베이지안 기반 후류 제어를 통한 풍력 단지 출력 향상

초록

This study proposes an innovative method for improving the energy production of an onshore wind farm by integrating Bayesian algorithm-based wake steering control. The proposed method dynamically adjusts turbine yaw angles to mitigate wake effects, thereby optimizing the wake across the wind farm. By applying the predictive optimization capabilities of Bayesian algorithms, the method significantly improves wind farm power output and shows strong potential for large-scale implementation. FLORIDyn, the dynamic version of the FLORIS model, is used to simulate wake interactions and to evaluate the performance of the proposed control strategy compared to existing methods. Bayesian optimization effectively strikes a good balance between computational effort and performance. Simulation results are presented that demonstrate a considerable increase in overall energy production by minimizing wake effects.

키워드