于此前研究者提出的机器学习预测框架，从输入输出特征选择和翼型绕流中数据分布特性两个方面出发，对机器学习预测结果的光滑性和准确性进行改善。提出了基于雷诺应力张量分析和流动特征辨识的输入特征选择准则；提出了局部区域建模方法以及基准模型和机器学习预测模型混合的代入计算方法。将改进方法应用于积冰翼型绕流问题之中，结果表明改进的方法能够准确给出训练集和预测集上的雷诺应力结果，并且代入平均流计算可以得到和真实分布更加接近的流动和机翼表面压力分布。

Abstract

Data-driven turbulence modeling has been considered as an effective method for improving the prediction accuracy of the Reynolds-averaged Navier-Stokes equations. By using machine learning algorithms such as the artificial neural network

features can be automatically extracted from high-fidelity data

and accurate prediction models from the mean flow characteristics to the Reynolds stress can be established. In this study

focused on the ice-accretion airfoil with the typical complex flow separation phenomenon under high Reynolds numbers

efforts are made in two aspects

i.e.

the input and output feature selection and the data distribution characteristics of flow around an airfoil

to improve the smoothness and accuracy of the machine learning assisted prediction results. An input feature selection criterion based on the Reynolds stress tensor analysis and the flow characteristics identification is proposed

and a zonal modeling method together with a hybrid approach of the benchmark model and the machine learning assisted prediction model is also proposed. Results obtained from the improved method show that the Reynolds stress on both the training dataset and the prediction dataset can be accurately predicted

and the flow separation and pressure distribution on the airfoil agree well with the real situation.

关键词

Keywords

references

CRAFT T J, LAUNDER B E, SUGA K. Development and application of a cubic eddy-viscosity model of turbulence[J]. International Journal of Heat and Fluid Flow, 1996, 17(2): 108-115.DOI: 10.1016/0142-727X(95)00079-6[LinkOut]

CELIC A, HIRSCHEL E H. Comparison of eddy-viscosity turbulence models in flows with adverse pressure gradient[J]. AIAA Journal, 2006, 44(10): 2156-2169.DOI: 10.2514/1.14902[LinkOut]

SPALART P R. Philosophies and fallacies in turbulence modeling[J]. Progress in Aerospace Sciences, 2015, 74: 1-15.DOI: 10.1016/j.paerosci.2014.12.004[LinkOut]

SHE Z S, TANG F, XIAO M. Structural ensemble dynamics theory for engineering turbulence models[J]. Acta Aerodynamica Sinica, 2019, 37(1): 1-18. (in Chinese)DOI: 10.7638/kqdlxxb-2018.0249[万方]

SPEZIALE C G. On nonlinear K-l and K-ε models of turbulence[J]. Journal of Fluid Mechanics, 1987, 178: 459-475.DOI: 10.1017/s0022112087001319[LinkOut]

BREUER M, PELLER N, RAPP C, et al. Flow over periodic hills - Numerical and experimental study in a wide range of Reynolds numbers[J]. Computers Fluids, 2009, 38(2): 433-457.DOI: 10.1016/j.compfluid.2008.05.002[LinkOut]

SPALART P R. Strategies for turbulence modelling and simulations[J]. International Journal of Heat and Fluid Flow, 2000, 21(3): 252-263.DOI: 10.1016/S0142-727X(00)00007-2[LinkOut]

DURAISAMY K, IACCARINO G, XIAO H. Turbulence modeling in the age of data[J]. Annual Review of Fluid Mechanics, 2019, 51(1): 357-377.DOI: 10.1146/annurev-fluid-010518-040547[LinkOut]

ZHANG W W, ZHU L Y, LIU Y L, et al. Progresses in the application of machine learning in turbulence modeling[J]. Acta Aerodynamica Sinica, 2019, 37(3): 444-454. (in Chinese)[万方]

EMORY M, LARSSON J, IACCARINO G. Modeling of structural uncertainties in Reynolds-averaged Navier-Stokes closures[J]. Physics of Fluids, 2013, 25(11): 1155.

XIAO H, WANG J X, GHANEM R G. A random matrix approach for quantifying model-form uncertainties in turbulence modeling[J]. Computer Methods in Applied Mechanics and Engineering, 2017, 313: 941-965.DOI: 10.1016/j.cma.2016.10.025[LinkOut]

LING J, TEMPLETON J. Evaluation of machine learning algorithms for prediction of regions of high Reynolds averaged Navier Stokes uncertainty[J]. Physics of Fluids, 2015, 27(8): 085103.DOI: 10.1063/1.4927765[LinkOut]

EDELING W N, CINNELLA P, DWIGHT R P. Predictive RANS simulations via Bayesian model-scenario averaging[J]. Journal of Computational Physics, 2014, 275: 65-91. doi: 10.1016/j.jcp.2014.06.052[LinkOut]

