基于点模式匹配的电力系统短期负荷预测  

A NEW APPROACH TO SHORT-TERM LOAD FORECASTING BASED ON POINT PATTERN MATCHING

作　　者：赵世伟[1] 吴捷[1] 刘永强[1] 杨苹[1] 

机构地区：[1]华南理工大学电力学院,广东省广州市510640

出　　处：《电力系统自动化》2003年第16期62-66,共5页Automation of Electric Power Systems

摘　　要：输电网短期负荷预测是电力调度的基本依据。文中基于点模式策略 ,并借助模拟退火算法 ,提出了一种短期负荷预测的新方法。从模式的角度看 ,负荷的变动也可以看成是模式的转换。该方法首先通过相关性分析找出输入输出数据关系以及各输入数据所占的权重 ;然后由点模式匹配法求得各输入数据的变换值 ;最后由变换值、输入数据及其权重得到短期负荷预测的结果。将该方法用于一个实际系统 。Short-term load forecasting is the basis of power dispatching. On the basis of point pattern matching and using simulated annealing algorithm, a new short-term load forecasting approach is proposed. From the point of view of the patterns, load variation can be regarded as conversion between the patterns. Therefore, firstly the relationship between the inputs and the outputs and their weights are founded by relativity analysis. Then the conversion values of the inputs are calculated with the method of point pattern matching. Finally, the forecasting results are obtained. The proposed approach is tested with an actual system, and the results show a rather accurate forecasting.

关 键 词：日负荷预测 点模式匹配 相关系数 多模式 电力系统 

分 类 号：TM715[电气工程—电力系统及自动化]