Wind farm cooperative control under unsteady inflow conditions considering dynamic wake interactions

Overlooking wake dynamics undermines the real-time control performance of wind farms. This paper proposes a novel dynamic wind farm control framework that integrates a mid-fidelity dynamic wake model, the FLORIDyn model, with surrogate model optimization, the DYCORS algorithm, to achieve optimal coordinated control settings within the yaw update interval accurately and efficiently. The framework is tested in a 6-turbine wind farm exposed to time-varying inflow conditions over 2400 s, with the conventional steady framework as the comparison. Additionally, parametric studies on yaw update interval and wind variability are conducted to explore the applicability of the dynamic framework under different inflow and operation conditions. Results indicate that achieving the anticipated power output of the steady wind farm control framework is challenging in a realistic wind farm setting. The proposed dynamic wind farm control framework enhances the power benefits of wake redirection compared to the steady framework, achieving a 2.22 % increase in power gains. The dynamic optimal control is more sensitive to yaw update interval variations than the greedy control. A smaller Hurst exponent, indicating increased stationarity of the inflow condition, reduces the power disparities between steady and dynamic control optimizations. Directional variability imposes a more distinct impact on control benefits than speed variability.