Optimization of Wind Farm Yaw Offset Angle using Online Genetic Algorithm with a Modified Elitism Strategy to Maximize Power Production

Abstract

The wake interaction in a wind farm occurs when the front turbines block the flow of wind to the turbines behind them, causing a total power loss of approximately 10–25%. Wake interactions can be redirected to reduce bad impacts by optimizing the yaw offset angles. Optimization of the yaw offset angle can increase the total power of the wind farm by approximately 6–9%. However, the fluctuating wind flow angle in the environment causes the behavior of the wake interaction to change, making it difficult to optimize the yaw offset angles. Therefore, this study proposes an online genetic algorithm with a modified elitism strategy to overcome this problem. The contribution of this study is to improve the performance of the genetic algorithm by modifying the elitism strategy in order to optimize the yaw offset angle for each turbine adaptively to a wind farm operating in a dynamic environment. The optimal yaw offset angles are stored in the elite population for various wind flow angles and then reinserted into the search population in each generation according to the actual wind flow angles. A Gaussian-based analytical wake interaction model under a yawed condition developed by Shapiro is employed in this study to evaluate the total power of a wind farm. This study resulted in a convergence speed that was 3.8 times faster than the classical elitism strategy. At several wind flow angles of 270°, 315°, and 360°, an average power increase of 10.52% was obtained. This study shows that the modification of the elitism strategy can increase the convergence speed to adaptively track the optimal yaw offset angle at various wind flow angles, so that the average increase in wind farm power is 1.94% higher than in previous studies.