非平均权重条件下电力负荷组合预测模型研究

在对电力负荷进行预测过程中,单一的预测模型往往会出现拟合能力低和预测精度不高的缺陷,而且多个预测模型的简单组合也是一种相对粗略的预测方法。本文将非平均权重法运用于电力负荷组合预测模型中,以某省2000~2011年电力负荷为例,结果表明非平均权重的电力负荷组合预测模型的拟合方差比单项预测模型以及平均权重下的组合预测模型都小,而且预测精度更高。

Research on Combined Power Load Forecasting Model under the Condition of Non-average Weight

In the process of power load forecasting,a single prediction model can not improve the fitting capability and forecast accuracy to a high level,so as the simple combination of a plurality of prediction models.In this paper, the non-average weight method is used in combined power load forecasting model. And then take the power load of a province from 2000 to 2011 for example ,the results show that the model this paper proposes not only has a smaller fitting variance than any single forecasting model and the combined model under average weight,but also has a higher forecast accuracy.