时间分辨平面PIV与单点速度时间序列有限观测下局部湍流流场重建物理依据、机理引导方法及适用范围:湍流边界层等参考湍流流动
Taylor冻结假设的适用性检验、随机扫掠修正与非侵入时空超采样
本组文献共同处理"时间信号能否等价为空间结构匀速对流"这一第1章最基础的物理前提:既有在管道流、光学湍流、地面层湍流与跨声速/高超场景下对TFTH的实测检验与误差来源分解,也有以随机扫掠(sweeping)假设、双特征时间等替代模型对冻结假设的扩展修正,还包括用当地瞬时速度替代平均对流速度的非侵入时空超采样测量流程。它们给出冻结假设成立的范围、失效率与可替代假设,为后面各组讨论尺度相关对流速度、恢复空间范围与时间窗口提供前提条件。
- Validation of Taylor’s Frozen Hypothesis for DAS-Based Flow(Shu-Chao Dai, Lei Liang, Ke Jiang, Hui Wang, Chen Zhong, 2025, Italian National Conference on Sensors)
- Spatiotemporal statistics of optical turbulence beyond Taylor's frozen turbulence hypothesis.(Darío G. Pérez, Hishan Farfán-Bachiloglu, M. Sepúlveda, 2024, Journal of The Optical Society of America A-optics Image Science and Vision)
- Beyond Taylor's frozen turbulence hypothesis in ground-layer turbulence(Darío G. Pérez, I. R. González, Hishan Farfán, 2023, Remote Sensing)
- Extended Taylor Frozen-Turbulence Hypothesis and Aero-Optical Phase Disturbances(S. Prasad, 2017, No journal)
- Non-Invasive Flow Measurement Technology Based on Taylor's Frozen Turbulence Hypothesis(Aleksey Krivonogov, 2021, International Conference on Industrial Engineering, Applications and Manufacturing)
- Beyond Taylor’s hypothesis: a novel volumetric reconstruction of velocity and density fields for variable-density and shear flows(Dominique Fratantonio, Chris C. K. Lai, J. Charonko, K. Prestridge, 2021, Experiments in Fluids)
- Application of Taylor's hypothesis to an unsteady convective field for the spectral analysis of turbulence in the aorta.(Takami Yamaguchi, S. Kikkawa, K. H. Parker, 1984, Journal of Biomechanics)
- Implication of Taylor’s hypothesis on measuring flow modulation(Xiang Yang, Michael F. Howland, 2017, Journal of Fluid Mechanics)
- Sweeping Effects Modify Taylor’s Frozen Turbulence Hypothesis for Scalars in the Roughness Sublayer(K. Everard, G. Katul, G. Lawrence, A. Christen, M. Parlange, 2021, Geophysical Research Letters)
尺度与壁法向位置相关的对流速度、壁法向速度相关与各向异性时空失相关标度
本组聚焦冻结假设中的关键参数本身:涡对流速度的定义与选取方式、其随尺度和壁法向位置的变化、极高雷诺数(含VLSM)下的内/外尺度标度,以及由两点时空相关函数与各向异性空间相关刻画的结构输运速率与失相关长度。它们回答"在多长的时间滞后、多大的空间邻域内,单点信号仍与周围流场保持确定性联系",是界定可恢复空间范围、尺度与时间跨度的定量依据,而非仅仅检验假设的好坏。
- EDDY CONVECTION VELOCITY AND TAYLOR'S HYPOTHESIS OF 'FROZEN' TURBULENCE IN A ROUGH-BED OPEN-CHANNEL FLOW(V. Nikora, D. Goring, 2000, Journal of hydroscience and hydraulic engineering)
- Evaluation of convective velocity choice in 3D velocity field reconstruction with Taylor’s hypothesis(S. Sheppard, J. Farnsworth, J. G. Brasseur, 2026, Experiments in Fluids)
- Wall-Normal Velocity Correlations in a Zero Pressure Gradient Turbulent Boundary Layer(Sylvain Morilhat, F. Chedevergne, F. Micheli, F. Simon, 2020, Journal of Fluids Engineering-transactions of The Asme)
- On the convection velocity in high-Reynolds-number pipe flow(Noriyuki Furuichi, Marie Ono, 2025, Journal of Fluid Mechanics)
- Space-time correlation measurements in a turbulent natural convection boundary layer(R. Cheesewright, K. Doan, 1978, International Journal of Heat and Mass Transfer)
- Development and Assessment of Non-Isotropic Spatial Resolution in PIV(F. Scarano, 2004, No journal)
TRPIV与热线实测的时空相关结构、内—外区相干及二阶/高阶应力交叉谱
本组以实验手段直接测量壁湍流时空相关与时空结构:时间分辨PIV对近壁时空相关场的直接测量、不同壁法向位置的时空相关与相干函数、内—外区大尺度信号相干幅值与相位的实验刻画、湍流时空结构的实验观测(含一篇DFD会议摘要性工作),以及由二阶/四阶应力—速度交叉谱推导的对流速度差与相干相似(CS)模型(随壁法向距离由单瓣向三瓣转变的时空耦合结构)。它们把前述标度关系落实为壁湍流(尤其近壁区)可用的定量时空耦合图像,是机理引导重建最直接的数据支撑。
- Structure of Velocity Space‐Time Correlations in a Boundary Layer(A. Favre, J. Gaviglio, R. Dumas, 1967, Physics of Fluids)
- Space-time correlations of near-wall velocity and pressure fluctuations in wall-resolved and wall-modeled large-eddy simulation(Guo-Qing Fan, Han-Qiao Han, Wei-Wen Zhao, De-Jun Wan, 2025, Journal of Hydrodynamics)
- TRPIV investigation of space-time correlation in turbulent flows over flat and wavy walls(Wei Wang, X. Guan, N. Jiang, 2014, Acta Mechanica Sinica)
- Experiment on Spatio-Temporal Turbulent Structures of a Natural Convection Boundary Layer(T. Tsuji, Y. Nagano, M. Tagawa, 1992, Journal of Heat Transfer-transactions of The Asme)
- Abstract Submitted for the DFD17 Meeting of The American Physical Society Implication of Taylor’s hypothesis on amplitude modulation1(Michael F. Howland, Xiang Yang, 2017, No journal)
- Experimental study on space–time correlations of velocity and passive scalar fields in turbulent pipe flows(Hui-Xin Li, Duo Xu, Guo-Wei He, 2025, Journal of Fluid Mechanics)
- The approximate similarity between higher-order and second-order stress–velocity cross-spectra and its significance for the convection velocity of Reynolds stress fluctuations(Huiying Zhang, Ting Wu, De-Zhi Ning, Guo-Wei He, 2025, Journal of Fluid Mechanics)
壁压、标量、可压缩与自然对流中的时空特性:亚对流成分、Schmidt数与Mach数效应
本组把时空耦合规律推广到速度场之外的观测量与更宽工况:壁压脉动的波数—频率交叉谱及其对流/亚对流成分来源识别、标量Schmidt数对对流速度与时空结构的影响、Mach数对时空相关的作用、自然对流边界层的时空结构。它们既用于判断"单点压力/标量/非壁面传感器信号能约束什么",也给出不同介质与可压缩性下对流速度与失相关规律的可迁移性边界,避免把不可压中性壁湍流结论简单外推。
- Identification of Sources of Wall Pressure Fluctuations Using Space–Time Pressure–Velocity Correlations(Humza Butt, Shishir Damani, W. Devenport, K. Lowe, 2026, AIAA Journal)
- Identification of Sources of Sub-Convective Wall Pressure Fluctuations Using Space-Time Pressure-Velocity Correlations(Humza Butt, Shishir Damani, W. Devenport, Todd Lowe, 2024, Advanced Courses)
- Space-time characteristics of wall pressure fluctuations on the surface of flexible extended cylinder(V. Voskoboinick, V. Grinchenko, O. Voskoboinyk, A. Voskobiinyk, 2025, Hydrodynamics and acoustics)
- The effect of Schmidt number on space–time correlation of scalar fluctuations in wall-bounded shear flow(Hui-Xin Li, Guo-Wei He, Duo Xu, 2026, Journal of Fluid Mechanics)
- Effects of Mach number on space-time characteristics of wall pressure fluctuations beneath turbulent boundary layers(Xingsheng Sun, Peng-Jun-Yi Zhang, Kun Zhao, Z. Wan, De-Jun Sun, 2024, The Physics of Fluids)
- Space-time structure of a turbulent natural convection boundary layer.(T. Tsuji, Y. Nagano, M. Tagawa, M. Aoyama, 1990, Transactions of the Japan Society of Mechanical Engineers. B)
线性/谱随机估计与条件统计框架:单点及少量探针时序可约束的低维分量与条件平均本质
本组是"少量点观测→周围场"的经典统计估计主线:线性随机估计(LSE)及其谱形式(尺度相关增益与相位核)、内—外区大尺度信号由对数区单点预测、Navier–Stokes线性模型(随机强迫+涡黏耗散)跨壁法向位置的估计及其模型项随壁距与尺度必须变化的证据、条件高斯框架给出的条件统计解析解与数据同化思路、缸内流实时全场估计、采样条件下识别方法与LSE的组合策略,以及涡量场到速度场的卷积反演。它们共同界定单点信息只能约束低维/大尺度相干分量,输出本质上是条件平均或统计相似流场而非瞬时真实结构,并量化观测位置、信号分量与稀疏度的影响。
- On spectral linear stochastic estimation(C. Tinney, F. Coiffet, J. Delville, André M. Hall, P. Jordan, M. Glauser, 2006, Experiments in Fluids)
- An Extended View of the Inner-outer Interaction Model for Wall-bounded Turbulence Using Spectral Linear Stochastic Estimation(I. Marusic, W. Baars, N. Hutchins, 2015, Procedia Engineering)
- Linking correlations and structure - Stochastic estimation and conditional averaging(R. Adrian, 1990, No journal)
- Towards In-Cylinder Flow Informed Engine Control Strategies Using Linear Stochastic Estimation(Daniel Butcher, A. Spencer, 2019, SAE technical paper series)
- Linear-model-based estimation in wall turbulence: improved stochastic forcing and eddy viscosity terms(Vikrant Gupta, Anagha Madhusudanan, M. Wan, S. J. Illingworth, M. Juniper, 2021, Journal of Fluid Mechanics)
- Conditional Gaussian Systems for Multiscale Nonlinear Stochastic Systems: Prediction, State Estimation and Uncertainty Quantification(N. Chen, A. Majda, 2018, Entropy)
- Combination of linear stochastic estimation and proper orthogonal decomposition: application in two-phase slug flow(S. Munir, Abdul Rashid Abdul Aziz, M. Heikal, M. I. Siddiqui, 2023, Journal of the Brazilian Society of Mechanical Sciences and Engineering)
- A new identification method in sampled quadrant analysis for wall-bounded turbulence(Yipeng Shi, Zhenhua Xia, Shi-Yi Chen, 2016, Physics of Fluids)
- Spectral stochastic estimation of high-Reynolds-number wall-bounded turbulence for a refined inner-outer interaction model(W. Baars, N. Hutchins, I. Marusic, 2016, No journal)
- Vortex-to-velocity reconstruction for wall-bounded turbulence via the field-based linear stochastic estimation(Cheng-Yue Wang, Qi Gao, Biao Wang, C. Pan, Jin-Jun Wang, 2021, Journal of Fluid Mechanics)
相位平均相干结构与POD—压缩感知降阶稀疏重构
本组从"相干结构"与"降阶模态"两个角度把稀疏观测压缩到低维表示:相位平均提取周期/相干结构并给出湍流特性(明确区分相干结构与随机脉动),以及POD模态配合压缩感知与传感器布置实现稀疏重构。它们代表不依赖神经网络的确定性降阶路线,其可恢复自由度上限、传感器位置敏感性与条件平均造成的细节丢失,是讨论"测量信息不足还是模型能力不足"的量化参照。
- Phase-averaged measurements of the turbulence properties in the near wake of a circular cylinder at high Reynolds number by 2C-PIV and 3C-PIV(R. Perrin, E. Cid, S. Cazin, A. Sevrain, M. Braza, F. Moradei, G. Harran, 2006, Experiments in Fluids)
- A novel thermal turbulence reconstruction method using proper orthogonal decomposition and compressed sensing coupled based on improved particle swarm optimization for sensor arrangement(Zhenhuan Zhang, Xiu-Yan Gao, Qi-Xiang Chen, Yuan Yuan, 2024, The Physics of Fluids)
近壁渐近标度、对数律与边界层分区作为硬约束嵌入的物理信息网络
本组代表"具体湍流机理引导"而非泛化方程约束:把对数律、von Kármán常数与截距、对数层上下界、黏性底层划分等近壁标度写成硬编码约束或分段物理先验,直接作用于网络结构、分区加权损失与特征表示;也包括边界层位置未知时以边界识别迭代训练替代固定分区,以及仅由y+>30统计量重建近壁场与动量输运的渐近硬约束方案。其价值在于避免网络在黏性底层施加错误物理、保证剖面及其导数的物理一致,同时暴露出参数标定依赖与适用雷诺数/壁法向区间的失效条件。
