摘要
针对因叶片安装误差、外界运行环境变化等因素导致机组运行最优桨距角与理论桨距角不一致而造成发电量损失的问题,提出了一种基于运行数据驱动的最优桨距角辨识方法。首先,根据风电机组发电机转速、功率和桨距角等运行数据进行工况识别,通过改进的DBSCAN(density-based spatial clustering of applications with noise)聚类方法剔除离群数据。然后,基于最小二乘法拟合“风速–功率–桨距角”模型特性曲面,通过Levenberg-Marquardt算法求解特性曲面变量之间的函数关系,进而辨识出不同平均风速下最优桨距角。通过算例进行验证,结果表明,所提出的风电机组最优桨距角辨识方法能够找到各平均风速下最优桨距角。
Aiming at the problem that the optimal pitch angle of the unit operation is inconsistent with the theoretical pitch angle due to the blade installation error and the continuous change of the external operating environment of the wind turbine,a method of identifying the optimal paddle pitch angle driven by the operation data is proposed.Firstly,the working conditions are identified and screened according to the operating data of the wind turbine generator speed,power and pitch angle,and the outlier data is eliminated by the improved DBSCAN(density-based spatial clustering of applications with noise)clustering method.Then,based on the least squares method,the characteristic surface of the"wind speed-power-pitch angle"model is fitted,and the function relationship between the characteristic surface variables is solved by Levenberg-Marquardt algorithm,and the optimal pitch angle under different average wind speeds is identified.Through the simulation,the results show that the proposed optimal pitch angle identification method for wind turbines can find the optimal pitch angle at each average wind speed.
作者
柏文超
刘颖明
王晓东
高兴
张书源
BAI Wenchao;LIU Yingming;WANG Xiaodong;GAO Xing;ZHANG Shuyuan(School of Electrical Engineering,Shenyang University of Technology,Shenyang 110870,China)
出处
《电力科学与工程》
2022年第9期38-44,共7页
Electric Power Science and Engineering
