基于误差校正的中长期负荷预测模型  

A Model for Medium-and Long-Term Power Load Forecasting Based on Error Correction

作　　者：刘达[1] 

机构地区：[1]华北电力大学经济与管理学院,北京市昌平区102206

出　　处：《电网技术》2012年第8期243-247,共5页Power System Technology

基　　金：国家自然科学基金项目(70901025);中央高校基本科研业务费专项资金资助(09MR39)~~

摘　　要：中国的中长期负荷呈现明显的增长趋势,大部分插值算法在预测建模时,对区间外预测结果的有效性得不到保证。文章建立了一个综合时间序列建模和回归建模优点的模型来预测湖南衡阳地区年度负荷。将总量数据转换成增速数,然后建立广义自回归条件异方差(general autoregressive conditional heteroscedasticity,GARCH)对负荷增速序列建模。建立回归模型分析GARCH模型的残差中未被GARCH模型解释的外界影响,然后根据回归预测的残差对GARCH模型误差进行校正。在选择回归模型变量时,引入格兰杰因果检验筛选适当的影响因素,引入主成分分析提取影响因素中包含的信息,降低自变量的维数,提高中长期负荷建模精度。实例研究表明该方法对于中国中长期负荷预测较为准确。The medium- and long-term power loads in China are evidently increasing. When interpolation algorithms are utilized to establish power load forecasting model, it is hard for the majority of them to ensure the effectiveness of the forecasting beyond the forecasted terms. In this paper, synthesizing the time series modeling with the regression modeling, a model to forecast annual power load in Hengyang region, Hunan province is established. The data of total amount is turned into growth data; then a general autoregressive conditional heteroscedasticity （GARCH） model for load growth series is built. A regression model is established to analyze the external influences in the residual error of the GARCH model, which is not explained by the GARCH model; then according to the residual error from regression forecasting the error of GARCH model is corrected. During the selection of variables for regression model the Granger causality test is introduced in to select the proper impacting factors for residual regress model, and the principal component analysis is introduced in to extract the information contained in impacting factors and to reduce the dimensionality of independent variables to improve the accuracy of medium- and long-term load forecasting. Results of case study show that the proposed method are more accurate for forecasting the medium- and long-term load in China.

关 键 词：负荷预测 误差校正 中长期负荷 广义自回归条件异方差 

分 类 号：TM714[电气工程—电力系统及自动化]