基于动态广义回归神经网络煤炭需求预测模型

煤炭需求预测是指导我国煤炭工业发展规划的重要依据之一。我国煤炭资源需求影响因素复杂,准确地预测需要考虑经济、社会等多方面因素。针对获取数据不足等特点,首次在煤炭的预测中运用动态广义回归神经网络(D-GRNN)的方法,预测了中国未来近十年内的煤炭预测值,得到了比较合理的预测值。运用预测数据对国家的煤炭宏观调控做出了合理评价。

Forecasting Model of Coal Demand Based on Dynamic D-GRNN

Forecasting of coal demand is one of the key basis to guide the development plan of China’s coal industry.Economic,social and other factors should been comprehensive considered when predict precisely due to the complexity of coal command in China.On account of the lack of the related data,this paper applies Dynamic GRNN(D-GRNN) to forecast the coal demand in nearly ten years firstly and get the quite reasonable result.Finally,we evaluated the macro-control of coal industry based on the forecasting data.