基于卷积神经网络的Bolund岛流场重构

李怡锋; 薛明; 吴定聪; 张伟; 倪章松

doi:10.7638/kqdlxxb-2023.0012

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 >

基于卷积神经网络的Bolund岛流场重构

研究论文 | 更新时间：2026-05-28

- 基于卷积神经网络的Bolund岛流场重构
- Flow field reconstruction for Bolund Island based on CNN
- 空气动力学学报 2023年41卷第9期页码：38-47
- 作者机构：
  
  1. 西南科技大学土木工程与建筑学院,绵阳,621010
  2. 成都流体动力创新中心,成都,610072
- 作者简介：
  
  李怡锋（1997-），男，四川攀枝花人，硕士研究生，研究方向：风工程. E-mail：liyifeng@mails.swuet.edu.cn
  倪章松*（1973-），男，安徽太湖人，博士，正高级工程师，研究方向：低速空气动力学，飞机结冰与防除冰技术. E-mail：nzscczx@163.com
- 基金信息：
  
  国家自然科学基金（11902278）
- DOI：10.7638/kqdlxxb-2023.0012
  中图分类号： TP183;P425.3
- 纸质出版：2023
- 稿件说明：
移动端阅览
李怡锋, 薛明, 吴定聪, 等. 基于卷积神经网络的Bolund岛流场重构[J]. 空气动力学学报, 2023, 41(9): 38−47.

LI Yifeng, XUE Ming, WU Dingcong, et al. Flow field reconstruction for Bolund Island based on CNN[J]. 2023, 41(9): 38-47.
李怡锋, 薛明, 吴定聪, 等. 基于卷积神经网络的Bolund岛流场重构[J]. 空气动力学学报, 2023, 41(9): 38−47. DOI： 10.7638/kqdlxxb-2023.0012

LI Yifeng, XUE Ming, WU Dingcong, et al. Flow field reconstruction for Bolund Island based on CNN[J]. 2023, 41(9): 38-47. DOI： 10.7638/kqdlxxb-2023.0012.

摘要

基于卷积神经网络的流场重构方法是风资源评估领域的一种新兴方法，具有成本低、周期短、精度高等优点，在风电场建设评估和风资源利用等领域具有较强的学术价值和工程应用背景。以风资源评估中常用的Bolund岛为研究对象，采用CFD数值模拟的输入风廓线和输出流场为样本标签，构建了以多层融合卷积神经网络为基础的流场重构方法。以缩短计算时间为目标，研究了样本量对流场重构方法精度的影响，同时对重构方法的外插能力进行了分析。研究结果表明：建立的基于卷积神经网络的流场重构方法可以较好地还原流场信息；当重构目标处于训练样本范围内，全场最大重构误差不高于2%，且在样本量从100减小到25时，全场最大重构误差仍在5%以内；构建的重构方法具有一定的外插能力，但其能力有限，当重构目标超出训练样本范围，全场最大重构误差增加到20%以上。

Abstract

Flow field reconstruction based on the convolutional neural network (CNN) is a new method in the field of wind resource assessment. It has the advantages of low cost

short evaluation and high precision for the wind farm construction and wind resource utilization

which provides values in both academic research and engineering practice. By selecting Bolund Island as a case study

the flow field reconstruction method based on a multi-layer fusion convolution neural network has been developed

which adopts sample labels of the input wind profile and the output flow field as used in the CFD. Aiming at shortening the computational time

this study examines the influence of the sample number on the accuracy of the flow field reconstruction method

and analyzes its extrapolation ability in the meantime. The results suggest that the flow field reconstruction method based on CNN can restore the flow field information well. The maximum reconstruction error is less than 2% when the reconstruction target is within the training sample range

and it keeps within 5% when the number of samples is reduced from 100 to 25. The reconstruction method has some extrapolation ability

but is limited within some range. When the reconstruction target is beyond the training sample range

the maximum reconstruction error rises from less than 5% to more than 20%.

