基于风电机组运行数据的筛选算法分析

Screening Algorithm Analysis Based on the Operating Data of Wind Turbine

为了更准确全面的计算风电场发电量,准确评估风电场效能情况,本文基于风电场中风电机组运行数据,同时考虑风电机组类型及其实际运行情况,提出3种筛选算法,区分正常数据与降容数据,同时创新性的利用标签分类的方法,对风电机组数据采集与监视控制系统(supervisory control and data acquisition,SCADA)得到的10min数据进行标记。据此计算风电机组的损失电量。该算法的提出完善了风电行业对于风电机组10min数据、异常申诉数据等运行数据划分的欠缺,并且更加全面有效的计算、描绘出了风电场发电水平。

In order to calculate the power generation of wind farmsmore accurately and comprehensively,and evaluate the efficiency of wind farms accurately,this paper proposes three screening algorithms based on the operating data of wind turbines in wind farms,taking into account the type of wind turbines and their actual operation,to distinguish between normal data and derating data,and at the same time innovative use of label classification methods to mark the 10-minute data obtained by the supervisory control and data acquisition system(SCADA).Based on this,the power loss of the wind turbine is calculated.The proposed algorithms complete the shortcomings in the wind power industry’s division of operating data such as wind turbine 10-minute data and abnormal data,and calculate and depict the level of wind farm power generation more comprehensively and effectively.