中文摘要：

以浙江省公路货运量历史数据为例，考虑到影响货运量主要因素，采用灰色神经网络模型GNNM（1，N）进行预测，并与灰色模型GM（1，N）和神经网络预测结果相比较。计算结果表明：该方法在预测公路物流需求量具有有效性；在灰色模型GM（1，N）预测时，通过比较紧邻均值生成序列的生成系数α对预测精度的影响，选取了最优值进行计算从而提高了灰色模型的预测精度。

英文摘要：

Taking the historical data of Zhejiang Province highway freight volume as an example, in view of main influence factors of freight volume, this paper predicts highway logistic demand by the gray neural network model GNNM （ 1, N） and compares the forecast results with ones of grey model and neural network. The test result shows that the prediction on the gray neural network model to predict highway logistic demand is effective. When using the grey model GM （ 1, N） to forecast precision influence of generating coefficient α, it tries to optimize the search for it and consequently improves the forecast precision of grey model.