基于径向基函数神经网络的交通量预测

采用了径向基函数神经网络进行未来年的交通量预测,它具有收敛速度快、唯一最佳逼近且无局部极小等优点。在进行交通量预测时,选取公路里程、汽车保有量、国民生产总值、国民收入和人口作为交通量影响因子,通过2个网络模型,在输出层输出该年的预测交通量。

Traffic Volume Forecast Based on Radial Basis Function Nerve-Network

The traffic volumes of the future years are forcasted by radial basis function nerve-network,whose merits include rapid constringency,exclusive best approach and non-partial minimum and so on.When traffic volumes are forecasted,the highway mileage.automobile tenure quantities,GNP,national income and population are taken as affection factors of traffic volume.Through two network models,the traffic volume forecasts of the year on the output layer are exported by computing the network connected values.