Abstract: Real-time and accurate dynamic wake information is essential for wind resource assessment and the optimization of wind farm operations. To further understand the wake characteristics of wind turbines, we propose a hierarchical learning approach integrated with a deep neural network-based prediction method. The integrated framework combines physical and mathematical models, enabling 3D spatiotemporal wind field predictions with minimal measured data requirements. Evaluation and validation results demonstrate that the proposed method achieves accurate ultra-short-term wake predictions across the entire domain with minimal prediction errors. Compared with conventional methods, the proposed hierarchical learning framework markedly lowers the training-data requirements of physics-informed neural networks for large-scale flow-field prediction while maintaining high accuracy. In addition, it demonstrates superior performance in both local and global wake forecasts, offering practical insights for efficient turbine operation and wake analysis.

基于分层学习与深度神经网络的阵列流场预测方法

[1]AmiriMM, ShadmanM, EstefenSF, 2024. A review of physical and numerical modeling techniques for horizontal-axis wind turbine wakes. Renewable and Sustainable Energy Reviews, 193:114279.

[2]BaiHM, WangNN, WanDC, 2023. Numerical study of aerodynamic performance of horizontal axis dual-rotor wind turbine under atmospheric boundary layers. Ocean Engineering, 280:114944.

[3]BastankhahM, Porté-AgelF, 2014. A new analytical model for wind-turbine wakes. Renewable Energy, 70:116-123.

[4]FrandsenS, BarthelmieR, PryorS, et al., 2006. Analytical modelling of wind speed deficit in large offshore wind farms. Wind Energy, 9(1-2):39-53.

[5]HeRY, YangHX, SunHY, et al., 2021. A novel three-dimensional wake model based on anisotropic Gaussian distribution for wind turbine wakes. Applied Energy, 296:117059.

[6]JensenNO, 1983. A Note on Wind Generator Interaction. Risø National Laboratory, Roskilde, Denmark.

[7]LiSY, ZhangMR, PiggottMD, 2023. End-to-end wind turbine wake modelling with deep graph representation learning. Applied Energy, 339:120928.

[8]LiuRL, PengLL, HuangGQ, et al., 2023. A Monte Carlo simulation method for probabilistic evaluation of annual energy production of wind farm considering wind flow model and wake effect. Energy Conversion and Management, 292:117355.

[9]LiuSY, LiQS, LuB, et al., 2025. Prediction of offshore wind turbine wake and output power using large eddy simulation and convolutional neural network. Energy Conversion and Management, 324:119326.

[10]PurohitS, NgEYK, Syed Ahmed KabirIF, 2022. Evaluation of three potential machine learning algorithms for predicting the velocity and turbulence intensity of a wind turbine wake. Renewable Energy, 184:405-420.

[11]SongJY, WangL, XinZQ, et al., 2025. Wake field prediction of a wind farm based on a physics-informed neural network with different spatiotemporal prediction performance improvement strategies. Theoretical and Applied Mechanics Letters, 15(2):100577.

[12]SunHY, QiuCY, LuL, et al., 2020. Wind turbine power modelling and optimization using artificial neural network with wind field experimental data. Applied Energy, 280:115880.

[13]SunSX, LiuS, LiuJ, et al., 2019. Wind field reconstruction using inverse process with optimal sensor placement. IEEE Transactions on Sustainable Energy, 10(3):1290-1299.

[14]SunSX, LiuS, ChenMX, et al., 2020. An optimized sensing arrangement in wind field reconstruction using CFD and POD. IEEE Transactions on Sustainable Energy, 11(4):2449-2456.

[15]SunSX, CuiSS, HeT, et al., 2024. An integrated deep neural network framework for predicting the wake flow in the wind field. Energy, 291:130400.

[16]WangJS, YinXX, LiuYK, et al., 2023. Optimal design of combined operations of wind power-pumped storage-hydrogen energy storage based on deep learning. Electric Power Systems Research, 218:109216.

[17]WangL, DongM, YangJ, et al., 2024. Wind turbine wakes modeling and applications: past, present, and future. Ocean Engineering, 309(Pt 1):118508.

[18]WangLY, ChenM, LuoZH, et al., 2024. Dynamic wake field reconstruction of wind turbine through physics-informed neural network and sparse LiDAR data. Energy, 291:130401.

[19]XuS, ZhuangTG, ZhaoWW, et al., 2023. Numerical investigation of aerodynamic responses and wake characteristics of a floating offshore wind turbine under atmospheric boundary layer inflows. Ocean Engineering, 279:114527.

[20]XuS, YangXL, ZhaoWW, et al., 2024. Numerical analysis of aero-hydrodynamic wake flows of a floating offshore wind turbine subjected to atmospheric turbulence inflows. Ocean Engineering, 300:117498.

[21]ZhanLY, WangZF, ChenYR, et al., 2024. Ada²MF: dual-adaptive multi-fidelity neural network approach and its application in wind turbine wake prediction. Engineering Applications of Artificial Intelligence, 137:109061.

[22]ZhangJC, ZhaoXW, 2021. Spatiotemporal wind field prediction based on physics-informed deep learning and LiDAR measurements. Applied Energy, 288:116641.

[23]ZhouL, WenJH, WangZK, et al., 2023. High-fidelity wind turbine wake velocity prediction by surrogate model based on d-POD and LSTM. Energy, 275:127525.

Similar articles