基于物理模型与机器学习求解叶栅流场的降阶模型

卢宇峰; 戴韧

doi:10.19805/j.cnki.jcspe.2026.250467

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 >

基于物理模型与机器学习求解叶栅流场的降阶模型

动力设备与系统 | 更新时间：2026-09-15

- 基于物理模型与机器学习求解叶栅流场的降阶模型
- Reduced-order Model for Solution of Cascade Flow Field Based on Physical Model and Machine Learning
- 动力工程学报 2026年46卷第9期页码：49-58
- 作者机构：
- 作者简介：
  
  卢宇峰(2001—),男,浙江台州人,硕士研究生,主要从事叶轮机械气体动力学方面的研究。戴韧(通信作者),男,教授,博士生导师,博士,E-mail:dairen@usst.edu.cn。
- 基金信息：
  
  国家自然科学基金资助项目(52276034）;国家科技重大专项资助项目（2017-V-0016-0069）
- DOI：10.19805/j.cnki.jcspe.2026.250467
  中图分类号： TK474
- 纸质出版：2026
- 稿件说明：
移动端阅览
卢宇峰, 戴韧. 基于物理模型与机器学习求解叶栅流场的降阶模型[J]. 动力工程学报, 2026, 46(9): 49-58.
卢宇峰, 戴韧. 基于物理模型与机器学习求解叶栅流场的降阶模型[J]. 动力工程学报, 2026, 46(9): 49-58. DOI： 10.19805/j.cnki.jcspe.2026.250467.

DOI： 10.19805/j.cnki.jcspe.2026.250467

摘要

提出一种求解叶栅三维流场的降阶模型，将RANS方程降阶为对应的Euler方程与黏性修正源项，求解Euler方程获得无黏的基本三维流场，应用多层感知神经网络，预测不同工况和几何参数条件下的黏性源项，修正Euler解。应用模型分别求解rotor37和1.5级Aachen透平流场。结果表明：机器学习修正方法不需要大量的学习样本，将Euler解的精度提高了一个量级，很好地反映了叶栅三维流动特征，对降低叶栅优化计算的硬件要求，提升叶轮机械通流优化设计的计算速度具有很好的应用价值。

Abstract

A reduced-order model for solving three-dimensional flow fields in cascades was proposed

which reduced RANS equations to corresponding Euler equations coupled with viscous correction source terms. The inviscid base three-dimensional flow field was obtained by solving Euler equations. A multilayer perceptron neural network was then employed to predict the viscous source terms under various operating conditions and geometric parameters

thereby correcting the Euler solutions. The model was applied to solve the flow fields of Rotor 37 and a 1.5-stage Aachen turbine. Results show that the machine learning correction method

without requiring extensive training samples

improves the accuracy of the Euler solutions by an order of magnitude

accurately capturing the three-dimensional flow characteristics within the cascade. This approach holds significant application value for reducing hardware requirements and enhancing computational speed for cascade optimization computations and throughflow design of turbomachinery.

关键词

Keywords

references

WU C H. A general theory of three-dimensional flow in subsonic and supersonic turbomachines of axial, radial, and mixed-flow types[J]. Journal of Fluids Engineering, 1952, 74(8): 1363-1380.

DENTON J D. The use of a distributed body force to simulate viscous effects in 3D flow calculations[C]//ASME 1986 International Gas Turbine Conference and Exhibit. Dusseldorf, West Germany: ASME, 1986.

张枚. 叶轮机械多叶片排三维粘性流动计算方法的研究[D]. 北京: 中国科学院工程热物理研究所, 2008.ZHANG Mei. The numerical study of three dimensional viscous flow of multi-blade rows in turbomachinery[D]. Beijing: The Institute of Engineering Thermophysics, Chinese Academy of Sciences, 2008.

陈晨铭, 郭雪岩, 常林森. 基于BP神经网络代理模型的翼型优化及领域自适应研究[J]. 动力工程学报, 2022, 42(7): 657-663.CHEN Chenming, GUO Xueyan, CHANG Linsen. Research on airfoil optimization and domain adaptation based on BP neural network surrogate model[J]. Journal of Chinese Society of Power Engineering, 2022, 42(7): 657-663.

尤孙钰, 靳智昊, 董祥瑞, 等. 基于深度强化学习的多攻角翼型流动控制研究[J]. 动力工程学报, 2024, 44(12): 1887-1897.YOU Sunyu, JIN Zhihao, DONG Xiangrui, et al. Flow control of multi-attack angle airfoils based on deep reinforcement learning[J]. Journal of Chinese Society of Power Engineering, 2024, 44(12): 1887-1897.

SEKAR V, JIANG Qinghua, SHU Chang, et al. Fast flow field prediction over airfoils using deep learning approach[J]. Physics of Fluids, 2019, 31(5): 057103.

WANG Jing, HE Cheng, LI Runze, et al. Flow field prediction of supercritical airfoils via variational autoencoder based deep learning framework[J]. Physics of Fluids, 2021, 33(8): 086108.

DURU C, ALEMDAR H, BARAN O U. A deep learning approach for the transonic flow field predictions around airfoils[J]. Computers & Fluids, 2022, 236: 105312.

