Optimization of wind farm power output using wake redirection control

Abstract

The wake effect, which is caused by the upstream turbines in a wind farm, adversely affects the efficiency of downstream turbines, leading to reduced energy generation and increased turbine fatigue loading. To mitigate this effect, a real-time wind farm control technique, i.e., wake redirection control (WRC), employing teaching learning-based optimization (TLBO) is introduced. This technique redirects the wakes away from the downstream turbines in real time, allowing them to generate more power by sacrificing some of the power generated by the upstream turbines. As a result, the total power generated by the wind farm is maximized. A low-fidelity 20-turbine real-life offshore wind farm is modeled and simulated in FLORISSE_M, the MATLAB version of the FLORIS (FLOw Redirection and Induction in Steady-state). The power produced by the wind farm model is maximized in real time by employing TLBO. The optimization results (i.e., the optimized yaw angles) are validated using the corresponding high-fidelity wind farm model developed in SOWFA (Simulator fOr Wind Farm Applications). Various results are presented to demonstrate that the TLBO-based WRC positively affects the power generated by the wind farm.