Self-consistent model for active control of wind turbine wakes

Abstract

Active wake control (AWC) has emerged as a promising strategy for enhancing wind turbine wake recovery, but accurately modelling its underlying fluid mechanisms remains challenging. This study presents a computationally efficient wake model that provides end-to-end prediction capability from rotor actuation to wake recovery enhancement by capturing the coupled dynamics of wake meandering and meanflow modification, requiring only two inputs: a reference wake without control and a user-defined AWC strategy. The model combines physics-based resolvent modelling for large-scale coherent structures and an eddy viscosity modelling for small-scale turbulence. A Reynolds stress model is introduced to account for the influence of both coherent and incoherent wake fluctuations, so that the time-averaged wake recovery enhanced by the AWC can be quantitatively predicted. Validation against large-eddy simulations (LES) across various AWC approaches and actuating frequencies demonstrates the model's predictive capability, accurately capturing AWC-specific and frequency-dependent mean wake recovery with less than 8% error from LES while reducing computational time from thousands of CPU hours to minutes. The efficiency and accuracy of the model makes it a promising tool for practical AWC design and optimization of large-scale wind farms.