建立了基于自回归移动平均、灰色预测和BP神经网络三种单项预测方法的最优加权法组合预测模型。通过预测误差评价指标分析,最优加权法组合预测能有效降低预测误差,预测结果表明了预测模型的有效性。
A combined coal price forecasting model was put forward which was based on optimal weighting of Autoregressive Moving Average (ARMA) , grey forecasting, and BP neural network. Error evaluation index tests proved that the combined forecasting model can reduce the prediction error effectively.