基于区域负荷的配电网超短期负荷预测  

SHORT-TERM LOAD FORECASTING OF DISTRIBUTION NETWORKS BASED ON PARTITION LOADS

作　　者：刘健 勾新鹏 徐精求[2] 党海锋 董海鹏 

机构地区：[1]陕西银河电力自动化股份有限公司,陕西省西安市710075 [2]西安科技大学自动化系,陕西省西安市710054

出　　处：《电力系统自动化》2003年第19期34-37,共4页Automation of Electric Power Systems

摘　　要：建立了以馈线开关为节点、以馈线为边的配电网简化模型 ,并采用负荷作为节点和边的权 ,给出了节点的负荷与边的负荷的关系。采用对负荷历史数据和温度历史数据进行统计分析的方法 ,对超出历史温度范围的情形采用最小二乘法拟合 ,得出了配电区域的超短期负荷预测结果 ,利用节点的负荷与边的负荷的关系进一步得出全配电网的超短期负荷预测结果。配电网运行方式的改变并不影响所提出的方法的预测精度。实际应用结果证实了所提出的方法的可行性。A simplified model of distribution networks is established, i.e., feeder switches are regarded as nodes while feeders are regarded as edges. The nodes and edges are weighted by load values. Relationships of node loads and edge loads are given. Based on a statistical method of the historical information of load and temperature, this paper describes the process of short-term load forecasting of individual distribution parts. A least squares approach is introduced to handle the case when the temperature exceeds the range of the historical data. Consequently, the short-term load forecasting results of the whole distribution network are obtained by using the relationships of node loads and edge loads. The proposed method has the advantage of that the change of topology of distribution network has no influence on the accuracy of forecasting result. The feasibility of the proposed method is proved by practice.

关 键 词：日负荷预测 配电网 配电自动化 

分 类 号：TM715[电气工程—电力系统及自动化] TM727.2