基于最小二乘组合灰色模型的中长期电力负荷预测  

Medium and Long-term Power Load Forecasting Based on Grey Prediction Models Combined with Least Squares Method

作　　者：李泓泽[1] 郭森[1] 王宝[1] 

机构地区：[1]华北电力大学经济与管理学院,北京102206

出　　处：《水电能源科学》2012年第8期187-189,共3页Water Resources and Power

基　　金：北京市哲学社会科学规划基金资助项目(11JGB070)

摘　　要：电力负荷预测是实现电力系统管理现代化的重要内容,有效的中长期电力负荷预测可为电网运行方式、电源建设规划等提供可靠的决策依据。针对中长期电力负荷预测,综合考虑负荷的历史数据和未来可能走势,在对三种灰色预测模型分析的基础上,以其预测值为解释变量,实际负荷值为被解释变量,通过最小二乘法建立多元回归方程,并通过计量经济检验,以此构造了负荷预测组合模型。将该模型应用于北京市年用电量预测中,结果表明所构造模型具有较高的预测精度,有效可行。Power load forecasting is an important part of management modernization of power system. Accurate medium and long-term load forecasting can provide reliable guidance for grid operation and power construction planning. Considering the historical data and future possible trend of power load, on the basis of analysis of three grey prediction models, taking their prediction values as variables and actual power load values as the dependent variable, a multivariate regression equation is established by using least squares method, which meets the need of econometric test. Thus, the combined prediction model is established to forecast annual power consumption of Beijing city. The result shows that the pro- posed model has higher forecasting accuracy, and it is effective and feasible.

关 键 词：中长期负荷预测 预测组合模型 GM(1 1)模型 DGM(1 1)模型 灰色VERHULST模型 最小二乘法 

分 类 号：TM714[电气工程—电力系统及自动化]