能源消费需求预测的Elman模型及应用

Elman Model and Its Application of Energy Consumption Forecasting

为提高能源消费需求预测精度,介绍了Elman网络的基本原理,提出了基于Elman网络的能源消费时间序列预测建模方法,以我国1996年至2011年能源消费对Elman网络预测模型训练获得最佳参数后,对2012年至2014年能源消费进行预测,预测相对误差在-0.61%~0.31%之间,结果表明,能源消费需求预测的Elman模型有较高的预测精度,能较好的反映能源消费的非线性特性与规律。

To improve the forecasting accuracy of energy consumption demand,the basic principle of Elman network was introduced,the prediction modeling method of energy consumption of the time series was proposed based on Elman network. Firstly,the best parameters of Elman network forecast model was get after training by the energy consumption from 1996 to 2011 in China,then,the energy consumption was forecast from 2012 to 2014,the relative error of prediction was between-0. 61% ~0. 31%. The results showed that energy consumption demand forecasting of Elman model has higher prediction accuracy, and can better reflect the nonlinear characteristics and law of energy consumption.