基于MATLAB统计学工具箱的电力负荷中期预测Middle Term Load Forecasting Based on MATLAB Statistics Toolbox
林国庆,黄民翔
LIN Guo-qing,HUANG Ming-xiang
摘要(Abstract)：
对电力负荷中期预测中常用的增长趋势预测法和回归模型预测法的优缺点进行分析,在此基础上提出一种电力负荷中期预测的新方法。该方法巧妙运用MATLAB统计学工具箱中的现有函数,并借助可决系数和预测相对标准偏差两个定量判别指标来选择预测模型,从而能更有效地对电力负荷进行中期预测。该方法在理论上兼有常用的增长趋势预测法和回归模型预测法的优点,在实践中又能很好地克服上述两种方法的缺点,且编程简单,预测结果精确度高,具有很强的实用性。最后以浙江省全社会用电量的中期预测为例,说明该方法的有效性。
A new method is developed by analysing the merits and drawbacks of two usual middle term load forecasting methods based on growing trend and regression analysis model.Using the coefficient of determination and the relative standard deviation as standard,this method can figure out a more effective forecasting model with MATLAB Statistics Toolbox.This method not only involves the merits of growing trend forecasting method and regression analysis model forecasting method in theory,but also avoids the drawbacks of the two methods in application.It is simple in programming and,has high forecasting accuracy and good practicability.The forecasting result of Zhejiang Province's electric power demand indicates this method's effectiveness.
关键词(KeyWords)：
MATLAB;负荷预测;模型;拟合精度;预测精度
MATLAB;load forecasting;model;fitting accuracy;forecasting accuracy
基金项目(Foundation):
作者(Author):
林国庆,黄民翔
LIN Guo-qing,HUANG Ming-xiang
DOI: 10.19585/j.zjdl.2005.06.001