基于物理约束深度学习的大跨柔性光伏阵列绕流场重构

张春伟; 柯世堂; 王伯洋; 余玮; 梁珂; 马文勇

doi:10.7638/kqdlxxb-2024.0191

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 >

基于物理约束深度学习的大跨柔性光伏阵列绕流场重构

研究专栏——新能源结构风工程 | 更新时间：2026-05-28

- 基于物理约束深度学习的大跨柔性光伏阵列绕流场重构
- Reconstruction of flow fields around large-span flexible PV array based on physics-informed deep learning
- 空气动力学学报 2025年43卷第4期页码：91-102
- 作者机构：
  
  1. 南京航空航天大学民航学院,南京,210016
  2. 南京航空航天大学土木工程动力多灾害防护江苏高校重点实验室,南京,210016
  3. 石家庄铁道大学土木工程学院,石家庄,050043
- 作者简介：
  
  张春伟（1993—），男，博士研究生，研究方向：风工程与结构抗风. E-mail：zcw19930212@sina.com
  柯世堂*（1982—），男，博士，教授，研究方向：风工程与结构抗风. E-mail：keshitang@163.com
- 基金信息：
  
  国家自然科学基金NSFC-RGC合作研究重点项目（52321165649）;江苏省杰出青年科学基金（BK20211518）;江苏省研究生科研与实践创新计划项目（KYCX22_0374）
- DOI：10.7638/kqdlxxb-2024.0191
  中图分类号： TU312+.1;TK81
- 纸质出版：2025
- 稿件说明：
移动端阅览
张春伟, 柯世堂, 王伯洋, 等. 基于物理约束深度学习的大跨柔性光伏阵列绕流场重构[J]. 空气动力学学报, 2025, 43(4): 91−102.

ZHANG Chunwei, KE Shitang, WANG Boyang, et al. Reconstruction of flow fields around large-span flexible PV array based on physics-informed deep learning[J]. 2025, 43(4): 91-102.
张春伟, 柯世堂, 王伯洋, 等. 基于物理约束深度学习的大跨柔性光伏阵列绕流场重构[J]. 空气动力学学报, 2025, 43(4): 91−102. DOI： 10.7638/kqdlxxb-2024.0191

ZHANG Chunwei, KE Shitang, WANG Boyang, et al. Reconstruction of flow fields around large-span flexible PV array based on physics-informed deep learning[J]. 2025, 43(4): 91-102. DOI： 10.7638/kqdlxxb-2024.0191.

摘要

大跨柔性光伏阵列因双向串联结构引发的排间复杂三维绕流效应是导致其发生风致破坏的重要原因之一，传统物理风洞试验难以直接捕捉排间干扰下流场流动全过程的分布特性，而深度学习方法为实现复杂流场的精确重构与预测提供了新思路。为了重构大跨柔性光伏阵列的速度场和压力场，以国电投江苏盐城柔性光伏示范基地40 m跨度五排三跨光伏阵列为研究对象，通过大涡模拟获取大跨柔性光伏阵列脉动风场数据，在此基础上提出了一种损失函数嵌入N-S方程和连续性方程约束的全连接神经网络深度学习方法，建立了数据驱动模型和数据-物理双驱动模型。结果表明：相较于数据驱动模型，数据-物理双驱动模型可以更加精确地捕捉大跨柔性光伏阵列绕流场特征，其中前两排及第四排光伏板后侧、上下缘及尾流区流向速度重构误差分别降低60.2%和36.6%，光伏板上下缘横向速度重构误差降低53.7%，流向速度和横向速度的全场重构误差分别为16.6%和18.5%；当损失函数缺乏压力项时，数据驱动模型无法从训练数据中捕捉压力信息，而数据-物理双驱动模型的N-S方程和连续性方程可引导模型通过速度场信息求解压力场，得到的排间绕流区压力场平均重构误差仅为16.1%。本研究为风荷载作用下复杂结构绕流场智能重构新方法提供了参考。

Abstract

The complex three-dimensional inter-row flow interference induced by bidirectional series of large-span flexible PV arrays is the primary reason for the wind-induced damage. Wind tunnel tests are difficult to capture the flow fields subjected to inter-row interference. However

deep learning methods offer promising solutions for accurate reconstruction and prediction of complex flow fields. The present study focuses on the 5-row by 3-span PV array with a span of 40 meters

located at the State Power Investment Group Flexible PV Demonstration Base in Yancheng

Jiangsu. Large-eddy simulations of wind fields were performed to provide training data for a fully connected neural network deep learning method that incorporated the loss function into Navier-Stokes equations. Based on this innovative method

a data-driven model and a data-physics dual driven model was established for accurately reconstructing the velocity and pressure fields around an array of large-span flexible PV. Compared to the data-driven model

the data-physics dual driven model demonstrates superior efficacy in capturing the flow characteristics. Specifically

the streamwise velocity reconstruction errors upstream of the first two rows and the fourth row

at the upper and lower edges of the PV panels and in the wake region

are reduced by 60.2% and 36.6%

respectively. The reconstruction errors of the lateral velocity at the upper and lower edges of the PV panels are reduced by 53.7%. Moreover

the model has full-field reconstruction errors of 16.6% and 18.5% for the streamwise and lateral velocities

respectively. Meanwhile

when the pressure term is not considered in the loss function

the data-driven model cannot learn pressure information from the training data

while the data-physics dual driven model can obtain the pressure field through the velocity field

yielding an average reconstruction error of the inter-row pressure fields being only 16.1%. This study provides a reference for a new intelligent reconstruction method for the flow fields around complex structures under wind loads.

