基于尾流判定的风电场偏航角分群优化控制

Clustered Optimization Control of Wind Farm Yaw Angle Based on Wake Determination

为了优化风电场整体的功率输出和疲劳载荷,提出分群优化方法。建立风机尾流干扰的神经网络判定模型,以风机的相对坐标和来流风速作为特征值来判断风机之间是否存在尾流干扰;引入连通分量分解算法,将风电场划分为多个互相无尾流干扰的气动解耦机群;采用改进的自适应评估粒子群算法,以提升功率和抑制载荷为目标对所有机群进行并行优化。结果表明:与贪婪算法和整体优化方法相比,所提分群优化方法在功率提升和载荷抑制方面的效果均较好,且其稳定性更好。

In order to optimize the overall power output and fatigue load of the wind farm,an optimization method of wind turbine yaw angle group was proposed.A neural network judgment model for wind turbine wake interference was established,and the relative coordinates of the wind turbine and the incoming wind speed were taken as characteristic values,to determine whether there was wake interference between the fans.The connected component decomposition algorithm was introduced to divide the wind farm into multiple aerodynamic decoupling groups without wake interference.With the goal of increasing power and suppressing load,the improved adaptive evaluation particle swarm algorithm was used to optimize all swarms in parallel.Results show that compared with the greedy algorithm and the overall optimization method,the group optimization method has better effects on power improvement and load suppression,and its stability is better.