基于信息熵与计量经济学模型的饱和负荷预测方法

The Saturated Load Forecasting Method Based on Information Entropy and Econometric Model

对于长期的负荷预测来说,采用单一的预测模型进行远景年饱和预测时显然是不够的,为了充分利用每种单一预测模型的优点,提出基于信息熵与计量经济学模型的饱和负荷预测方法。首先选用计量经济学模型分别预测了全社会用电量的饱和值、三产计算用电量的饱和值以及人均用电量的饱和值,并通过上述3种途径分别预测出全社会用电量的饱和值,然后计算上述3种方法预测偏离误差的熵值,并确定每种方法的权重,建立最终的组合预测模型,并用实例分析验证了所提方法的有效性。

For the long term load forecasting,it is obviously not enough to use a single prediction model to make a saturation prediction of a prospective year.In order to make full use of the advantages of each single prediction model,this paper proposes a method of forecasting saturated load based on information entropy and econometric model.First of all,the econometric model has been chosen to predict the saturation value of the total electricity consumption in the whole society,the saturation value of the three production electricity capacity,and the saturation value of the per capita electricity consumption.Through the above three ways,the saturation value of electricity consumption in the whole society has been predicted.Then,the deviation of the error entropy value got from the above three predicting methods has been calculated to determine the weight of each method.Finally,aportfolio forecasting model has been established,and the effectiveness of the proposed method has been analyzed and verified by the practical case.