- Turbulent channel flow: A physics-informed neural network approach with embedded parameter optimization(A. Palasis, G. Sofiadis, F. Sofos, A. Liakopoulos, 2025, The Physics of Fluids)
- A unified hard-constraint physics-informed neural network approach for sparse-observation airflow-field reconstruction under complex boundary conditions(Gang Jing, Huan Wang, Xian-Ting Li, Gui-Jin Wang, Jian Gao, 2026, Engineering applications of artificial intelligence)
- Near-wall flow reconstruction in turbulent channel flow using a physics-informed neural network(Xiao-Jing Cui, Z. Yao, 2026, The Physics of Fluids)
- Solving ultra-high-load, low-pressure turbine cascade flow using physics-informed neural networks with boundary identification strategy(Ruo-Yu Chen, Zi-Liang Li, Jianshe Zhang, Hang Yuan, Xin'gen Lu, 2025, The Physics of Fluids)
以方程残差与PDE先验为约束的边界层时空重建、压力推断与不确定度量化
本组以N-S/能量方程残差或PDE先验作为主要物理约束,重点在湍流边界层类流动:稀疏数据下的时空重建与压力推断、RANS残差惩罚配合内部测点/界面数据分布对二阶统计量的影响、PINN与LSTM在时空外推上的物理保真度对比、贝叶斯神经网络给出带不确定度的物理一致预测、域分解分布式PINN的压力规范固定、速度到温度场的方程约束反演,以及PINN与扩展卡尔曼滤波耦合的平均量重建。通过与上一组对照,可分离"一般方程残差约束"与"具体机理引导"各自的独立收益与失效条件。
- Physics-informed neural network for turbulent flow reconstruction in composite porous-fluid systems(Seohee Jang, M. Jadidi, S. Rezaeiravesh, Alistair Revell, Y. Mahmoudi, 2024, Machine Learning: Science and Technology)
- Deep learning for turbulent boundary layers: spatio-temporal reconstruction and prediction(X. Dong, Wei Zhou, T. Cai, X. Cai, 2026, Proceedings of the 22nd International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics)
- Bayesian Neural Networks With Physics-Informed Priors With Application to Boundary Layer Velocity(Luca Menicali, David H. Richter, S. Castruccio, 2025, IEEE Transactions on Neural Networks and Learning Systems)
- Physics-informed deep learning for turbulent boundary layers: Spatio-temporal reconstruction and prediction(J. Zhu, Xiang-Rui Dong, Wu Zhou, Lin Ding, T. Cai, Yin-Lin Dong, 2026, Journal of Hydrodynamics)
- Distributed physics-informed neural networks via domain decomposition for fast flow reconstruction(Yi-Xiao Qian, Jia-Xu Liu, Zewei Xia, Song Chen, Chao Xu, Sheng-Ze Cai, 2026, arXiv.org)
- Flow temperature field reconstruction based on physics-informed neural network(Yu-Bao Yang, Yang Xu, Hong-Ping Wang, Yiping Liu, Jin-Jun Wang, 2025, The Physics of Fluids)
- Physics-Informed Neural Network Coupled with an Extended Kalman Filter for Turbulent Channel Flow Reconstruction Across Multiple Reynolds Numbers(Gaetan France Andriamahollsoa, Arimonjy Ramelina, F. Randimbindrainibe, 2026, European Journal of Artificial Intelligence and Machine Learning)
工程与大气场景下稀疏含噪观测的PINN数据同化重建
本组是方程约束路线在复杂几何与真实观测条件下的应用集合:水下航行器、周期山与曲面后台阶、多孔介质复合系统、可压缩进气道、四维大气边界层风场(无人机沿轨迹采样)等,输入均为稀疏、含噪、部分分量的PIV/探针/轨迹测量。它们的共同研究重点是稀疏度与噪声敏感性、未知/待标定参数过多导致重建不可靠、边界条件与观测算子如何进入损失,可作为判定"可恢复范围"与"输出超出可靠范围"的工程级证据。
- Reconstruction of the mean velocity, pressure, and turbulent viscosity fields from background-oriented schlieren measurements using a physics-informed neural network(Yulia K. Rudenko, N. Vinnichenko, A. V. Pushtaev, Y. Plaksina, A. V. Uvarov, 2025, The Physics of Fluids)
- Reconstruction of the turbulent flow field with sparse measurements using physics-informed neural network(N. Chaurasia, Shubhankar Chakraborty, 2024, The Physics of Fluids)
- Physics Informed Neural Network in Turbulent Porous Flow: Pore-scale Flow Reconstruction(Seohee Jang, M. Jadidi, Y. Mahmoudi, 2024, Proceedings of the 9th World Congress on Momentum, Heat and Mass Transfer)
- Flow field reconstruction of underwater vehicle based on physics-informed neural network(Wei Guo, Jun Ding, Cheng Cheng, Tianqi Wu, Xuesen Chu, 2025, No journal)
- Physics Informed Reconstruction of Four-Dimensional Atmospheric Wind Fields Using Multi-UAS Swarm Observations in a Synthetic Turbulent Environment(A. Tasim, Wei Sun, 2026, arXiv.org)
- Hypersonic Inlet Flow Field Reconstruction Dominated by Shock Wave and Boundary Layer Based on Small Sample Physics-informed Neural Networks(Ming-Ming Guo, Xue Deng, Yue Ma, Ye Tian, Jia-Ling Le, Hua Zhang, 2024, Aerospace Science and Technology)
将对流位移显式编码为跨位置时间滞后边的时空图模型与平流超采样
本组直接回答第4章的核心问题:把"结构以对流速度平移"写入连接与时间对齐。一是基于冻结湍流/平流方程的三维时空超采样,用空间信息换时间分辨率并降低采样率需求;二是把对流位移编码为带时间滞后的显式跨位置边参与联合消息传递,并给出相对静态图、可学习邻接与"空间图+独立时间模块"的定量比较与失效条件(仅当对流位移落在观测网络内才有收益,且相当部分收益来自把运动矢量作为输入特征)。二者共同刻画沿对流方向跨位置时间连接所表达的物理关系及其在剪切变形与失相关下的局限。
- An advection-based model to increase the temporal resolution of PIV time series(F. Scarano, P. Moore, 2011, Experiments in Fluids)
- When Does Advection-Aware Graph Nowcasting Help? A Controlled Study of Distributed Solar Ramp Forecasting with a Self-Supervised Cloud-Motion Estimator(Phillip Jiang, 2026, arXiv.org)
显式空间边、注意力与图Transformer的稀疏传感时空重建(非对流感知近邻方法)
本组是前一组的主要对照与近邻:非定常桥梁绕流的因果自注意力时空预测、把激波检测等局部流场信息嵌入图表示与多跳消息传递的FFMI-GNN、以图结构编码空间连接并用图Transformer做稀疏传感重建(系统分析测点数目与位置影响)、以及激光雷达观测驱动的Transformer时空估计。它们大多将时间连接独立建模或隐式交给注意力,可用来辨析固定位置时间连接、可学习关联与物理对流边在表达结构移动、剪切变形与失相关时的实质差异。
- Spatiotemporal reconstruction of unsteady bridge flow field via hierarchical graph neural networks with causal attention(C.Z. Cai, Jun Xiao, Yun-Feng Zou, Xu-Hui He, 2025, The Physics of Fluids)
- Physics-constrained and flow-field-message-informed graph neural network for solving unsteady compressible flows(Si-Ye Li, Zhen-Sheng Sun, Yujie Zhu, Chi Zhang, 2024, The Physics of Fluids)
- High-fidelity flow-field reconstruction from sparse sensors using a hybrid graph neural network-transformer architecture(Yong-Qi Zhang, Jiang-Zhou Peng, Zhi-Qiao Wang, Wen-Xuan Yuan, Yu-Bai Li, Zhi-Fu Zhou, Hao-Ran Xie, Wei-Tao Wu, 2026, The Physics of Fluids)
- LiDAR-informed transformer framework for spatiotemporal estimation of rotor-effective wind speed beyond Taylor’s frozen turbulence hypothesis(Meng Zhang, J. Zuluaga, M. Yousif, Jung-Hyun Choi, HeeChang Lim, 2026, Energy)
- Flow field reconstruction and prediction of the two-dimensional cylinder flow using data-driven physics-informed neural network combined with long short-term memory(Yehao Dou, Xun Han, Pengzhi Lin, 2025, Engineering applications of artificial intelligence)
时空耦合深度网络、条件生成模型与多方法比较:非时间分辨场+时序信号的时序重建
本组是"用时序信号驱动空间场恢复"的网络化实现与横向评估:CAE+Transformer时空耦合网络以非时间分辨速度场与非定常壁面压力序列重建圆柱绕流时间分辨速度场(含变频采样增益)、条件流匹配生成模型从壁面稀疏/低分辨率观测合成瞬时脉动场并显式量化预测不确定度(直接触及"统计相似但瞬时错误"的识别),以及五种不同方法的比较研究。它们提供时空对齐、模态压缩与序列学习能否替代显式物理联系的经验证据,也给出方法间的取舍依据。
- Spatiotemporal coupling deep neural network for time-resolved flow field reconstruction around a circular cylinder(Xiao-Wei Jin, Meng-Jie Xu, Ye-Fei Yang, Wen-Li Chen, 2024, The Physics of Fluids)
- Conditional flow matching for generative modelling of near-wall turbulence with quantified uncertainty(Xin-Yang Liu, Xian-Tao Fan, Jian-Xun Wang, 2026, Journal of Fluid Mechanics)
- A comparative study of five different PIV interrogation algorithms(M. Piirto, H. Eloranta, P. Saarenrinne, R. Karvinen, 2005, Experiments in Fluids)
相干结构演化、象限分析及动量/耗散输运特征的定量评价
本组提供速度误差与能谱之外的物理特征评价量:雷诺剪切应力(含二次剪应力)的象限分析揭示喷射/扫掠事件对动量输运的贡献及其随位置变化、高时空分辨率壁面剪切应力测量中的多尺度结构能量贡献划分、小波+FTLE+POD追踪尾迹多尺度涡结构的生成—增长—破碎、湍流耗散率的准确估计、涨落驱动的粒子/通量估计。它们构成评价相干结构位置、尺度、相位、寿命与演化过程的方法学基础,并为"场误差较小但结构/输运特征恢复不足"提供判据来源,同时界定哪些物理量可由平面二维速度测量可靠支持。
- Quadrant Analysis of Reynolds Shear Stresses in a Turbulent Square-Duct Flow.(G. Kawahara, K. Ayukawa, J. Ochi, E. Kamada, Fumihiro Ono, 1999, No journal)
- Quadrant Analysis of the Reynolds Shear Stress in a Two-Height Canopy(L. Shig, V. Babin, Ron Shnapp, Eyal Fattal, Alexander Liberzon, Y. Bohbot-Raviv, 2023, Flow Turbulence and Combustion)
- High spatiotemporal resolution wall shear stress measurement and multi-scale energy characterization in turbulent boundary layers(Jizhong Chen, Haiping Tian, Guozhen Ma, Jun-Fei Ding, Shengkun Chen, 2025, The Physics of Fluids)
- Dynamics of multi-scale vortical structures behind a barchan dune(Xiaolei Han, Jia-Wei Li, Hiroka Rinoshika, Yuyang Zhou, Yan Zheng, Lin Dong, A. Rinoshika, 2023, The Physics of Fluids)
- Accurate estimate of turbulent dissipation rate using PIV data(Duo Xu, Jun Chen, 2013, Experimental Thermal and Fluid Science)
- Estimation of Particle Flux Driven by Coherent Mode Using of Statistical Conditional Averaging(Y. Kawachi, S. Inagaki, K. Hasamada, K. Yamasaki, F. Kin, Y. Kosuga, M. Sasaki, Y. Nagashima, N. Kasuya, T. Yamada, Bo-Yu Zhang, A. Fujisawa, 2019, Plasma and Fusion Research)
PIV空间分辨率、空间滤波与欠分辨/随机误差对谱与统计量的衰减及参考场可信度
本组从测量传递函数层面界定参考真值与可恢复尺度:空间分辨率与滤波各向异性的理论、二维/三维谱与统计量在空间滤波下的衰减、随机误差最小化、TR-PIV在回转体湍流边界层中的谱响应与有效光片厚度随壁法向的截止波长、欠分辨条件下湍流强度的正误评估、双平面PIV测量评估以及壁面摩擦与近壁量的相关关系。它们直接支撑"PIV空间滤波与查问窗口如何污染DNS/实验参考场与训练标签",也是区分测量信息不足与模型能力不足的物理量来源。
- Theory of non-isotropic spatial resolution in PIV(F. Scarano, 2003, Experiments in Fluids)
- Appropriate spatial filtering of 2D and 3D PIV measurements of wall-bounded turbulent flows(C. Atkinson, N. Buchmann, O. Amili, J. Soria, 2013, No journal)
- On PIV random error minimization with optimal POD-based low-order reconstruction(M. Raiola, S. Discetti, A. Ianiro, 2015, Experiments in Fluids)