关键词

Keywords

references

LI W J, KE S T, CAI Z B, et al. Instability mechanism and failure criteria of large-span flexible PV support arrays under severe wind[J]. Solar Energy, 2023, 264: 112000. doi: 10.1016/j.solener.2023.112000

CAI Q G, KE S T, WANG L S, et al. Evolution of wind-induced vibration form of large-span flexible PV aeroelastic arrays[J]. Solar Energy, 2024, 277: 112684. doi: 10.1016/j.solener.2024.112684

YEMENICI O, AKSOY M O. An experimental and numerical study of wind effects on a ground-mounted solar panel at different panel tilt angles and wind directions[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2021, 213: 104630. doi: 10.1016/j.jweia.2021.104630

WANG J X, YANG Q S, TAMURA Y. Effects of building parameters on wind loads on flat-roof-mounted solar arrays[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2018, 174: 210−224. doi: 10.1016/j.jweia.2017.12.023

Ministry of Housing and Urban-Rural Development of the People's Republic of China. Load code for the design of building structures: GB 50009—2012[S]. Beijing: China Architecture Building Press, 2012(in Chinese).

中国电机工程学会. 柔性光伏支架结构设计规程: T/CSEE 0394-2023[S]. 2023.

American Society of Civil Engineers. Minimum design loads and associated criteria for buildings and other structures[M]. American Society of Civil Engineers, 2017. doi: 10.1061/9780784414248

WANG H, KE S T. Research on non-stationary characteristics of wind-induced vibration of extra-large cooling tower in coastal area based on field measurements[J]. Journal of Vibration and Shock, 2020, 39(10): 206−214 (in Chinese).

ZHANG X B, MA W Y, KANG X H, et al. Experimental study of the torsional aeroelastic instability of single-axis solar trackers under different turbulence intensities[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2023, 240: 105486. doi: 10.1016/j.jweia.2023.105486

SUÁREZ J L, CADENAS D, RUBIO H, et al. Vortex shedding dynamics behind a single solar PV panel over a range of tilt angles in uniform flow[J]. Fluids, 2022, 7(10): 322. doi: 10.3390/fluids7100322

XIONG H L, EGUSQUIZA M, ALBERG ØSTERGAARD P, et al. Multi-objective optimization of a hydro-wind-photovoltaic power complementary plant with a vibration avoidance strategy[J]. Applied Energy, 2021, 301: 117459. doi: 10.1016/j.apenergy.2021.117459

WATINE Y, GABILLET C, LOSSOUARN B, et al. Vortex-induced vibrations of a cantilevered blunt plate: POD of TR-PIV measurements and structural modal analysis[J]. Journal of Fluids and Structures, 2023, 117: 103832. doi: 10.1016/j.jfluidstructs.2022.103832

MIN X W, CHEN W L. Influence of gap thickness on aerodynamic forces and flow structures of circular cylinders under passive-suction-jet control[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2024, 246: 105677. doi: 10.1016/j.jweia.2024.105677

DU M H, TIAN L L, ZHU C L. IDDES simulations of flow separation around the iced airfoil and its modal analysis[J]. Acta Aerodynamica Sinica, 2023, 41(11): 46−55(in Chinese). doi: 10.7638/kqdlxxb-2022.0140

ZENG P P, CHEN S S, FENG C, et al. Effect of lateral jet on local aerodynamic heating characteristics of missile rudders[J]. Acta Aerodynamica Sinica, 2023, 41(9): 70−81(in Chinese). doi: 10.7638/kqdlxxb-2022.0168

LIU S J, ZHANG Z, LIU Q K. Analysis of unsteady flow field around square cylinders based on dynamic modal decomposition[J]. Engineering Mechanics, 2024, 41(S1): 326−331(in Chinese). doi: 10.6052/j.issn.1000-4750.2023.05.S016

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

SRINIVASAN P A, GUASTONI L, AZIZPOUR H, et al. Predictions of turbulent shear flows using deep neural networks[J]. Physical Review Fluids, 2019, 4(5): 054603. doi: 10.1103/physrevfluids.4.054603

ESTEPHAN J, GAN CHOWDHURY A, IRWIN P. A new experimental-numerical approach to estimate peak wind loads on roof-mounted photovoltaic systems by incorporating inflow turbulence and dynamic effects[J]. Engineering Structures, 2022, 252: 113739. doi: 10.1016/j.engstruct.2021.113739

ALRAWASHDEH H, STATHOPOULOS T. Experimental investigation of the wind loading on solar panels: Effects of clearance off flat roofs[J]. Journal of Structural Engineering, 2022, 148(12): 04022202. doi: 10.1061/jsendh.steng-10957

RAISSI M, PERDIKARIS P, KARNIADAKIS G E. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations[J]. Journal of Computational Physics, 2019, 378: 686−707. doi: 10.1016/j.jcp.2018.10.045

CHEN Z, LIU Y, SUN H. Physics-informed learning of governing equations from scarce data[J]. Nature Communications, 2021, 12(1): 6136. doi: 10.1038/s41467-021-26434-1

LOU Q, MENG X H, KARNIADAKIS G E. Physics-informed neural networks for solving forward and inverse flow problems via the Boltzmann-BGK formulation[J ] . Journal of Computational Physics, 2021, 447: 110676. doi: 10.1016/j.jcp.2021.110676

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

基于卷积神经网络的Bolund岛流场重构

基于U-Net的磁浮飞行风洞气动力干扰剥离算法

基于条件扩散模型的火箭气动外形快速设计方法

定常来流下钝体二维组合断面气动特性与流场的智能预测方法

基于深度神经网络的高速翼型流场降阶模型