基于虚拟变量与时间序列法的电量需求预测  

Forecasting for electric demand based on dummy variables and times series model

作　　者：杨云莹[1] 任玉珑[2] 段锴[1] 

机构地区：[1]重庆市电力公司江北供电局,重庆渝北401147 [2]重庆大学经济与工商管理学院,重庆沙坪坝400030

出　　处：《电力需求侧管理》2011年第5期17-20,共4页Power Demand Side Management

摘　　要：在使用时间序列法预测电量需求的过程中,引入虚拟变量,量化关键不确定性因素对电量需求的影响,构建基于时间序列法和虚拟变量法的预测模型。该模型能够对参与回归的电量数据进行较好的平滑处理,从而更好地拟合电量需求的变化趋势。算例分析表明,使用该模型预测电量需求的准确性较高,在实际工作中具有可行性和有效性。Currently, the uncertainties which affect electric demand such as economic situation,climate anomalies,eet.happen frequently.In view of this fact,the dummy variables is introduced in time series model in this paper, in order to quantify these key uncer- tainties and build a forecasting modei based on the time series and dummy variables.The model can deal with the datum moe smoothly during regression analysis;consequently,it is able to better fit the trend of electric demand.This model is applied in electric demand forescasting,with can improve forecasting precision significantly and prove the feasibility and validity of it in practical work.

关 键 词：虚拟变量 时间序列法 电量需求预测 不规则变动 

分 类 号：TM715[电气工程—电力系统及自动化]