叶旖茵, 韩巍, 李济超, 等. 基于机理修正的小型燃气轮机变工况神经网络模型构建方法研究[J]. 动力工程学报, 2025, 45(4): 505-513.YE Yiyin, HAN Wei, LI Jichao, et al. Research on construction method of neural network model for small gas turbine under variable working conditions based on mechanism modification[J]. Journal of Chinese Society of Power Engineering, 2025, 45(4): 505-513.

宋家豪, 曹文博, 张伟伟. FD-PINN: 频域物理信息神经网络[J]. 力学学报, 2023, 55(5): 1195-1205.SONG Jiahao, CAO Wenbo, ZHANG Weiwei. FD-PINN: frequency domain physics-informed neural network[J]. Chinese Journal of Theoretical and Applied Mechanics, 2023, 55(5): 1195-1205.

冯云阳, 宋西镇, 袁巍, 等. 基于物理嵌入神经网络的叶栅损失模型[J]. 航空动力学报, 2023, 38(7): 1615-1625.FENG Yunyang, SONG Xizhen, YUAN Wei, et al. Physics-informed neural networks based cascade loss model[J]. Journal of Aerospace Power, 2023, 38(7): 1615-1625.

张振, 苏欣荣, 袁新. 基于物理信息神经网络的气膜冷却湍流模型反演学习[J]. 动力工程学报, 2024, 44(9): 1459-1465.ZHANG Zhen, SU Xinrong, YUAN Xin. PINN-based inversion learning of turbulence model for film cooling[J]. Journal of Chinese Society of Power Engineering, 2024, 44(9): 1459-1465.

SPODNIAK M, SEMRD K, DRAGANOV K. Turbine blade temperature field prediction using the numerical methods[J]. Applied Sciences, 2021, 11(6): 2870.

吴跃腾, 巴顿, 杜娟, 等. 基于深度注意力网络的压气机流场重构方法[J]. 航空学报, 2024, 45(24): 630580.WU Yueteng, BA Dun, DU Juan, et al. Compressor flow field reconstruction method based on deep attention networks[J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(24): 630580.

崔晟, 苏纬仪, 孙斐, 等. E3高压涡轮三维流场快速预测方法研究[J]. 推进技术, 2022, 43(11): 54-66.CUI Sheng, SU Weiyi, SUN Fei, et al. Rapid prediction method of three dimensional flow field for E3 high pressure turbine[J]. Journal of Propulsion Technology, 2022, 43(11): 54-66.

QIAO Tianyang, HU Chenxing, YANG Ce. Reduced-order prediction of unsteady spatial-temporal aerodynamics in a turbine cascade[J]. Aerospace Science and Technology, 2024, 145: 108896.

董锦昊, 董锦洋, 唐怀平. 基于POD-神经网络的格尼襟翼压力分布预测[J]. 力学学报, 2025, 57(4): 829-842.DONG Jinhao, DONG Jinyang, TANG Huaiping. Prediction of pressure distribution on Gurney flaps based on POD-neural network model[J]. Chinese Journal of Theoretical and Applied Mechanics, 2025, 57(4): 829-842.

XU Lianchao, LIU Zhengxian, LI Xiaojian, et al. Dynamic mode characteristics of flow instabilities in a centrifugal compressor impeller[J]. Aerospace Science and Technology, 2023, 142: 108707.

NAKAMURA T, FUKAMI K, HASEGAWA K, et al. Convolutional neural network and long short-term memory based reduced order surrogate for minimal turbulent channel flow[J]. Physics of Fluids, 2021, 33(2): 025116.

ZHANG Weiwei, XIA Mingkun, KOU Jiaqing. A scientometric investigation of artificial intelligence for fluid mechanics: emerging topics and active groups[J]. Progress in Aerospace Sciences, 2025, 157: 101130.

EMUNDS R, JENNIONS I K, BOHN D, et al. The computation of adjacent blade-row effects in a 1.5-stage axial flow turbine[J]. Journal of Turbomachinery, 1999, 121(1): 1-10.

GIUNTA A, WOJTKIEWICZ S, ELDRED M. Overview of modern design of experiments methods for computational simulations[C]//41st Aerospace Sciences Meeting and Exhibit. Reno, Nevada: AIAA, 2003: 649.

HIRSCH C. Numerical computation of internal and external flows: the fundamentals of computational fluid dynamics[M]. London, UK: Elsevier Limited, 2007.

LONGO J M A. Simulation of complex inviscid and viscous vortex flow[C]//Fluid Dynamics of High Angle of Attack: IUTAM Symposium Tokyo. Tokyo: Springer, 1993: 363-373.

ZHANG Haideng, WU Yun, LI Yinghong. Evaluation of RANS turbulence models in simulating the corner separation of a high-speed compressor cascade[J]. Engineering Applications of Computational Fluid Mechanics, 2015, 9(1): 477-489.

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

风功率预测关键技术及其研究应用综述

重型燃气轮机动叶片蠕变寿命预测的轻量化建模方法

基于物理单调最小二乘支持向量机的锅炉燃烧NOx生成浓度预测方法

基于国产流程模拟软件的气流床气化炉专业模型建模

基于T-S模糊模型的风力机系统事件触发控制