数据驱动的风电机组偏航参数适应性优化

为了增强风电机组偏航系统自适应水平,提升风能利用率,提出一种基于K近邻聚类（KNN）算法风电机组偏航控制参数优化方法。为准确描述风向变化,建立改进Weibull概率分布建立风向评估模型,即以风向波动的幅值（A）和波动持续时间（T）作为风况的数据标签来描述风向。对比风电机组不同偏航参数下的运行数据确定聚类中心（已知风况下的最佳偏航参数）,通过基于KNN算法的风电机组偏航控制参数优化模型,得到不同风况下风电机组最佳的偏航参数。通过对风电机组运行数据进行算例分析表明,该方法高风速时可提升风电机组发电效率,并在低风速时减少偏航启动次数。

DATA-DRIVEN ADAPTIVE OPTIMIZATION OF WIND TURBINE YAW PARAMETERS

The frequency and range of wind direction change under different geographical conditions are different, and the adaptive level of wind turbine yaw control is poor, which significantly affects the power generation. In order to enhance the adaptive level of wind turbine yaw system and improve the utilization of wind energy, a wind turbine yaw control parameter optimization method based on K-nearest neighbor clustering （KNN） algorithm is proposed. In order to accurately describe wind direction changes, an improved Weibull probability distribution was established to establish a wind direction evaluation model. The amplitude of wind direction fluctuations （A） and the duration of fluctuations （T） are used as data labels for wind conditions to describe the wind direction. The operating data of wind turbines under different yaw parameters are compared to determine the clustering center （the best yaw parameter under known wind conditions）. The wind turbine yaw control parameters based on the KNN algorithm are optimized. Therefore, the best yaw parameters of wind turbines under different wind conditions are obtained. Case studies show that the proposed method can improve the power generation efficiency of wind turbines at high wind speeds and reduce the number of yaw starts at low wind speeds.