Search

• The machine learning framework achieves favorable accuracy and efficiency in the wake prediction. • ADM-R/RANS coupling model can efficiently produce big wake datasets for machine learning model. • For an aligned turbine row, power drops at the second turbine and gradually stabilizes afterwards. • Turbulence model plays an important role in accurate power prediction of wind farm. In the paper, a novel framework that employs the machine learning and CFD (computational fluid dynamics) simulation to develop new wake velocity and turbulence models with high accuracy and good efficiency is proposed to improve the turbine wake predictions. An ANN (artificial neural network) model based on the back-propagation (BP) algorithm is designed to build the underlying spatial relationship between the inflow conditions and the three-dimensional wake flows. To save the computational cost, a reduced-order turbine model ADM-R (actuator disk model with rotation), is incorporated into RANS (Reynolds-averaged Navier-Stokes equations) simulations coupled with a modified k - ε turbulence model to provide big datasets of wake flow for training, testing, and validation of the ANN model. The numerical framework of RANS/ADM-R simulations is validated by a standalone Vestas V80 2 MW wind turbine and NTNU wind tunnel test of double aligned turbines. In the ANN-based wake model, the inflow wind speed and turbulence intensity at hub height are selected as input variables, while the spatial velocity deficit and added turbulence kinetic energy (TKE) in wake field are taken as output variables. The ANN-based wake model is first deployed to a standalone turbine, and then the spatial wake characteristics and power generation of an aligned 8-turbine row as representation of Horns Rev wind farm are also validated against Large Eddy Simulations (LES) and field measurement. The results of ANN-based wake model show good agreement with the numerical simulations and measurement data, indicating that the ANN is capable of establishing the complex spatial relationship between inflow conditions and the wake flows. The machine learning techniques can remarkably improve the accuracy and efficiency of wake predictions. [ABSTRACT FROM AUTHOR]

In the design of wind energy farms, the loss of power should be seriously considered for the second wind turbine located inside the wake region of the first one. The rotation of the first wind-front rotor generates a high-vorticity wake with turbulence, and a suitable model is required in computational fluid dynamics (CFD) to predict the deficit of energy of the second turbine for the given configuration. A simplified numerical model based on the classical momentum theory is proposed in this study for multiple wind turbines, which is proposed with a couple of tuning parameters applied to Reynolds-averaged Navier-Stokes (RANS) analysis, resulting in a remarkable reduction of computational load compared with advanced methods, such as large eddy simulation (LES) where two parameters reflect on axial and rotational wake motion, simply tuned with the wind-tunnel test and its corresponding LES result. As a lumped parameter for the figure of merit, we regard the normalized efficiency on the kinetic power output of computational domain, which should be directed to maximize for the optimization of wind farms. The parameter surface is plotted in a dimensionless form versus intervals between turbines, and a simple correlation is obtained for a given hub height of 70% diameter and a fixed rotational speed tuned from the experimental data in a wide range. [ABSTRACT FROM AUTHOR]