摘要
实际电网中监测数据存在不确定性噪声、通信丢包导致的异常数据项,将会给光伏发电规律的总结与电网运行与调度的决策产生带来困难。光伏电站输出功率受到多种因素影响,包括太阳辐射度、环境温度、太阳辐射面积等,而光伏电站输出功率在相同气象条件下存在相似性。本文考虑光伏电站输出功率的多种影响因素,训练RBF人工神经网络作为状态转换方程。然后基于Sigma点卡尔曼滤波理论对光伏电站输出功率信息进行滤波。算例结果表明,所提方法能够有效修复光伏电站监测数据。
Measurement data may be delayed, reordered or even lost in the actual power system, which will affect power system planning anddispatch decision-making. The output of a PV station is affected by many factors, including solar radiation , temperature and solar radiation area. Theoutput data of PV stations is similar to historical data in the same weather condition situations. This paper proposes a novel method to repair the outputdata of PV station. The method considers many factors affecting the output of PV stations and establish the state transition equation by training the artificialneural network. Then the output data of PV station is repaired based on Sigma Kalman filter. Lastly, this method is examined by the measured data ofa PV station in Wuxi and results show effectiveness in data repairing.
作者
俞娜燕
李向超
费科
倪晓宇
任佳琦
YU Na-yan;LI Xiang-chao;FEI Ke;NI Xiao-yu;REN Jia-qi(State Grid Jiangsu Electric Power Co.,Ltd.Wuxi Power Supply Branch,Wuxi Jiangsu 214000;Wuxi Yangsheng Polytron Technologies Inc,Wuxi Jiangsu 214106;School of Energy and Electrical Engineering,Hohai University,Nanjing Jiangsu 210098)
出处
《数字技术与应用》
2018年第8期32-34,共3页
Digital Technology & Application
