陕西省能源消费的组合预测

组合预测理论及建模技术对于信息不完备的复杂经济系统具有一定的实用性。鉴于能源消费系统的复杂性及非线性的特征,文章首先利用陕西能源消费量的历史数据,分别采用指数回归模型、能源需求弹性回归模型及灰色模型建立了陕西省能源消费系统的单项预测模型。其次,采用标准差法进行非负权重分配,建立了陕西省能源消费量的组合预测模型。结果表明,加入时间虚拟变量和分段建模的预测精度明显提高,且组合预测模型的精度高于单项预测模型。最后,应用该模型对陕西未来10年的能源消费量进行了预测。

Combinatory Forecasting on Energy Consumption of Shaanxi Province

Theory of combination forecasting and skill of modeling are practicable in complex economic system with uncompleted information.Because energy consumption system is of complexity and non-linearity,the exponential regressing model,energy demand elasticity regressing model and grey forecasting model were used to construct the models for Shaanxi energy consumption depending on historical data.Combination forecasting model of energy consumption was proposed by using standard variance to allocate non-negative weights.Results showed that the forecast precision improved greatly by the model including time dummy variables and divisional mathematical modeling,the precision of combination forecasting model is better than the single one.The model was also used to forecast Shaanxi energy consumption in the following 10years.