- The spectral response of time-resolved PIV in a turbulent boundary layer(P. Manovski, W. Rowin, Henry Ng, P. Gulotta, Matteo Giacobello, C. D. de Silva, N. Hutchins, I. Marusic, 2025, Experiments in Fluids)
- Validating under-resolved turbulence intensities for PIV experiments in canonical wall-bounded turbulence(J. Lee, Kevin, J. Monty, N. Hutchins, 2016, Experiments in Fluids)
- Assessment of dual plane PIV measurements in wall turbulence using DNS data(N. Saikrishnan, I. Marusic, E. Longmire, 2006, Experiments in Fluids)
- Correlation of the Wall Skin-Friction and Streamwise Velocity Fluctuations in a Turbulent Boundary Layer(I) -Analysis of Long-Time Averaged Space-Time Correlation-(Jun-Mo Yang, Jeong-yeol Yu, Hae-Cheon Choe, 1997, Transactions of The Korean Society of Mechanical Engineers B)
自适应查问窗口、迭代多网格、稀疏矢量修复、近壁预处理与平面压力反演的窗口/采样准则
本组处理观测链路与派生量的工程实现及其可靠性阈值:迭代多网格与窗口变形、自适应细长/拉伸查问窗口以提高近壁高梯度与涡/双曲区信噪比、基于有限时间Lyapunov指数自适应调整窗口形状、鲁棒多分辨率稀疏PIV处理、用线性随机估计替代插值修复成簇伪矢量、近壁与运动界面图像预处理、软边界条件下边界层测量的固有局限,以及由PIV速度求瞬时平面压力场时对查问窗口(小于结构尺度数倍)与采样频率(高于流动特征频率数倍)的定量准则。它们共同界定"观测能否支撑某尺度恢复"的硬指标,避免把测量伪迹误判为模型能力不足。
- Iterative multigrid approach in PIV image processing with discrete window offset(F. Scarano, M. Riethmuller, 1999, Experiments in Fluids)
- Improve the Spatial Resolution of PIV Results by Using Hierarchical Recursive Operation(H. Hui, T. Saga, Toshio Kobayashi, N. Taniguchi, Shigeki Segawa, 2000, No journal)
- Adjustable interogation window for 2D PIV estimation based on local Lagrangian coherency(Ali Rahimi Khojasteh, D. Heitz, Yin Yang, 2021, No journal)
- Adjustable window for 2D PIV estimation based on local Lagrangian coherency(Ali Rahimi Khojasteh, D. Heitz, Yin Yang, 2021, 14th International Symposium on Particle Image Velocimetry)
- Robust and multiresolution sparse processing particle image velocimetry for improvement in spatial resolution(Chihaya Abe, Naoki Kanda, Kumi Nakai, Taku Nonomura, 2024, Journal of Vision)
- Spurious PIV Vector Correction Using Linear Stochastic Estimation(Daniel Butcher, A. Spencer, 2019, Fluids)
- Image pre-processing method for near-wall PIV measurements over moving curved interfaces(L. Jia, Yi-Ding Zhu, Yong-Xia Jia, Hui-Jing Yuan, Cun-Biao Lee, 2017, Measurement Science and Technology)
- Boundary-layer limitations of a standard soft-boundary physics-informed neural network: A Legendre wavelet collocation benchmark for a singularly perturbed transport problem(S. Nayak, Arun Kumar Gupta, 2026, Advances in Computational Science and Engineering)
- Instantaneous planar pressure determination from PIV in turbulent flow(R. de Kat, B. V. van Oudheusden, 2012, Experiments in Fluids)
合并后的 89 篇文献按报告七个方向重排为 15 个并列分组,前四组构成\"时间—空间对应的物理依据\"链条:冻结假设的检验与随机扫掠修正 → 尺度/壁法向相关的对流速度与失相关标度 → TRPIV与热线实测的时空相关、内—外区相干及应力交叉谱 → 壁压/标量/可压缩/自然对流中的时空特性与亚对流成分。第五、六组为单点/少测点重建能力与限制的方法谱系:线性与谱随机估计、条件高斯与识别策略(明确其输出为条件平均/统计相似场),以及相位平均与POD—压缩感知降阶稀疏重构。第七至九组区分三类物理进入学习模型的方式:近壁渐近标度与对数律作为硬约束/分区先验(具体机理引导)、以方程残差与PDE先验约束的边界层重建与不确定度量化(一般约束基线)、以及工程与大气场景的稀疏数据同化应用(可观察稀疏度、噪声与参数冗余导致的不可靠性)。第十、十一、十二组对应时空连接与近邻模型:对流位移显式编码为带时间滞后的跨位置边与平流超采样(含失效条件)、GNN/图Transformer/注意力的稀疏传感重建、以及CAE+Transformer与条件流匹配生成模型及多方法比较(用于识别统计相似但瞬时错误)。第十三组汇总相干结构、象限分析与动量/耗散输运特征的定量评价;第十四、十五组从测量侧界定可恢复范围与参考场可信度(空间分辨率与滤波传递函数、欠分辨与随机误差;自适应查问窗口、稀疏矢量修复、近壁预处理与压力反演准则)。整体上方向1—3得到最密集的文献支撑;方向5与方向6(可观测性/可辨识性与误差判据)多以分散指标与误差分析形式散落于第十三至十五组,仍属需重点指出的研究空白;跨几何与广泛跨工况泛化仅作为次要线索见于第九组。
本组文献共同处理"时间信号能否等价为空间结构匀速对流"这一第1章最基础的物理前提:既有在管道流、光学湍流、地面层湍流与跨声速/高超场景下对TFTH的实测检验与误差来源分解,也有以随机扫掠(sweeping)假设、双特征时间等替代模型对冻结假设的扩展修正,还包括用当地瞬时速度替代平均对流速度的非侵入时空超采样测量流程。它们给出冻结假设成立的范围、失效率与可替代假设,为后面各组讨论尺度相关对流速度、恢复空间范围与时间窗口提供前提条件。相关文献:Shu-Chao Dai et. al, 2025 等 9 篇文献
本组聚焦冻结假设中的关键参数本身:涡对流速度的定义与选取方式、其随尺度和壁法向位置的变化、极高雷诺数(含VLSM)下的内/外尺度标度,以及由两点时空相关函数与各向异性空间相关刻画的结构输运速率与失相关长度。它们回答"在多长的时间滞后、多大的空间邻域内,单点信号仍与周围流场保持确定性联系",是界定可恢复空间范围、尺度与时间跨度的定量依据,而非仅仅检验假设的好坏。相关文献:V. Nikora et. al, 2000 等 6 篇文献
本组以实验手段直接测量壁湍流时空相关与时空结构:时间分辨PIV对近壁时空相关场的直接测量、不同壁法向位置的时空相关与相干函数、内—外区大尺度信号相干幅值与相位的实验刻画、湍流时空结构的实验观测(含一篇DFD会议摘要性工作),以及由二阶/四阶应力—速度交叉谱推导的对流速度差与相干相似(CS)模型(随壁法向距离由单瓣向三瓣转变的时空耦合结构)。它们把前述标度关系落实为壁湍流(尤其近壁区)可用的定量时空耦合图像,是机理引导重建最直接的数据支撑。相关文献:A. Favre et. al, 1967 等 7 篇文献
本组把时空耦合规律推广到速度场之外的观测量与更宽工况:壁压脉动的波数—频率交叉谱及其对流/亚对流成分来源识别、标量Schmidt数对对流速度与时空结构的影响、Mach数对时空相关的作用、自然对流边界层的时空结构。它们既用于判断"单点压力/标量/非壁面传感器信号能约束什么",也给出不同介质与可压缩性下对流速度与失相关规律的可迁移性边界,避免把不可压中性壁湍流结论简单外推。相关文献:Humza Butt et. al, 2026 等 6 篇文献
本组是"少量点观测→周围场"的经典统计估计主线:线性随机估计(LSE)及其谱形式(尺度相关增益与相位核)、内—外区大尺度信号由对数区单点预测、Navier–Stokes线性模型(随机强迫+涡黏耗散)跨壁法向位置的估计及其模型项随壁距与尺度必须变化的证据、条件高斯框架给出的条件统计解析解与数据同化思路、缸内流实时全场估计、采样条件下识别方法与LSE的组合策略,以及涡量场到速度场的卷积反演。它们共同界定单点信息只能约束低维/大尺度相干分量,输出本质上是条件平均或统计相似流场而非瞬时真实结构,并量化观测位置、信号分量与稀疏度的影响。相关文献:C. Tinney et. al, 2006 等 10 篇文献
本组从"相干结构"与"降阶模态"两个角度把稀疏观测压缩到低维表示:相位平均提取周期/相干结构并给出湍流特性(明确区分相干结构与随机脉动),以及POD模态配合压缩感知与传感器布置实现稀疏重构。它们代表不依赖神经网络的确定性降阶路线,其可恢复自由度上限、传感器位置敏感性与条件平均造成的细节丢失,是讨论"测量信息不足还是模型能力不足"的量化参照。相关文献:R. Perrin et. al, 2006 等 2 篇文献
本组代表"具体湍流机理引导"而非泛化方程约束:把对数律、von Kármán常数与截距、对数层上下界、黏性底层划分等近壁标度写成硬编码约束或分段物理先验,直接作用于网络结构、分区加权损失与特征表示;也包括边界层位置未知时以边界识别迭代训练替代固定分区,以及仅由y+>30统计量重建近壁场与动量输运的渐近硬约束方案。其价值在于避免网络在黏性底层施加错误物理、保证剖面及其导数的物理一致,同时暴露出参数标定依赖与适用雷诺数/壁法向区间的失效条件。相关文献:A. Palasis et. al, 2025 等 4 篇文献
本组以N-S/能量方程残差或PDE先验作为主要物理约束,重点在湍流边界层类流动:稀疏数据下的时空重建与压力推断、RANS残差惩罚配合内部测点/界面数据分布对二阶统计量的影响、PINN与LSTM在时空外推上的物理保真度对比、贝叶斯神经网络给出带不确定度的物理一致预测、域分解分布式PINN的压力规范固定、速度到温度场的方程约束反演,以及PINN与扩展卡尔曼滤波耦合的平均量重建。通过与上一组对照,可分离"一般方程残差约束"与"具体机理引导"各自的独立收益与失效条件。相关文献:Seohee Jang et. al, 2024 等 7 篇文献
本组是方程约束路线在复杂几何与真实观测条件下的应用集合:水下航行器、周期山与曲面后台阶、多孔介质复合系统、可压缩进气道、四维大气边界层风场(无人机沿轨迹采样)等,输入均为稀疏、含噪、部分分量的PIV/探针/轨迹测量。它们的共同研究重点是稀疏度与噪声敏感性、未知/待标定参数过多导致重建不可靠、边界条件与观测算子如何进入损失,可作为判定"可恢复范围"与"输出超出可靠范围"的工程级证据。相关文献:Yulia K. Rudenko et. al, 2025 等 6 篇文献
本组直接回答第4章的核心问题:把"结构以对流速度平移"写入连接与时间对齐。一是基于冻结湍流/平流方程的三维时空超采样,用空间信息换时间分辨率并降低采样率需求;二是把对流位移编码为带时间滞后的显式跨位置边参与联合消息传递,并给出相对静态图、可学习邻接与"空间图+独立时间模块"的定量比较与失效条件(仅当对流位移落在观测网络内才有收益,且相当部分收益来自把运动矢量作为输入特征)。二者共同刻画沿对流方向跨位置时间连接所表达的物理关系及其在剪切变形与失相关下的局限。相关文献:F. Scarano et. al, 2011 等 2 篇文献
本组是前一组的主要对照与近邻:非定常桥梁绕流的因果自注意力时空预测、把激波检测等局部流场信息嵌入图表示与多跳消息传递的FFMI-GNN、以图结构编码空间连接并用图Transformer做稀疏传感重建(系统分析测点数目与位置影响)、以及激光雷达观测驱动的Transformer时空估计。它们大多将时间连接独立建模或隐式交给注意力,可用来辨析固定位置时间连接、可学习关联与物理对流边在表达结构移动、剪切变形与失相关时的实质差异。相关文献:C.Z. Cai et. al, 2025 等 5 篇文献
本组是"用时序信号驱动空间场恢复"的网络化实现与横向评估:CAE+Transformer时空耦合网络以非时间分辨速度场与非定常壁面压力序列重建圆柱绕流时间分辨速度场(含变频采样增益)、条件流匹配生成模型从壁面稀疏/低分辨率观测合成瞬时脉动场并显式量化预测不确定度(直接触及"统计相似但瞬时错误"的识别),以及五种不同方法的比较研究。它们提供时空对齐、模态压缩与序列学习能否替代显式物理联系的经验证据,也给出方法间的取舍依据。相关文献:Xiao-Wei Jin et. al, 2024 等 3 篇文献
本组提供速度误差与能谱之外的物理特征评价量:雷诺剪切应力(含二次剪应力)的象限分析揭示喷射/扫掠事件对动量输运的贡献及其随位置变化、高时空分辨率壁面剪切应力测量中的多尺度结构能量贡献划分、小波+FTLE+POD追踪尾迹多尺度涡结构的生成—增长—破碎、湍流耗散率的准确估计、涨落驱动的粒子/通量估计。它们构成评价相干结构位置、尺度、相位、寿命与演化过程的方法学基础,并为"场误差较小但结构/输运特征恢复不足"提供判据来源,同时界定哪些物理量可由平面二维速度测量可靠支持。相关文献:G. Kawahara et. al, 1999 等 6 篇文献
本组从测量传递函数层面界定参考真值与可恢复尺度:空间分辨率与滤波各向异性的理论、二维/三维谱与统计量在空间滤波下的衰减、随机误差最小化、TR-PIV在回转体湍流边界层中的谱响应与有效光片厚度随壁法向的截止波长、欠分辨条件下湍流强度的正误评估、双平面PIV测量评估以及壁面摩擦与近壁量的相关关系。它们直接支撑"PIV空间滤波与查问窗口如何污染DNS/实验参考场与训练标签",也是区分测量信息不足与模型能力不足的物理量来源。相关文献:F. Scarano et. al, 2003 等 7 篇文献
本组处理观测链路与派生量的工程实现及其可靠性阈值:迭代多网格与窗口变形、自适应细长/拉伸查问窗口以提高近壁高梯度与涡/双曲区信噪比、基于有限时间Lyapunov指数自适应调整窗口形状、鲁棒多分辨率稀疏PIV处理、用线性随机估计替代插值修复成簇伪矢量、近壁与运动界面图像预处理、软边界条件下边界层测量的固有局限,以及由PIV速度求瞬时平面压力场时对查问窗口(小于结构尺度数倍)与采样频率(高于流动特征频率数倍)的定量准则。它们共同界定"观测能否支撑某尺度恢复"的硬指标,避免把测量伪迹误判为模型能力不足。相关文献:F. Scarano et. al, 1999 等 9 篇文献
总计 89 篇相关文献
The efficacy of reconstructing three-dimensional (3D) volumes of time-evolving three-component velocity from planar experimental measurements is explored within strong shear-distorting turbulent flows using Taylor’s frozen turbulence hypothesis. In flows with a strong mean shear-rate the instantaneous turbulence structure is spuriously distorted with the classical Taylor’s hypothesis method, where the local ensemble average velocity is used as the convective velocity to reconstruct 3D volumes. In the current study, two additional models for the convective velocity that extend the classical Taylor’s hypothesis approach are explored with varied levels of mean shear-rate and turbulence intensity in order to reconstruct 4D velocity fields for time-resolved analysis of turbulence structure. The classical Taylor’s frozen turbulence hypothesis is compared with local time average and instantaneous convective velocities, with and without a Poisson solver step to maintain continuity. Direct numerical simulation data of a turbulent channel flow from the Johns Hopkins Turbulence Database are used to assess the accuracy of the three methods while varying mean shear-rate and turbulence intensity independently. The three methods are also applied to time-resolved sPIV measurements on transverse planes within the near-wall surface layer of a canonical flat-plate turbulent boundary layer to assess the statistical means of the reconstructions. The instantaneous convective velocity method is generally found to be most accurate at reconstructing instantaneous velocity fields, although systematic biases are observed in mean statistics. In flows with lower turbulence intensities, the local time average convective velocity is comparable for significantly lower computation and implementation costs.