DURAISAMY K, ZHANG Z J, SINGH A P. New approaches in turbulence and transition modeling using data-driven techniques[C ] // 53rd AIAA Aerospace Sciences Meeting, Kissimmee, Florida. Reston, Virginia: AIAA, 2015. doi: 10.2514/6.2015-1284

SINGH A P, MEDIDA S, DURAISAMY K. Machine-learning-augmented predictive modeling of turbulent separated flows over airfoils[J]. AIAA Journal, 2017, 55(7): 2215-2227.DOI: 10.2514/1.J055595[LinkOut]

XIAO H, WU J L, WANG J X, et al. Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier-Stokes simulations: a data-driven, physics-informed Bayesian approach[J]. Journal of Computational Physics, 2016, 324: 115-136. doi: 10.1016/j.jcp.2016.07.038[LinkOut]

TRACEY B D, DURAISAMY K, ALONSO J J. A machine learning strategy to assist turbulence model development[C ] // 53rd AIAA Aerospace Sciences Meeting, Kissimmee, Florida. Reston, Virginia: AIAA, 2015. doi: 10.2514/6.2015-1287

ZHU L Y, ZHANG W W, KOU J Q, et al. Machine learning methods for turbulence modeling in subsonic flows around airfoils[J]. Physics of Fluids, 2019, 31(1): 015105.DOI: 10.1063/1.5061693[LinkOut]

SINGH A P, DURAISAMY K. Using field inversion to quantify functional errors in turbulence closures[J]. Physics of Fluids, 2016, 28(4): 045110.DOI: 10.1063/1.4947045[LinkOut]

PARISH E J, DURAISAMY K. A paradigm for data-driven predictive modeling using field inversion and machine learning[J]. Journal of Computational Physics, 2016, 305: 758-774.DOI: 10.1016/j.jcp.2015.11.012[LinkOut]

LING J L, 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[LinkOut]

SCHMELZER M, DWIGHT R P, CINNELLA P. Discovery of algebraic Reynolds-stress models using sparse symbolic regression[J]. Flow, Turbulence and Combustion, 2020, 104(2-3): 579-603.DOI: 10.1007/s10494-019-00089-x[LinkOut]

WANG J X, WU J L, XIAO H. Physics-informed machine learning approach for reconstructing Reynolds stress modeling discrepancies based on DNS data[J]. Physical Review Fluids, 2017, 2(3): 1-22.

WU J L, XIAO H, PATERSON E. Physics-informed machine learning approach for augmenting turbulence models: A omprehensive framework[J]. Physical Review Fluids, 2018, 3(7): 074602.DOI: 10.1103/physrevfluids.3.074602[LinkOut]

YIN Y H, YANG P, ZHANG Y F, et al. Feature selection and processing of turbulence modeling based on an artificial neural network[J]. Physics of Fluids, 2020, 32(10): 105117.DOI: 10.1063/5.0022561[LinkOut]

LI H R, ZHANG Y F, CHEN H X. Aerodynamic prediction of iced airfoils based on modified three-equation turbulence model[J]. AIAA Journal, 2020, 58(9): 3863-3876.DOI: 10.2514/1.J059206[LinkOut]

TRACEY B, DURAISAMY K, ALONSO J. Application of supervised learning to quantify uncertainties in turbulence and combustion modeling[C ] // 51st AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition, Grapevine (Dal las/Ft. Worth Region), Texas. Reston, Virginia: AIAA, 2013. doi: 10.2514/6.2013-259

LING J L, JONES R, TEMPLETON J. Machine learning strategies for systems with invariance properties[J]. Journal of Computational Physics, 2016, 318: 22-35. doi: 10.1016/j.jcp.2016.05.003[LinkOut]

POPE S B. A more general effective-viscosity hypothesis[J]. Journal of Fluid Mechanics, 1975, 72(2): 331-340.DOI: 10.1017/s0022112075003382[LinkOut]

SPALART P R, DECK S, SHUR M L, et al. A new version of detached-eddy simulation, resistant to ambiguous grid densities[J]. Theoretical and Computational Fluid Dynamics, 2006, 20(3): 181-195.DOI: 10.1007/s00162-006-0015-0[LinkOut]

RISTORCELLI J R. Diagnostic Statistics for the assessment and characterization of complex turbulent flows[R]. NASA-CR 198221, 1995.

ADDY H, BROEREN A, ZOECKLER J, et al. A wind tunnel study of icing effects on a business jet airfoil[C ] // 41st Aerospace Sciences Meeting and Exhibit, Reno, Nevada. Reston, Virginia: AIAA, 2003. doi: 10.2514/6.2003-727

PEDREGOSA F, VAROQUAUX G, GRAMFORT A, et al. Scikit-learn: Machine learning in Python[J]. Journal of Machine Learning Research, 2011, 12(1): 2825−2830.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

考虑近距离海洋下垫面影响的对称风机翼型静气动特性

基于条件扩散模型的火箭气动外形快速设计方法

基于机器学习的极端风场短时预测研究进展与思考

大型结构风效应流固耦合机器学习研究进展

基于机器学习的房颤病人左心耳血流动力学研究