短期电力负荷的灰色GM(1,1)模型群组合预测

组合预测方法是一种性能优越的预测方法.由于电力负荷具有很多不确定因素,用单一预测模型进行预测时,其预测精度不高,为提高预测精度,提出组合预测模型.灰色GM(1,1)模型群能够很好地反映电力负荷的周期变化特性,而用不同时期的历史数据可反映不同的信息特征,因此,提出了基于远近数据的GM(1,1)模型群预测组合预测法.利用此方法对某地区的日电力负荷进行预测的算例结果表明:此方法的预测精度高于各单一模型的预测精度,且能够很好地反映日负荷变化的随机性和周期性.

Combination Forecasting Based on Gray GM (1,1) Model Groups for Short-term Electric Power Load

Combination forecasting method is a superior performance forecasting method.Because the power load has many uncertain factors,the prediction accuracy is not high when power load is forecasted by use of single forecast model.For this reason,combination forecasting method is put forward in order to improve the prediction accuracy.Gray GM(1,1) model groups can reflect the cycle characteristics of power load change effectively and different historical data provide different information.Therefore,combination forecasting based on gray GM(1,1) model groups is put forward in this paper.Power load in a certain area is forecasted by using this method.The results show that the prediction accuracy of this method is higher than the prediction accuracy of single forecast model and it can reflect the randomness and periodicity of power load change effectively.