Abstract This work presents a novel numerical procedure for reconstructing volumetric density and velocity fields from planar laser-induced fluorescence (PLIF) and stereoscopic particle image velocimetry (SPIV) data. This new method is theoretically and practically demonstrated to provide more accurate 3D vortical structures and density fields in high shear flows than reconstruction methods based on the mean convective velocity. While Taylor’s hypothesis of frozen turbulence is commonly applied by using the local mean streamwise velocity, the proposed algorithm uses the measured local instantaneous velocity for data convection. It consists of a step-by-step reconstruction based on a mixed Lagrangian–Eulerian solver that includes the 3D interpolation of scattered flow data and that relaxes the Taylor’s hypothesis by iterative enforcement of the incompressibility constraint on the velocity field. This methodology provides 3D fields with temporal resolution, spatial resolution, and accuracy comparable to that of real 3D snapshots, thus providing a practical alternative to tomographic measurements. The procedure is validated using numerical data of the constant-density channel flow available on the Johns Hopkins University Turbulence Database (JHTDB), showing the accurate reconstruction of the 3D velocity field. The algorithm is applied to an experimental dataset of PLIF and SPIV measurements of a variable-density jet flow, demonstrating its capability to provide 3D velocity and density fields that are more consistent with the Navier–Stokes equations compared to the mean flow convective method. Graphic abstract
No abstract available
No abstract available
No abstract available
A numerical implementation of the advection equation is proposed to increase the temporal resolution of PIV time series. The method is based on the principle that velocity fluctuations are transported passively, similar to Taylor’s hypothesis of frozen turbulence. In the present work, the advection model is extended to unsteady three-dimensional flows. The main objective of the method is that of lowering the requirement on the PIV repetition rate from the Eulerian frequency toward the Lagrangian one. The local trajectory of the fluid parcel is obtained by forward projection of the instantaneous velocity at the preceding time instant and backward projection from the subsequent time step. The trajectories are approximated by the instantaneous streamlines, which yields accurate results when the amplitude of velocity fluctuations is small with respect to the convective motion. The verification is performed with two experiments conducted at temporal resolutions significantly higher than that dictated by Nyquist criterion. The flow past the trailing edge of a NACA0012 airfoil closely approximates frozen turbulence, where the largest ratio between the Lagrangian and Eulerian temporal scales is expected. An order of magnitude reduction of the needed acquisition frequency is demonstrated by the velocity spectra of super-sampled series. The application to three-dimensional data is made with time-resolved tomographic PIV measurements of a transitional jet. Here, the 3D advection equation is implemented to estimate the fluid trajectories. The reduction in the minimum sampling rate by the use of super-sampling in this case is less, due to the fact that vortices occurring in the jet shear layer are not well approximated by sole advection at large time separation. Both cases reveal that the current requirements for time-resolved PIV experiments can be revised when information is poured from space to time. An additional favorable effect is observed by the analysis in the frequency domain whereby the spectrum becomes significantly less prone to aliasing error for the super-sampled data series.
Currently, limitations in modeling the temporal behavior of light propagating through atmospheric turbulence stem from the Taylor's frozen turbulence hypothesis (TFTH). Indeed, under certain conditions it has been reported to be unreliable, often leading to inaccurate predictions. On the other hand, in fluid dynamics an alternative has been validated: the random sweeping hypothesis. Nevertheless, its applicability to optical turbulence has remained unexplored. This work introduces the first, to the best of our knowledge, controlled experiment testing this hypothesis on the spatiotemporal properties from image wander. The existence of two characteristic times is observed, one associated with TFTH decorrelation and a second potentially linked to the sweeping hypothesis.
No abstract available
Taylor’s frozen turbulence hypothesis (FTH) is investigated in the roughness sublayer of a sloped vineyard canopy using a spatial array of fine‐wire thermocouples and ultrasonic anemometers. The Ellipse Approximation (EA) method is applied to the measured space‐time temperature correlation function to delineate sweeping effects from advection velocity. Sweeping effects are explained primarily by the turbulence kinetic energy. Upon the removal of sweeping effects, the advection velocity is found to be commensurate with the mean velocity.
Taylor’s Frozen Turbulence Hypothesis (TFTH) has been used extensively in theoretical studies to model the temporal fluctuations of optical quantities affected by atmospheric turbulence. It has been relied upon to provide temporal-frequency spectra under varying propagation conditions and for different atmospheric refractive index models. However, experimental works have revealed its limitations, such as systematic inaccuracies in estimating cross winds during calm nights in scintillation measurements at astronomical sites, scintillation discrepancies in ground-layer measurements, and broad estimates of the coherence time in phase fluctuation measurement techniques. This highlights the need to recognize the limitations of the TFTH and seek alternatives that can provide a more reliable description of atmospheric turbulence’s temporal fluctuations. Here, we propose a spatio-temporal statistics for refractive index fluctuations through fluid dynamics models and evaluate the complex phase propagation under weak turbulence. Then, we test its ability to reproduce experimental observations under different ground-layer turbulence conditions.
No abstract available
Accurate measurement of pipeline flow is of great significance for industrial and environmental monitoring. Traditional intrusive methods have the disadvantages of high cost and damage to pipeline structure, while non-intrusive techniques can circumvent such issues. Although Taylor’s frozen hypothesis has a theoretical advantage in non-intrusive velocity detection, current research focuses on planar flow fields, and its applicability in turbulent circular pipes remains controversial. Moreover, there is no precedent for combining it with distributed acoustic sensing (DAS) technology. This paper constructs a circular pipe turbulence model through large eddy simulation (LES), revealing the spatiotemporal distribution characteristics of turbulent kinetic energy and the energy propagation rules of FK spectra. It proposes a dispersion feature enhancement algorithm based on cross-correlation, which combines a rotatable elliptical template with normalized cross-correlation coefficients to suppress interference from non-target directions. An experimental circulating pipeline DAS measurement system was set up to complete signal denoising and compare two principles of flow velocity verification. The results show that the vortex structure of turbulent flow in circular pipes remains stable in the convection direction, conforming to theoretical premises; the relative error of average flow velocity by this method is ≤3%, with significant improvements in accuracy and stability in high-flow zones. This study provides innovative methods and experimental basis for non-intrusive flow detection using DAS.
No abstract available
The article is devoted a new method of flow measurement using an original mathematical algorithm and piezo film sensor. We describe the first part of an investigative project to develop a new flow measurement system. The advantage of the presented method is that it supports noninvasive flow measurement through the pipe wall. We present different results of studying mathematical methods for signal processing and numerical simulations using LES turbulence models such as Stress-Blended Eddy Simulation (SBES). The physical effect based on Taylor's frozen turbulence hypothesis, which serves as the foundation of this measurement method, is described. The mathematical algorithm and signal processing method contain a two-dimensional Fourier transform to build the frequency-wavenumber spectrum and cross-correlation function. We provide a short description about the chosen sensor and justify its use for the proposed measured system. In conclusion, the findings of our investigation and a real prototype scheme are presented.
With the development of offshore wind turbine single power toward levels beyond 10 MW, the increase in heat loss of components in the nacelle leads to a high local temperature in the nacelle, which seriously affects the performance of the components. Accurate reconstruction and control of thermal turbulence in the nacelle can alleviate this problem. However, the physical environment of thermal turbulence in the nacelle is very complex. Due to the intermittent and fluctuating nature of turbulence, the turbulent thermal environment is highly nonlinear when coupled with the temperature field. This leads to large reconstruction errors in existing reconstruction methods. Therefore, we improve the sparse reconstruction method for compressed sensing (CS) based on the concept of virtual time using proper orthogonal decomposition (POD). The POD-CS method links the turbulent thermal environment reconstruction with matrix decomposition to ensure computational accuracy and computational efficiency. The improved particle swarm optimization (PSO) is used to optimize the sensor arrangement to ensure stability of the reconstruction and to save sensor resources. We apply this novel and improved PSO-POD-CS coupled reconstruction method to the thermal turbulence reconstruction in the nacelle. The effects of different basis vector dimensions and different sensor location arrangements (boundary and interior) on the reconstruction errors are also evaluated separately, and finally, the desired reconstruction accuracy is obtained. The method is of research value for the reconstruction of conjugate heat transfer problems with high turbulence intensity.
This study introduces a physics-informed neural network (PINN) framework for the spatio-temporal reconstruction and prediction of turbulent boundary layer flow fields over a broad range of Reynolds numbers. By incorporating the governing Navier-Stokes (N-S) equations into the loss function, the proposed model effectively integrates physical constraints with sparse flow field data, enabling high-fidelity spatial reconstruction, resolution optimization, and temporal prediction of velocity and pressure fields. The method is validated using simulated shock-induced turbulent boundary layer data and experimental turbulent channel flow data. Results show that the PINN model accurately reconstructs flow fields with 60% missing data, infers pressure fields without pressure measurements, and provides reliable short-term temporal predictions. The framework exhibits strong generalizability and robustness, highlighting its potential for complex turbulent flow analysis in both computational and experimental fluid dynamics.
