Finite Control Set MPC Yaw Control Method of Wind Farms Based on a Dynamic Wake Model

The wake effect inside wind farms reduces the inflow wind speed and increases the turbulence intensity of downstream turbines, resulting in power loss and increased fatigue loads. Active yaw control can mitigate wake interference through collaborative optimization of turbine yaw angles. However, most existing methods rely on steady-state wake models, which fail to capture the dynamic delay characteristics of wakes and usually lead to excessive yaw actuation losses. To address these issues, this paper constructs a dynamic wake model suitable for real-time control based on the OFF dynamic wake framework (OnWARDS, FLORIDyn, and FLORIS), adopting an improved three-dimensional analytical wake model at the lowest level. On this basis, a finite control set model predictive control (MPC) active yaw controller is designed. Aiming to maximize power generation and minimize yaw loss, the controller realizes rolling optimization of yaw actions combined with ARIMA-based wind direction prediction and particle swarm optimization. Simulations on the 4 × 4 turbine array of the Horns Rev I wind farm show that the proposed method increases the total power by 2.25%, which is 0.79% higher than that obtained by the deadband controller. It results in lower power loss for upstream turbines and higher power gain for downstream turbines, reduces the total yaw travel by nearly 1000° compared with the deadband controller, and produces smaller power fluctuations under sharply changing wind directions.