关键词

Keywords

references

HAN Y, ZHAO Y, TONG B, et al. Review and outlook of wind resource assessment methods for wind site located at complex terrain[J]. Thermal Power Generation, 2020, 49(11): 139-146. (in Chinese)

DHUNNY A Z, LOLLCHUND M R, RUGHOOPUTH S D D V. Wind energy evaluation for a highly complex terrain using Computational Fluid Dynamics (CFD)[J]. Renewable Energy, 2017, 101: 1-9. DOI: 10.1016/j.renene.2016.08.032

SHARMA P K, WARUDKAR V, AHMED S. Application of a new method to develop a CFD model to analyze wind characteristics for a complex terrain[J]. Sustainable Energy Technologies and Assessments, 2020, 37: 100580. DOI: 10.1016/j.seta.2019.100580

YANG J C, BAO Z H, YU M C. Study on location selection of anemometer tower in mountain wind farm[J]. Wind Energy, 2017(12): 62-67. (in Chinese)

LI Z C, ZHU R, HE X F, et al. Study on the assessment technology of wind energy resource[J]. Acta Meteorologica Sinica, 2007, 65(5): 708-717. (in Chinese)

邱珩, 张双益. 风能资源评估中常用数值气象模式模拟结果的对比分析[J]. 中国电力教育, 2012, (z2): 418-423.

CHEN Z G, LI K, SHI H H, et al. Research on wind energy resource evaluation method of high resolution complex terrain based on nested CFD technology of mesoscale atmospheric model[J]. Wind Energy, 2014(8): 88-92. (in Chinese)

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)

LI R, ZHANG J, ZHAO X. Multi-fidelity modeling of wind farm wakes based on a novel super-fidelity network[J]. Energy Conversion and Management, 2022, 270: 116185. doi: 10.1016/j.enconman.2022.116185

LI R, ZHANG J, ZHAO X. Dynamic wind farm wake modeling based on a bilateral convolutional neural network and high-fidelity LES data[J]. Energy, 2022, 258: 124845. DOI: 10.1016/j.energy.2022.124845

ZHANG J, ZHAO X. Three-dimensional spatiotemporal wind field reconstruction based on physics-informed deep learning[J]. Applied Energy, 2021, 300: 117390. DOI: 10.1016/j.apenergy.2021.117390

VERSTEEG H K, MALALASEKERA W. An introduction to computational fluid dynamics: the finite volume method[M]. 2nd ed. Harlow, England: Pearson Education Ltd. , 2007.

ZHANG X D. Theory and method of flow field calculation in wind farm[M]. Beijing: China Electric Power Press, 2016 (in Chinese).

LI D S, DONG Y B, FU X G, et al. Numerical simulation of bolund island’s influence on wind flow[J]. Hydraulics Pneumatics Seals, 2021, 41(5): 6-9. (in Chinese)

BECHMANN A, SØRENSEN N N, BERG J, et al. The bolund experiment, part II: blind comparison of microscale flow models[J]. Boundary-Layer Meteorology, 2011, 141(2): 245-271. DOI: 10.1007/s10546-011-9637-x

BERG J, MANN J, BECHMANN A, et al. The bolund experiment, part I: flow over a steep, three-dimensional hill[J]. Boundary-Layer Meteorology, 2011, 141(2): 219-243. DOI: 10.1007/s10546-011-9636-y

LIANG S C. Numerical studies of the wind resource characteristic of wind farm on complex terrain[D]. Beijing: North China Electric Power University, 2015 (in Chinese).doi: 10.7666/d.Y2878671

ZHANG X B. CFD numerical simulation of wind field over complex terrain[D]. Hefei: Hefei University of Technology, 2019 (in Chinese).

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

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

基于物理约束深度学习的大跨柔性光伏阵列绕流场重构

基于深度学习的超分辨率重构方法在CAARC标模绕流流场重构中的应用

基于深度学习的非定常周期性流动预测方法

基于深度神经网络的粒子图像测速算法

基于U-Net的磁浮飞行风洞气动力干扰剥离算法