Spatio-temporal prediction of turbulent flow fields remains a challenging task due to complex multi-scale dynamics and high-dimensional data. Recent advances in deep learning offer promising data-driven approaches, yet their comparative performance under physical constraints warrants further investigation. In this study, a physics-informed neural network (PINN) framework and a long short term memory (LSTM) network are implemented and systematically compared for the spatio-temporal prediction of turbulent boundary layer flows over a wide range of Reynolds numbers. Both models are trained and tested using simulated data from a shock-induced turbulent boundary layer and experimental data from particle image velocimetry (PIV) measurements of turbulent channel flow. Prediction accuracy is evaluated across multiple time horizons, with particular attention to short-term and extended temporal extrapolation. Results indicate that the PINN framework achieves higher physical fidelity and robustness in reconstructing velocity and pressure fields from sparse or incomplete data, especially in scenarios requiring alignment with fundamental flow physics. The comparative analysis highlights the complementary strengths and limitations of physics-embedded and purely data-driven deep learning approaches for turbulent flow prediction, providing guidance for method selection based on prediction horizon, data availability, and physical accuracy requirements.
A data assimilation technique for turbulent flows, based on the use of the physics-informed neural network (PINN), is presented. Mean temperature fields, measured with the nonintrusive background-oriented schlieren (BOS) method, and sparse velocity measurements are used as input data. The velocity, pressure, turbulent viscosity, and thermal conductivity fields are reconstructed using PINN. PINN searches for the flow field, which approximately satisfies the governing equations while remaining close to the available experimental data. Reconstruction is performed for the free and impinging subsonic axisymmetric turbulent jets of hot air. The results are compared with conventional Reynolds-averaged Navier–Stokes simulations. The feasibility of taking into account the spatial variation of the turbulent Prandtl number is demonstrated. However, due to a large number of unknown parameters, accurate reconstruction of the turbulent Prandtl number field requires velocity data measured with high spatial resolution.
Near-wall flow reconstruction from outer-region observations is challenging because the unobserved region contains the transition from viscous- to Reynolds-stress-dominated momentum transport and the buffer-layer maximum in turbulent kinetic energy production. We develop a physics-informed neural network with adaptive asymptotic hard constraints (AAHC-PINN) for a restricted statistical inverse problem in fully developed turbulent channel flow. Only statistics at y+> 30 are observed, while near-wall fields and pressure are withheld and no empirical turbulence closure is prescribed. The AAHC representation imposes the wall-limit values and leading asymptotic orders of the mean velocity and Reynolds stresses, while their amplitudes and wall-normal distributions are inferred from outer-region observations, the unclosed Reynolds-averaged Navier–Stokes equations, and additional physical constraints. At Reτ= 1000, the mean streamwise velocity and Reynolds-stress components are reconstructed within y+≤ 30 with errors below 5%. The reconstructed fields capture the transition from viscous-dominated to Reynolds-stress-dominated momentum transport and the transfer of kinetic energy from the mean flow to turbulent fluctuations through buffer-layer mean-shear production. The characteristic locations of the momentum-transport transition and production maximum differ from direct numerical simulation by 4.6% and 3.8%, respectively, while the production maximum is underpredicted by 7.5%. Cases at Reτ= 550, 950, and 2000 use the same formulation and settings, showing robustness across the tested range. The results show that the reconstruction recovers both the missing near-wall statistics and their coupled momentum-transport and energy-transfer organization. Under the conditions examined, the characteristic locations of the momentum-transport transition and production maximum are recovered more accurately than the local production intensity.
Accurate high-resolution flow field prediction based on limited experimental data is a complex task. This research introduces an innovative framework leveraging physics-informed neural network (PINN) to reconstruct high-resolution flow fields using sparse particle image velocimetry measurements for flow over a periodic hill and high-fidelity computational fluid dynamics data for flow over a curved backward-facing step. Model training utilized mean flow measurements, with increased measurement sparsity achieved through various curation strategies. The resulting flow field reconstruction demonstrated marginal error in both test cases, showcasing the ability of the framework to reconstruct the flow field with limited measurement data accurately. Additionally, the study successfully predicted flow fields under two different noise levels, closely aligning with experimental and high-fidelity results (experimental, direct numerical simulation, or large eddy simulation) for both cases. Hyperparameter tuning conducted on the periodic hill case has been applied to the curved backward-facing step case. This research underscores the potential of PINN as an emerging method for turbulent flow field prediction via data assimilation, offering reduced computational costs even with sparse, noisy measurements.
This study explores the implementation of physics-informed neural networks (PINNs) to analyze turbulent flow in composite porous-fluid systems. These systems are composed of a fluid-saturated porous medium and an adjacent fluid, where the flow properties are exchanged across the porous-fluid interface. The segregated PINN model employs a novel approach combining supervised learning and enforces fidelity to flow physics through penalization by the Reynolds-averaged Navier-Stokes (RANS) equations. Two cases were simulated for this purpose: solid block, i.e. porous media with zero porosity, and porous block with a defined porosity. The effect of providing internal training data on the accuracy of the PINN predictions for prominent flow features, including flow leakage, channeling effect and wake recirculation was investigated. Additionally, L2 norm error, which evaluates the prediction accuracy for flow variables was studied. Furthermore, PINN training time in both cases with internal training data was considered in this study. Results showed that the PINN model predictions with second-order internal training data achieved high accuracy for the prominent flow features compared to the RANS data, within a 20% L2 norm error of second-order statistics in the solid block case. In addition, for the porous block case, providing training data at the porous-fluid interface showed errors of 18.04% and 19.94% for second-order statistics, representing an increase in prediction accuracy by 7% compared to without interface training data. The study elucidates the impact of the internal training data distribution on the PINN training in complex turbulent flow dynamics, underscoring the necessity of turbulent second-order statistics variables in PINN training and an additional velocity gradient treatment to enhance PINN prediction.
- Turbulence modelling in porous media presents challenges in Computational Fluid Dynamics (CFD). While various modelling approaches have been employed to analyze turbulent flow properties, achieving both precision and cost-effectiveness remains a significant hurdle. In recent years, Deep Learning (DL), with its capacity for solving nonlinear model, has emerged as a promising solution to address these challenges. The advent of Physics-Informed Neural Networks (PINN) has expanded the scope of Deep learning applications in turbulent flow modelling. However, applying PINN to complex flow physics within porous media remains an underexplored territory. This study employs PINN to solve the Reynolds-Averaged Navier-Stokes (RANS) equations in a composite porous-fluid system, guided by supervised learning and penalized by the RANS equation to ensure fidelity to flow physics. The research aims to enhance flow prediction accuracy and explore the influence of data distribution on PINN performance in complex flow scenarios in composite porous-fluid systems. Results showed that using porous-fluid interface training data provides better accuracy, with improvements of 40% and 2% in second-order statistics. This research contributes to advancing our understanding of turbulent flows in porous media and highlights the potential of PINN as a valuable tool for exploring complex flow physics.
No abstract available
We present a hybrid data-assimilation framework that couples a Physics- Informed Neural Network (PINN) with an Extended Kalman Filter (EKF) for the reconstruction of mean turbulent statistics in incompressible channel flow. The PINN is parameterised by the friction Reynolds number Reτ and maps the wall-normal coordinate y+ to four mean turbulent quantities (U+ , k+ , ε+ , uv+), using Fourier feature embeddings and physics-based residuals derived from the Reynolds-averaged Navier–Stokes (RANS) equations. The network is trained exclusively on Direct Numerical Simulation (DNS) data at Reτ = 5200 and evaluated without retraining at Reτ ∈ {550, 2000, 5200, 10000}. The EKF performs a sequential spatial correction of the PINN predictions along the wall-normal direction, using adaptive noise-covariance tuning. Across all Reynolds numbers, the PINN+EKF framework reduces the Mean-Squared Error (MSE) by 46.6% to 51.2% relative to the standalone PINN, with a mean improvement of 49.1%. These results indicate that coupling a physics-constrained surrogate model with sequential Bayesian filtering is a promising route toward scalable and accurate turbulence reconstruction from sparse or out-of-distribution data.
No abstract available
One of the most popular recent areas of machine learning predicates the use of neural networks (NNs) augmented by information about the underlying process in the form of partial differential equations (PDEs). These physics-informed NNs (PINNs) are obtained by penalizing the inference with a PDE and have been cast as a minimization problem currently lacking a formal approach to quantify the uncertainty. In this work, we propose a novel model-based framework that regards the PDE as a prior information of a deep Bayesian NN (BNN), physics-informed prior (PIP)-BNN. The prior is calibrated without data to resemble the PDE solution in the prior mean, while our degree of confidence in the PDE with respect to the data is expressed in terms of the prior variance. The information embedded in the PDE are then propagated to the posterior yielding physics-informed forecasts with uncertainty quantification. We apply our approach to a simulated viscous fluid and to an experimentally obtained turbulent boundary layer velocity in a water tunnel using an appropriately simplified Navier–Stokes (NS) equation. Our approach requires very few observations to produce physically consistent forecasts as opposed to nonphysical forecasts stemming from noninformed priors, thereby allowing forecasting complex systems, where some amount of data as well as some contextual knowledge is available.
No abstract available
Wall-bounded turbulence modeling is fundamentally limited by the representation of the logarithmic law, yet its parameters such as the von Kármán constant, κ, and intercept, B, and its spatial bounds (ylow+, yhigh+) are often assumed a priori. Here we present a hybrid optimization – physics-informed neural network (PINN) framework that directly addresses this point. The method's novelty is a two-stage process: first, an optimization pre-processing step uses direct numerical dimulation data to calibrate the log-law constants (κ,B) and (ylow+, yhigh+). Next, these values are hard-coded into a PINN, whose loss function constrains the solution to obey both the governing Reynolds-averaged Navier–Stokes equations and the log-law within its identified region. By accurately delineating the log-layer we prevent the PINN from enforcing incorrect physics in the viscous sublayer, which is essential for obtaining physically consistent velocity profiles and their derivatives. This method suggests a more universal and accurate model across various Reynolds numbers, thereby enabling the development of data-informed, physics-grounded turbulence closures.
This paper constructs a physics-informed neural network with boundary identification strategy (BI-PINN) to reconstruct the flow field of an ultra-high-load, low-pressure turbine cascade using sparse data. Since the boundary layer position is unknown before training, an iterative method is required to determine the boundary layer range. The BI-PINN trains the network based on a dual convergence criterion: achieving flow field reconstruction convergence in both the boundary layer and mainstream regions to extract the current boundary layer range and then ensuring the convergence of the boundary layer range. To improve neural network convergence while maintaining high training speed, this study assigns different weights to the governing equation terms in the loss function for the boundary layer and mainstream regions. The BI-PINN method demonstrates computational accuracy comparable to computational fluid dynamics in inverse problems with incomplete boundary conditions. This study finds that for high-load, low-pressure turbine cascades, selecting pressure observations only on the blade surface is sufficient for high-accuracy flow field reconstruction. However, for ultra-high-load, low-pressure turbine cascades, this paper proposes a method of selecting multiple normal-direction pressure observations along the blade surface to achieve high-accuracy flow field reconstruction with a small observation set. The BI-PINN method balances the conflict between the low-gradient mainstream region and the high-gradient boundary layer region in flow field reconstruction. By leveraging both physical governing equations and observation data, it enables precise full flow field reconstruction of ultra-high-load, low-pressure turbine cascades using a small dataset and a neural network with a limited number of parameters.
When calculating the transient flow around a bridge structure, traditional computational fluid dynamics methods are extremely time-consuming, especially for multiparameter optimization analyses. Inspired by the development of deep graph neural networks with a mesh structure, this paper describes a spatiotemporal prediction framework for the rapid reconstruction and prediction of transient flows on large-scale unstructured grids. To ensure stability and reliability during self-supervised training, a causal self-attention mechanism is employed in the temporal model. The framework is trained and tested on a dataset containing 40 000 snapshots of bridge flow fields with Reynolds numbers ranging from 104 to 105. The relative mean square error of the model in predicting the velocity and pressure fields is found to be in the order of 10−3 and the relative error does not exceed 10%. This demonstrates that the model is capable of reconstructing high-dimensional flow field information from low-dimensional data. Furthermore, the proposed model achieves a computational speedup by two orders of magnitude compared with traditional computational fluid dynamics methods with respect to the temporal inference. To validate its ability to infer bridge aerodynamic characteristics, the model is used to predict the bridge surface pressure, aerodynamic coefficients, streamlines, and vorticity. The results demonstrate that the proposed model has reliable accuracy, representation, and stability in predicting bridge flow fields and identifying multiscale characteristics.
