灰色BP神经网络在河南煤炭需求预测中的应用

在灰色GM（1，1）模型的基础上，利用BP神经网络修正残差，建立灰色BP神经网络组合预测模型，对河南省煤炭消费总量进行拟合和预测。此组合模型既克服了数据的非线性关系及随机波动大对预测精度的影响，也增强了预测的自适应性。实例结果证明该组合模型具有较好的预测精度，预测结果是可信的。

Application of grey BP neural network in Henan coal demand forecast

In this paper, based on the grey GM （ 1, 1 ） model, using BP neural network to correct residual er- ror to build grey BP neural network combination model in order to simulate and predict the total coal consump- tion, This model not only overcomes the effect of nonlinear relationship between data and random fluctuation on predictive accuracy, but also enhances the adaptive prediction. Finally, the results show that this model has better forecast precision and the forecasting result is reliable.