摘要
城市道路交通状态判别是动态导航系统中的关键技术之一,文章从城市道路交通流系统的高度复杂性特点出发,提出了一种基于神经网络的城市道路交通状态判别方法.首先,利用灰关联熵分析方法选取交通状态的关键性特征指标;然后,建立交通状态判别的神经网络模型并利用实测数据对其进行离线训练;最后,应用训练后的神经网络进行城市道路交通状态在线判别.实验表明,用于城市快速路的交通流状态判别方法效果良好。
Nonlinear traffic flow of urban road is one of the key technologies of dynamic navigation system. From the point of highly nonlinear traffic flow of urban road, an identification method of urban road traffic state based on neural network was discussed. The grey relation entropy method was used to select the key characteristic indexes of traffic state. Then, a neural network was established as traffic state identification model and trained by measured data off-line. Finally, the trained neural network was used to identify the urban road traffic state on-line. Moreover, a simulation experiment proves the good performance of the proposed method in case of urban freeway.
出处
《交通信息与安全》
2009年第2期73-76,共4页
Journal of Transport Information and Safety
关键词
交通状态
动态导航
灰关联熵
BP神经网络
traffic state
dynamic navigation
grey relation entropy
back propagation neural network(BPNN)