We propose a spatiotemporal coupling deep neural network approach for time-resolved reconstruction of the velocity field around a circular cylinder. The neural network leverages two distinct data types: (1) non-time-resolved velocity field around the cylinder, consisting of fixed frequency sampling and variable frequency sampling velocity field, and (2) the time-resolved surface pressure sequence around the cylinder. The deep neural network comprises two sub-networks: a convolutional autoencoder (CAE) for nonlinear mode extraction and a Transformer for sequence-to-sequence learning. We refer to this architecture as CTNet (CAE-Transformer Network). The encoder in the CAE maps non-time-resolved velocity field to a latent vector, enabling the extraction of nonlinear modal coefficients. An appropriate time window length for the surface pressure sequence is then selected to establish a Transformer sequence learning model, using the chosen sequence as input to predict the corresponding nonlinear modal coefficients. Once the Transformer is well trained, the time-resolved nonlinear modal coefficients of velocity field can be achieved. Along with the well-trained decoder in the CAE, the time-resolved velocity field can be reconstructed from the output of the Transformer. We verify the performance of CTNet by a simulated dataset at a representative Reynolds number of 3900. The results show a relative reconstruction error of just 6.3% for the time-resolved velocity field, demonstrating high reliability in the reconstruction. We further compare the reconstructed velocity field obtained with and without the utilization of variable frequency sampling velocity field. Notably, the inclusion of variable frequency sampling velocity field significantly improves the reconstruction quality.
Reconstructing flow fields from sparse observations constitutes a fundamental challenge in aerodynamic analysis, since existing methods often fail to reconcile accuracy with computational efficiency under severe data constraints. We introduce a Graph Transformer Reconstruction Network (GTREN), a novel model designed to infer latent data patterns from extremely sparse measurements and to reconstruct high-fidelity, full-field flow solutions. The approach represents discrete grid points as nodes within a graph data structure, enabling a principled encoding of spatial relationships. We integrate a Transformer-style attention mechanism into the graph network's message-passing operations, allowing the model to selectively emphasize salient neighbor interactions while attenuating irrelevant signals; this facilitates efficient and stable information propagation. After training, GTREN uncovers implicit correlations among sparse sensor measurements and successfully generalizes these relationships to reconstruct the entire flow field. Results indicate that, with only ten measurement points, the model accurately reconstructs both pressure and velocity distributions across the field. When compared to Computational Fluid Dynamics simulations, the mean squared error (MSE) of the reconstruction is as low as 0.05%. By contrast, a conventional graph neural network yields an MSE of 18.31%. We also provide a systematic analysis of how both the number and spatial arrangement of measurement points affect reconstruction accuracy. Our analysis reveals that, rather than merely increasing sensor count, strategically locating sensors within the core vortex-shedding region of the wake substantially improves reconstruction accuracy—offering practical guidance for flexible sensor deployment. Overall, GTREN achieves high-precision full-field reconstruction with very few pressure sensors, enabling real-time sparse-sensing flow-field monitoring.
Aiming at the technical challenge of reconstructing high-resolution flow fields from sparse flow field data, research on flow field reconstruction based on the physics-informed neural network (PINN) for underwater vehicles has been carried out. To begin with, the Lattice Boltzmann Method (LBM) was used to construct the benchmark flow field data (Timesequenced velocity and pressure fields in three-dimensional space) of the SUBOFF model, and the sparse flow field dataset is generated by spatial down-sampling of the benchmark flow field data. Subsequently, the fully data-driven algorithm and the PINN deep learning algorithm for SUBOFF flow reconstruction (with spatio-temporal coordinate points as inputs and velocity and pressure fields as outputs) were developed, the fully data-driven model and the PINN were obtained by training with the sparse flow data set. Eventually, only the spatiotemporal coordinate points from the benchmark flow field data were input into the two models, and the complete flow field was reconstructed using model inference inversion, and the results were compared and analyzed with the benchmark flow field data generated by the LBM, to validate the correctness of the reconstructed flow field for the two models. The results demonstrate that the PINN model is significantly more accurate than the fully data-driven model in reconstructing the flow field from sparse flow field data. The study provides a new flow field reconstruction solution that could contribute to the optimal design of the hydrodynamic performance of underwater vehicles and similar structures.
With the increasing use of deep neural networks as surrogate models for accelerating computational simulations in mechanics, the application of artificial intelligence in computational fluid dynamics has seen renewed interest in recent years. However, the application of deep neural networks for flow simulations has mainly concentrated on relatively simple cases of incompressible flows. The strongly discontinuous structures that appear in compressible flows dominated by convection, such as shock waves, introduce significant challenges when approximating the nonlinear solutions or governing equations. In this work, we propose a novel physics-constrained, flow-field-message-informed (FFMI) graph neural network for spatiotemporal flow simulations of compressible flows involving strong discontinuities. To enhance the nonlinear approximation capability of strong discontinuities, a shock detector method is leveraged to extract the local flow-field messages. These messages are embedded into the graph representation to resolve the discontinuous solutions accurately. A new FFMI sample-and-aggregate-based message-passing layer, which aggregates the edge-weighted attributes with node features on different hop layers, is then developed to diffuse and process the flow-field messages. Furthermore, an end-to-end paradigm is established within the encoder–decoder framework to transform the extracted information from the flow field into latent knowledge about the underlying fluid mechanics. Finally, a variety of one- and two-dimensional cases involving strong shock waves are considered to demonstrate the effectiveness and generalizability of the proposed FFMI graph neural network.
No abstract available
Flow temperature fields are more challenging to measure than the velocity fields. Accurate reconstruction of temperature distributions from the velocity data is crucial in heat transfer problems, especially for experimental data analysis. We apply a physics-informed neural network (PINN) to accurately infer temperature fields at arbitrary space-time points within the measurement domain, using only the velocity data and initial temperature field. The PINN minimizes a composite loss function that incorporates residuals from both the Navier–Stokes and energy equations. We validate this approach using three representative two-dimensional heat transfer cases: steady flow around a heated cylinder, synthetic jet impinging on a heated wall, and Rayleigh–Bénard convection. Recognizing that the energy and momentum equations are decoupled under incompressible conditions, we introduce an alternative network architecture in which the temperature is predicted via a dedicated subnetwork. This modification improves reconstruction accuracy by approximately 10%.
No abstract available
No abstract available
Abstract Representing complex flows by evolving vortex structures is an important principle in many investigations of wall-bounded turbulence. The practice of this principle benefits from the bi-directional transformation between the velocity field and the corresponding vortex field. While the velocity-to-vortex transformation could be implemented by various vortex identification criteria, few efforts have been devoted to the inverse process. This work develops a linear reconstruction method, which allows an effective reconstruction for the velocity field of wall turbulence based on a given vortex field. The vortex field is defined as a vector field by combining the swirl strength and the real eigenvector of the velocity gradient tensor. The reconstructed velocity fields are calculated by convolution operations on the vortex fields, with the kernel functions derived by the field-based linear stochastic estimation. The method can effectively recover the turbulent motions in a large scale range, showing clear advantages over the Biot–Savart formula in the near-wall region. The method is also employed to investigate the inducing effects of vortices at different heights. The wall-bounding effect on the induced motions is observed from the contribution spectra of vortices. The higher-order moments of the reconstructed streamwise velocity component present larger deviations from the original data, which is discussed and explained reasonably. At last, the vortex fields filtered by prescribed thresholds are employed to reconstruct the velocity fields. It is found that the strongest vortex components occupying 5 % of the total volume can reasonably recover the main flow features including both the near-wall streaks and the large-scale motions.
Abstract Reconstructing near-wall turbulence from wall-based measurements is a critical yet inherently ill-posed problem in wall-bounded flows, where limited sensing and spatially heterogeneous flow–wall coupling challenge deterministic estimation strategies. To address this, we introduce a novel generative modelling framework based on conditional flow matching for synthesising instantaneous velocity fluctuation fields from wall observations, with explicit quantification of predictive uncertainty. Our method integrates continuous-time flow matching with a probabilistic forward operator trained using stochastic weight-averaging Gaussian, enabling zero-shot conditional generation without model re-training. We demonstrate that the proposed approach not only recovers physically realistic, statistically consistent turbulence structures across the near-wall region but also effectively adapts to various sensor configurations, including sparse, incomplete and low-resolution wall measurements. The model achieves robust uncertainty-aware reconstruction, preserving flow intermittency and structure even under significantly degraded observability. Compared with classical linear stochastic estimation and deterministic convolutional neural network methods, our stochastic generative learning framework exhibits superior generalisation for unseen realisations under same flow conditions and resilience under measurement sparsity with quantified uncertainty. This work establishes a robust semi-supervised generative modelling paradigm for data-consistent flow reconstruction and lays the foundation for uncertainty-aware, sensor-driven modelling of wall-bounded turbulence.
No abstract available
Abstract We use Navier–Stokes-based linear models for wall-bounded turbulent flows to estimate large-scale fluctuations at different wall-normal locations from their measurements at a single wall-normal location. In these models, we replace the nonlinear term by a combination of a stochastic forcing term and an eddy dissipation term. The stochastic forcing term plays a role in energy production by the large scales, and the eddy dissipation term plays a role in energy dissipation by the small scales. Based on the results in channel flow, we find that the models can estimate large-scale fluctuations with reasonable accuracy only when the stochastic forcing and eddy dissipation terms vary with wall distance and with the length scale of the fluctuations to be estimated. The dependence on the wall distance ensures that energy production and energy dissipation are not concentrated close to the wall but are evenly distributed across the near-wall and logarithmic regions. The dependence on the length scale of the fluctuations ensures that lower wavelength fluctuations are not excessively damped by the eddy dissipation term and hence that the dominant scales shift towards lower wavelengths towards the wall. This highlights that, on the one hand, energy extraction in wall turbulence is predominantly linear and thus physics-based linear models give reasonably accurate results. On the other hand, the absence of linearly unstable modes in wall turbulence means that the nonlinear term still plays an essential role in energy extraction and thus the modelled terms should include the observed wall distance and length scale dependencies of the nonlinear term.
Streamwise velocity fluctuations in the inner-region of wall-bounded turbulent flows can be predicted by the model of Marusic, Mathis & Hutchins (2010). Only a single large-scale velocity signal from an outer position in the logarithmic region is needed for the model, and all other parameters are determined from a once-off calibration experiment. Here we elucidate part of the model by investigating the scale-dependent coherence magnitude and phase throughout the boundary layer. The collection of coherent scales exhibits a shift with respect to the reference position that is shown to be independent of scale; thus the large-scales are non-dispersive. Because these scales comprise a strong coherence, their signature in the inner-region is predicted from an input signal acquired at the geometric center of the log-region. Previously this was achieved via single-time stochastic estimation. Here we leverage the inherent advantages of spectral linear stochastic estimation for the prediction of these large-scales.
For wall-bounded flows, the model of Marusic et al. [Science 329, 193 (2010)] allows one to predict the statistics of the streamwise fluctuating velocity in the inner region, from a measured input signal in the logarithmic region. Normally, a user-defined large-scale portion of the input forms the large-scale content in the prediction by scaling its amplitude, as well as temporally shifting the signal to account for the physical inclination of these scales. Incoherent smaller scales are then fused to the prediction via universally expressed fluctuations that are subject to an amplitude modulation. Here we present a refined version of the model using spectral linear stochastic estimation, which eliminates a user-defined scale separation of the input. Now, an empirically derived transfer kernel comprises an implicit filtering via a scale-dependent gain and phase; this kernel captures the coherent portion in the prediction. An additional refinement of the model embodies a relative shift between the stochastically estimated scales in the prediction and the modulation envelope of the universal small scales. Predictions over a three-decade span of Reynolds numbers, Re-tau similar to O(10(3)) to O(10(6)), highlight promising applications of the refined model to high-Reynolds-number flows, in which coherent scales become the primary contributor to the fluctuating energy.
Interdisciplinary Graduate School of Engineering Sciences, Kyushu University, 6-1 Kasuga-koen, Kasuga 816-8580, Japan 1) Research Center for Plasma Turbulence, Kyushu University, 6-1 Kasuga-koen, Kasuga 816-8580, Japan 2) Research Institution for Applied Mechanics, Kyushu University, 6-1 Kasuga-koen, Kasuga 816-8580, Japan 3)Faculty of Arts and Science, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan
A conditional Gaussian framework for understanding and predicting complex multiscale nonlinear stochastic systems is developed. Despite the conditional Gaussianity, such systems are nevertheless highly nonlinear and are able to capture the non-Gaussian features of nature. The special structure of the system allows closed analytical formulae for solving the conditional statistics and is thus computationally efficient. A rich gallery of examples of conditional Gaussian systems are illustrated here, which includes data-driven physics-constrained nonlinear stochastic models, stochastically coupled reaction–diffusion models in neuroscience and ecology, and large-scale dynamical models in turbulence, fluids and geophysical flows. Making use of the conditional Gaussian structure, efficient statistically accurate algorithms involving a novel hybrid strategy for different subspaces, a judicious block decomposition and statistical symmetry are developed for solving the Fokker–Planck equation in large dimensions. The conditional Gaussian framework is also applied to develop extremely cheap multiscale data assimilation schemes, such as the stochastic superparameterization, which use particle filters to capture the non-Gaussian statistics on the large-scale part whose dimension is small whereas the statistics of the small-scale part are conditional Gaussian given the large-scale part. Other topics of the conditional Gaussian systems studied here include designing new parameter estimation schemes and understanding model errors.
