A double Gaussian wake model for single and multiple yawed wind turbines

Accurate prediction of turbine wake is essential for optimal control of wind turbine yaw in wind farms. In this study, a new double Gaussian yawed wake model for wind turbines is proposed, and computational fluid dynamic (CFD) simulations with the Realizable k-ε turbulence model and a modified actuator disk method (ADM) are performed to compare the performance of this model with other wake models. The reliability of CFD is accomplished by comparison with wind tunnel tests. A single wind turbine considering three incoming wind velocities and four yawed angles is adopted to validate the accuracy of the proposed wake model, and the results show that the RMSE and MAE errors are reduced by 1.3 %∼6.1 % and 1.3 %∼6.1 %, respectively. The wake recovery length becomes shorter with increasing thrust coefficient for smaller yaw angles and remains constant for larger yaw angles. Furthermore, the superposition effect of multiple wakes is studied by square-array and rhombic-array multiturbines with four wind directions and four yawed angles. The results suggest that the sum-of-squares superposition (SS) model has the highest accuracy compared to the other three superposition models, reducing the RMSE and MAE errors of wake prediction by 1.1 %∼8.6 % and 0.6 %∼4.3 %, respectively. Therefore, the proposed wake model can predict the wind distribution in real wind farms, which is of great significance for engineering applications.