天津职业技术师范大学学报

以某市公路货运量的历史数据为例,分别采用灰色模型和径向基神经网络模型对公路货运量进行预测。在此基础上,运用有效度理论,构建并联型灰色神经网络模型,并对两模型的预测结果进行加权作为最终预测值,计算结果表明了该方法在公路货运量预测中的有效性。文中还应用马尔可夫链预测模型对组合模型的预测结果进行了分析,增加了预测结果的可信度。

Taking the historical data of one city highway freight volume as an example,by using the radialgray model and neural network to predict the model freight volume on the highway,based on this,by using validity theory,parallel grey neural network model is constructed and the result of the two models can be used as the final result.The test results show that the prediction on the basis of the combined forecast methods is effective.At last,Marqov chain prediction model is used to verify the result.