Techniques for the experimental determination of velocity fields such as particle image velocimetry (PIV) can often be hampered by spurious vectors or sparse regions of measurement which may occur due to a number of reasons. Commonly used methods for detecting and replacing erroneous values are often based on statistical measures of the surrounding vectors and may be influenced by further poor data quality in the region. A new method is presented in this paper using Linear Stochastic Estimation for vector replacement (LSEVR) which allows for increased flexibility in situations with regions of spurious vectors. LSEVR is applied to PIV dataset to demonstrate and assess its performance relative to commonly used bilinear and bicubic interpolation methods. For replacement of a single vector, all methods performed well, with LSEVR having an average error of 11% in comparison to 14% and 18% for bilinear and bicubic interpolation respectively. A more significant difference was found in replacement of clusters of vectors which showed average vector angle errors of 10°, 9° and 6° for bilinear, bicubic and LSEVR respectively. Error in magnitude was 3% for both interpolation techniques and 1% for LSEVR showing a clear benefit to using LSEVR for conditions that require multiple clustered vectors to be replaced.
Many modern I.C. engines rely on some form of active control of injection, timing and/or ignition to help combat tailpipe out emissions, increase the fuel economy and engine drivability. However, development of these strategies is often optimised to suit the average cycle at each condition; an assumption that can lead to sub-optimal performance, especially an increase in particulate (PN) emissions as I.C. engine operation, and in-particular it’s charge motion is subject to cycle-to-cycle variation (CCV). Literature shows that the locations of otherwise repeatable large-scale flow structures may vary by as much 25% of the bore dimension; this could have an impact on fuel break-up and distribution and therefore subsequent combustion performance and emissions. In the presented work, a method is presented that allows full-field flow velocity information to be estimated in real-time from only a limited number of point velocity measurements using linear stochastic estimation (LSE). Three sensor arrangements – single bisecting ‘line-of-sight’, a central cluster and a circumferential ring - which are deemed applicable to implementation in an I.C. engine are compared over all test flow conditions; with all providing useful estimations of the flow field. It is shown how with even a modest number of point measurements it is possible to achieve at least 85% correlation between estimates and original data allowing cycle characterisation to be achieved. Information gathered from this technique could provide inputs to engine control strategies to account for the CCV of the in-cylinder flow.
No abstract available
The space-time structure of a turbulent natural convection boundary layer along a vertical flat plate in air has been examined using a thermocouple rake and a pair of hot-wire/cold-wire arrangements. The instantaneous temperature profile clearly indicates the evolution of a large-scale fluid motion originating in the outer region of the boundary layer. However, the space-time correlation measurements have not suggested any existence of quasi-coherent structures such as low-speed streaks and bursts. The convection velocity and characteristic length scales of turbulence have also been investigated. The convection velocity exceeds the local mean velocity in the near wall and outermost regions. The microscale of the thermal field shows a strongly anisotropic structure, and the integral scale indicates a large-scale structure normal to the wall.
No abstract available
Abstract Content of image described in text. Scalar turbulence, involving the chaotic transport of passive scalars such as temperature, pollution and chemical concentration, plays a critical role in scientific and engineering applications. Transport of scalars is largely affected by its inherent characteristic molecular diffusivity through the Schmidt number ( italic Sc Sc $\textit{Sc}$ , the ratio of fluid viscosity and diffusivity). This study explores the scalar convection speed and Taylor microscale in fully developed pipe flow with a localised scalar plume at Reynolds number italic Re equals 10 000 Re=10000 $\textit{Re} = 10\,000$ for 10 Superscript negative 3 Baseline less than or slanted equals italic Sc less than or slanted equals 10 10−3⩽Sc⩽10 $10^{-3} \leqslant \textit{Sc} \leqslant 10$ . We employ a hybrid approach combining direct numerical simulations for the velocity field with Lagrangian scalar tracking to capture the scalar field’s evolution under turbulent convection and molecular diffusion. The scalar concentration data are obtained in the Eulerian cylindrical coordinates. In a fixed range of streamwise positions, the scalar convection speed aligns with the convection speed of velocity fluctuations at italic Sc equals 1 Sc=1 $\textit{Sc} = 1$ , but it is smaller and larger than the velocity convection speed for italic Sc less than 1 Sc<1 $\textit{Sc} \lt 1$ and italic Sc greater than 1 Sc>1 $\textit{Sc} \gt 1$ , respectively, and also for the radial positions close to the pipe wall. The Taylor microscale scales approximately as italic Sc Superscript negative 1 divided by 2 Sc−1/2 $\textit{Sc}^{-1/2}$ for italic Sc greater than or slanted equals 1 Sc⩾1 $\textit{Sc} \geqslant 1$ , while for italic Sc less than 1 Sc<1 $\textit{Sc} \lt 1$ , it approximately follows italic Sc Superscript negative 0.03 Sc−0.03 $\textit{Sc}^{-0.03}$ . Along the streamwise positions, the scalar convection speed is found to gradually approach the velocity convection speed, regardless of whether the Schmidt number is larger or smaller than unity, along with the decay of scalar fluctuations. Overall, both the scalar convection velocity and scalar Taylor microscale depend on italic Sc Sc $\textit{Sc}$ and the mixing state.
No abstract available
Wall pressure fluctuations beneath turbulent boundary layers are a fundamental source of aerodynamic noise by exciting the wall structure, with their space-time characteristics serving as the basic ingredient for predicting the wall structural response. To this end, direct numerical simulations of fully developed compressible turbulent boundary layers at Mach numbers of 0.5, 1.2, and 2.0 are conducted to investigate wall pressure fluctuations comprehensively. The effects of Mach number on the single-point statistics of wall pressure fluctuations, such as the root mean square, skewness and flatness factors, probability density function, and frequency spectrum, are assessed to be very weak. Regarding the space-time characteristics, the convection velocity Uc determined by the space-time correlation of wall pressure fluctuations increases slightly with the Mach number, which only reflects the convective behavior of turbulent vortices. On the wavenumber–frequency spectrum, characteristic peaks of both the acoustic wave and convective vortices are identified. At Mach 0.5, the peaks of the fast (Uc+c) and slow (Uc−c) acoustic waves are unattached to others with c denoting acoustic speed, while only the peak of the fast acoustic wave is distinguishable from the convective peak at Mach 1.2 and 2.0. Due to the aerodynamic heating at supersonic conditions, the thermal effect on acoustic speed should be taken into account in determining the acoustic wavenumber. By introducing a convective Prandtl–Glauert parameter, a refined relation is proposed to provide a more accurate depiction of the acoustic domain in the wavenumber–frequency spectrum.
The origin and characteristics of large-scale pressure fluctuations have been attributed to the large-scale coherent motions in turbulent boundary layers, often termed “superstructures.” Existing studies lack a few key aspects, including limitations to lower Reynolds numbers, significant spatiotemporal aliasing, and limited measurement domains. This study uses synchronous wall-parallel stereoscopic particle image velocimetry (PIV) and wall-pressure measurements to quantify the characteristics of large-scale flow structures and their relationship with wall pressure fluctuations in a smooth-wall turbulent boundary-layer flow. These measurements are acquired at a friction Reynolds number [Formula: see text] in the upper part of the logarithmic layer ([Formula: see text]) under the influence of a small pressure gradient ([Formula: see text]). Space–time correlations reveal the presence of a correlation between wall pressure and velocity fluctuations. The wavenumber–frequency cross-spectra between wall pressure and velocity components are presented, showing the direct contribution of the streamwise and wall-normal velocity components in generating wall-pressure fluctuations, while the spanwise velocity component is considered insignificant.
Abstract Space–time correlations of velocity and high-Schmidt-number ( $Sc \approx 2000$ ) passive scalar fields are investigated in turbulent pipe flow using particle image velocimetry and planar laser-induced fluorescence, respectively. Both the velocity and scalar fields exhibit characteristic elliptical patterns in their respective space–time correlations. The elliptic approximation model, originally developed for the velocity field, is applied to estimate convection and sweeping velocities for both fields. In both fields, the convection velocity decreases, while the sweeping velocity increases, along the pipe radius. The convection velocity ratio between the scalar and velocity fields shows that high-Schmidt-number scalar fluctuations are advected faster than the velocity fluctuations. Similarly, the sweeping velocity of the scalar fluctuations is found to be larger than that of the velocity fluctuations. Furthermore, the high-Schmidt-number scalar is found to decorrelate more rapidly than the corresponding velocity, with the scalar Taylor microscale distinctly smaller than the velocity Taylor microscale.
No abstract available
Generation of the hydrodynamic or aerodynamic noise is a well-known problem for vehicles and structures moving in water or air media. Therefore, predicting and controlling the properties of the turbulent hydrodynamic pseudoacoustic sources is of great importance. This paper presents the results of an experimental study of the cross-correlations and cross-spectra of wall pressure fluctuations on the surface of a longitudinally streamlined flexible cylinder. The degree of conversion of flow energy into the energy of the wall pressure fluctuation field is shown to be limited by a certain threshold. Its value corresponds to the ratio of the root-mean-square wall pressure fluctuation to the dynamic pressure of the order of 0.01 on a hydraulically smooth streamlined surface under a turbulent boundary layer with zero pressure gradient. The space-time correlation of wall pressure fluctuations along the generatrix of a flexible extended cylinder decreases with increasing distance between observation points, and the maximum values of the cross-correlation are observed for longer delay times. Convection of coherent vortex structures over the streamlined cylinder surface leads to an increase in the coherence levels of the wall pressure fluctuation field. Moreover, changes in the phase spectrum presented by sloping curves are observed. Their slope decreases with increasing convective velocity. The rate of degeneration of the maximum values of the space-time correlation coefficient in a wide frequency range on a flexible cylinder is higher than on a plate. Small-scale vortices that generate high-frequency pressure oscillations degenerate faster and are transported more slowly than large-scale coherent vortical structures from the outer region of the boundary layer. The obtained results are useful for the designers of the towed linear receiving hydroacoustic arrays.
No abstract available
No abstract available
No abstract available
No abstract available
Abstract Stress–velocity cross-spectra provide critical insights into the wall turbulence dynamics, where second-order cross-spectra have been used to characterise the amplitude modulation of large-scale motions on smaller scales. Here, we investigate the higher-order stress–velocity cross-spectra. Through theoretical analysis, we derive an exact relationship demonstrating that the difference in convection velocity between streamwise Reynolds normal stress fluctuations ( $r$ ) and streamwise velocity fluctuations ( $u$ ) – termed the $r{-}u$ convection velocity difference – is governed jointly by the second- and fourth-order cross-spectra. A new ‘coherence similarity’ (CS) model is proposed, which reveals an approximate similarity between higher-order and second-order cross-spectra. As a result, the $r{-}u$ convection velocity difference can be explained in terms of second-order cross-spectral properties. Numerical validation confirms that the CS model predicts higher-order cross-spectra and the convection velocity difference accurately. Furthermore, the contours of stress–velocity cross-spectra undergo a structural transition from single-lobe to triple-lobe patterns with increasing wall distance, suggesting the presence of complex space–time coupling between $r$ and $u$ .
Abstract The convection velocity in high-Reynolds-number pipe flow was investigated using two-point correlations obtained from two laser Doppler velocimetry systems. The Reynolds number ranged from ${\textit{Re}}_{{\tau}}=3000$ to 20 800, and profiles were obtained from $y/R=0.002$ up to the pipe centre, where $R$ is the pipe radius. This study examines the scaling behaviour of convection velocity profiles derived from raw velocity signals, and the convection characteristics of very large-scale motions (VLSMs) and large-scale motions extracted via scale-separated or time-resolved velocity signals. The profiles show that convection velocities from raw signals exceed the local mean velocity near the wall and gradually approach it toward the centre. These profiles can be scaled using inner variables, namely $y^+$ and $\Delta x^+$ , where $\Delta x^+$ represents the measurement distance. Scale-separated convection velocities for VLSM-scale structures – defined as those larger than $5R$ – were higher than the unfiltered values and remained nearly constant up to $y^+ \leq 2000$ at ${\textit{Re}}_{{\tau}} \approx 20\,000$ . In this constant region, the convection velocity of VLSMs scaled well with the bulk velocity $U_{\textit{b}}$ , taking values of approximately $0.85U_{\textit{b}}$ . Furthermore, analysis of the time-resolved data highlights that, when applying Taylor’s frozen turbulence hypothesis, it is essential to consider both the scale dependence and the temporal fluctuations of the convection velocity, which reflect the underlying spatio-temporal dynamics of the flow structures. The present study provides valuable data for discussions on converting frequency-domain measurements into wavenumber space using Taylor’s hypothesis.
An experimental campaign dedicated to the characterization of the wall-normal velocity correlations in a zero pressure gradient turbulent boundary layer was performed. A double set of Laser Doppler Velocimetry benches were used to access two-point two-time correlations of the wall-normal velocity. The measurements analysis confirms several important hypotheses classically made to model wall pressure spectra from the velocity correlations. In particular, the ratio of the wall-normal Reynolds stress to the turbulent shear stress is confirmed to exhibit a large plateau in the logarithmic region. In addition, Taylor's hypothesis of frozen turbulence is well recovered for the wall-normal velocity fluctuations. The convection velocity for the wall-normal velocity fluctuations is also shown to evolve across the boundary layer, according to the mean velocity profile. Furthermore, the decorrelation time scale of velocity correlations appears to be increasing throughout the boundary layer thickness in accordance with the increase of the convection velocity. The results obtained with this original campaign will help improving models for wall pressure spectra, especially those based on the resolution of the Poisson equation for the pressure for which the wall pressure correlations are related to the wall-normal velocity correlations.
