摘要
根据交通量形成的原因,分析了交通量和相关影响因素之间的关系以及交通量预测的特点,建立了交通量预测的广义回归神经网络(GRNN)模型。以某一公路交通吸引区1985-1995年的交通量和相关经济指标的历史统计数据作为学习样本,通过拟合训练和外推预测分析,验证了GRNN用于交通量预测的有效性。
According to the formation of a traffic volume, the relationship between traffic volume and relative factors and the characteristics of traffic forecast are analyzed. A general regression neural network (GRNN) model for traffic volume forecast is established. Adapting and extrapolation forecast are made with specimens out of the historical statistical date of the traffic volume and other economic indices of a typical section of a highway from 1985 to 1995. The effectiveness of using GRNN to foreeast traffic volume is demonstrated.
出处
《长沙交通学院学报》
2006年第2期46-50,共5页
Journal of Changsha Communications University