No abstract available
Primary and secondary Reynolds shear stresses, -(u'v')^^- and -(v'w')^^-, have been measured with a two-component LDV in a fully developed turbulent square-duct flow. Quadrant analysis has been performed for these Reynolds stresses in order to demonstrate typical flow structures responsible for their generation. In the near wall region close to the wall bisector, the primary Reynolds stress is generated by streamwise vortices through both ejection and sweep motions in a similar manner to turbulent boundary layers and two-dimensional channel flows. In the near-wall region apart from the wall bisector, on the other hand, the contribution of a sweep motion to the primary Reynolds stress is significantly reduced. In the corner region, apart from the wall bisector, the secondary Reynolds shear stress is induced by fluid motions from one wall toward another wall via an antisymmetric streamwise vortex. These motions transport the streamwise momentum near one wall, which is lower than that close to the wall bisector because of the no-slip constraint by the wall, toward another wall. As a result, in the corner region relatively-low-speed fluids inrush toward the wall, which leads to the reduction of a sweep motion apart from the wall bisector.
No abstract available
In this study, multi-scale vortical structures and vortex dynamics around a fixed-bed barchan dune have been experimentally investigated based on the particle image velocimetry (PIV) technique, wavelet transform, and the finite-time Lyapunov exponent (FTLE) method. It was found that the dynamic characteristics of a dune wake are predominated by large- and intermediate-scale coherent structures. Quadrant analysis of the Reynolds-stress distribution for the corresponding wavelet components revealed that ejection and sweep events are the main contributors to the whole field, while outward and inward interaction events just dominate the region near the dune crest. In addition, the process of ejection and sweeping motions associated with the turbulent bursting sequence can also be captured by applying proper orthogonal decomposition (POD) analysis of the decomposed velocity filed for the different wavelet components. Finally, a continuous development process of the different wavelet scale structures in the shear layer was visualized in the Lagrangian framework. The small-scale waves grow exponentially and gradually develop into larger-scale vortices when convected downstream until the reattachment point, and larger-scale vortices break into the smaller ones.
We present a novel approach to adjust shapes of the interrogation windows (IW) in Particle Image Velocimetry (PIV) measurements as a function of temporal and spatial local coherent motions. Lagrangian Coherent Structures (LCS) has been widely utilized to determine local flow boundaries. We propose using Finite-Time Lyapunov Exponent (FTLE) to quantify LCS separatrix boundaries (i.e. ridges) and adjust the interrogation window. We integrated the proposed method with a local optical flow PIV algorithm. The evaluation was performed using synthetic particle images of 2D homogeneous isotropic turbulence obtained from Direct Numerical Simulation (DNS). The results showed significant improvements in regions with complex flow behaviours, particularly shear, vortex and hyperbolic motions. We studied improvements of the velocity estimation in a real experiment of the wake flow behind a cylinder at Reynolds number equal to 3900. It was found that optical flow featured by coherency based interrogation window (coherent optical flow) reveals detailed vector field estimations in regions with complex behaviours inside the wake flow.
No abstract available
No abstract available
No abstract available
This study presents the application of time-resolved particle image velocimetry (TR-PIV) to measure the mean and fluctuating velocity components in a turbulent boundary layer (TBL) over an axisymmetric body of revolution. A narrow wall-normal strip of the flow was captured using a synchronised high-speed laser and camera at a recording frequency of up to 80 kHz. The resulting streamwise and wall-normal velocity TR-PIV data were validated against hot-wire anemometry measurements and direct numerical simulations (DNS) of a flat plate under matched flow conditions. The mean flow results showed good agreement between all methods, while the expected attenuation due to the spatial averaging was found in the TR-PIV turbulence statistics closer to the wall. A key outcome of this study is the establishment of an effective laser sheet thickness for the TR-PIV using DNS as a reference. This study fills a gap in understanding the spectral response and limitations of TR-PIV in such complex flows, particularly how spatial resolution and noise influence the accuracy of turbulence measurements. The TR-PIV streamwise velocity energy spectra were compared with DNS data that were spatially filtered to match the resolution of the TR-PIV and hot-wire. A transfer function was derived to determine cut-off wavelengths as a function of wall-normal distance. The cut-off wavelengths enabled the quantification of the resolvable turbulence scales within the TBL, revealing that the spatial resolution is a limiting factor for TR-PIV. The methodology was applied to locations with zero and favourable pressure gradients, providing insights into how pressure gradients influence spectral content and the limitations of TR-PIV in capturing the full range of turbulence scales. The outlined methodology is applicable more broadly and can be used to enhance the accuracy of experimental techniques in future boundary layer investigations.
No abstract available
No abstract available
No abstract available
No abstract available
No abstract available
This paper deals with the determination of instantaneous planar pressure fields from velocity data obtained by particle image velocimetry (PIV) in turbulent flow. The operating principles of pressure determination using a Eulerian or a Lagrangian approach are described together with theoretical considerations on its expected performance. These considerations are verified by a performance assessment on a synthetic flow field. Based on these results, guidelines regarding the temporal and spatial resolution required are proposed. The interrogation window size needs to be 5 times smaller than the flow structures and the acquisition frequency needs to be 10 times higher than the corresponding flow frequency (e.g. Eulerian time scales for the Eulerian approach). To further assess the experimental viability of the pressure evaluation methods, stereoscopic PIV and tomographic PIV experiments on a square cylinder flow (ReD = 9,500) were performed, employing surface pressure data for validation. The experimental results were found to support the proposed guidelines.
Obtaining accurate wall shear stress (WSS) measurements with high spatiotemporal resolution remains a significant challenge in turbulent boundary layers (TBLs) research. We present the adjustable spatio-temporal resolution cross correlation (ASTRCC) algorithm. This method combines temporal ensemble averaging with elongated interrogation windows (IWs). It resolves the competing demands between spatial and temporal resolution in PIV-based approaches. When validated against direct numerical simulation (DNS) data at friction Reynolds number Reτ=360, the ASTRCC algorithm demonstrates a 34% reduction in mean absolute error and a 36% reduction in root mean square error compared to the traditional single-row cross correlation (SRCC) method. Our experimental implementation in TBLs at Reτ=519 achieves high spatiotemporal resolution (Δx+≈8.5, Δt+≈0.6), facilitating detailed characterization of instantaneous WSS dynamics. Multi-scale decomposition analysis reveals three distinct structural contributions to WSS fluctuations. Large-scale structures (λ+≈103–104) account for 46% of total energy. Medium-scale structures (λ+≈102–103) contribute 37%. Small-scale structures (λ+≈101) represent 17%. Our quadrant analysis establishes that sweep events exhibit strong correlations with high-WSS regions, while ejection events predominate in low-WSS areas, with WSS disturbances convecting at 0.53U∞ (where U∞ is the free-stream velocity). These results provide quantitative evidence for the multi-scale nature of WSS generation and establish the pronounced influence of large-scale structures on near-wall WSS dynamics, thereby offering experimental foundations for enhanced turbulence modeling and flow control strategies.
No abstract available
No abstract available
PIV measurements near a moving interface are always difficult. This paper presents a PIV image pre-processing method that returns high spatial resolution velocity profiles near the interface. Instead of re-shaping or re-orientating the interrogation windows, interface tracking and an image transformation are used to stretch the particle image strips near a curved interface into rectangles. Then the adaptive structured interrogation windows can be arranged at specified distances from the interface. Synthetic particles are also added into the solid region to minimize interfacial effects and to restrict particles on both sides of the interface. Since a high spatial resolution is only required in high velocity gradient region, adaptive meshing and stretching of the image strips in the normal direction is used to improve the cross-correlation signal-to-noise ratio (SN) by reducing the velocity difference and the particle image distortion within the interrogation window. A two dimensional Gaussian fit is used to compensate for the effects of stretching particle images. The working hypothesis is that fluid motion near the interface is ‘quasi-tangential flow’, which is reasonable in most fluid-structure interaction scenarios. The method was validated against the window deformation iterative multi-grid scheme (WIDIM) using synthetic image pairs with different velocity profiles. The method was tested for boundary layer measurements of a supersonic turbulent boundary layer on a flat plate, near a rotating blade and near a flexible flapping flag. This image pre-processing method provides higher spatial resolution than conventional WIDIM and good robustness for measuring velocity profiles near moving interfaces.
No abstract available
No abstract available
Accurate reconstruction of atmospheric wind fields is essential for applications such as weather forecasting, hazard prediction, and wind energy assessment, yet conventional instruments leave spatio-temporal gaps within the lower atmospheric boundary layer. Unmanned aircraft systems (UAS) provide flexible in situ measurements, but individual platforms sample wind only along their flight trajectories, limiting full wind-field recovery. This study presents a framework for reconstructing four-dimensional atmospheric wind fields using measurements obtained from a coordinated UAS swarm. A synthetic turbulence environment and high-fidelity multirotor simulation are used to generate training and evaluation data. Local wind components are estimated from UAS dynamics using a bidirectional long short-term memory network (Bi-LSTM) and assimilated into a physics-informed neural network (PINN) to reconstruct a continuous wind field in space and time. For local wind estimation, the bidirectional LSTM achieves root-mean-square errors (RMSE) of 0.064 and 0.062 m/s for the north and east components in low-wind conditions, increasing to 0.122 to 0.129 m/s under moderate winds and 0.271 to 0.273 m/s in high-wind conditions, while the vertical component exhibits higher error, with RMSE values of 0.029 to 0.091 m/s. The physics-informed reconstruction recovers the dominant spatial and temporal structure of the wind field up to 1000 m altitude while preserving mean flow direction and vertical shear. Under moderate wind conditions, the reconstructed mean wind field achieves an overall RMSE between 0.118 and 0.154 m/s across evaluated UAS configurations, with the lowest error obtained using a five-UAS swarm. These results demonstrate that coordinated UAS measurements enable accurate and scalable four-dimensional wind-field reconstruction without dedicated wind sensors or fixed infrastructure.
Short-term forecasting of cloud-induced power ramps across a network of distributed photovoltaic (PV) or irradiance sensors is a recognised pain point for grid operators. A natural idea is to make the graph neural network (GNN) advection-aware: connect each site to the sites upwind of it, with edge time-lags set by the cloud-motion vector (CMV), so that a ramp is propagated forward before it physically arrives. Using a controlled synthetic testbed with a known wind field, we show that (i) with a realistic cross-correlation CMV estimate, an explicit advection graph does not beat a plain static or learned-adjacency spatiotemporal GNN; (ii) roughly half of the benefit available from a perfect CMV comes simply from providing an accurate motion vector as an input feature, not from graph structure; and (iii) advection helps only when the advective displacement over the forecast horizon, v*H, fits inside the sensor network. Motivated by (ii), we introduce a small self-supervised cloud-motion estimator -- a position-aware encoder trained only on a multi-lag optical-flow reconstruction objective with an annealed kernel -- that recovers the true wind vector to 2-4 degrees median angular error, 2-4x better than the classical cross-correlation method across every wind regime. Freezing this estimator and feeding its vector to the forecaster closes about 60% of the oracle-CMV RMSE gap at moderate wind (8-15% RMSE reduction over no advection), with no external wind data. We also report a negative result for a spatially-coherent probabilistic head. All claims are established on a single synthetic simulator; we discuss why real-network validation is the necessary next step and outline it.
Physics-Informed Neural Networks (PINNs) offer a powerful paradigm for flow reconstruction, seamlessly integrating sparse velocity measurements with the governing Navier-Stokes equations to recover complete velocity and latent pressure fields. However, scaling such models to large spatiotemporal domains is hindered by computational bottlenecks and optimization instabilities. In this work, we propose a robust distributed PINNs framework designed for efficient flow reconstruction via spatiotemporal domain decomposition. A critical challenge in such distributed solvers is pressure indeterminacy, where independent sub-networks drift into inconsistent local pressure baselines. We address this issue through a reference anchor normalization strategy coupled with decoupled asymmetric weighting. By enforcing a unidirectional information flow from designated master ranks where the anchor point lies to neighboring ranks, our approach eliminates gauge freedom and guarantees global pressure uniqueness while preserving temporal continuity. Furthermore, to mitigate the Python interpreter overhead associated with computing high-order physics residuals, we implement a high-performance training pipeline accelerated by CUDA graphs and JIT compilation. Extensive validation on complex flow benchmarks demonstrates that our method achieves near-linear strong scaling and high-fidelity reconstruction, establishing a scalable and physically rigorous pathway for flow reconstruction and understanding of complex hydrodynamics.