城市交通网元胞自动机全局流量预测方法  

Prediction method of global transportation network flow with cellular automata

作　　者：张凤琴[1] 王梦非[1] 管桦[1] 李小青[1] 陈大武[2] 陈桂茸[1] 

机构地区：[1]空军工程大学信息与导航学院,西安710077 [2]空军工程大学图书馆,西安710077

出　　处：《计算机应用研究》2017年第1期42-45,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(71503260);陕西省自然科学基金资助项目(2014JM8345)

摘　　要：面对城市交通日益拥堵,针对交通管理中全局流量预测难的问题,通过将路口的各个方向的车流以不同的边表示,将双向道路的两侧分别拆分为两个端点,提出了精准描述进入和离开交通道路各方向负载的方法。以节点作为元胞自动机的元胞,以边关系作为元胞的邻域,利用阻塞—转移方法建立元胞自动机的局部规则,并采用递归算法计算每一时刻交通网中各个节点元胞上的交通流量,构建城市交通网络流量预测模型。最后以某市交通主干道的实测流量数据为例,验证了该预测模型的准确性和高效性,从而证实了该模型对于城市的交通管理能力有一定的提升作用。In order to solve the issues of traffic managment in the global flow forecast, the paper proposed a new approach to precisely describe the bidirectional traffic load based on vehicle entering and leaving, which driving to different direction at an intersection can be represented as different sides and both ends of a bidirectional road can be represented as two endpoints in a graph. Further, the paper built new local rules based on Cellular Automata that nodes were unit cells and edges between those unit cells were their neighborhoods to improve the blocking-transfer method. Then, an urban traffic network prediction model was proposed by a recursive algorithm to caclulate the traffic flow on each Cellular node. Finally, the actual flow data measured from a city traffic trunk was used to verify the accuracy and efficiency of the prediction model. The result of simulation experiments show that the model can significantly promote the capabilities of urban transport management.

关 键 词：城市交通网 元胞自动机 流量预测 阻塞—转移 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] TP301.6[自动化与计算机技术—控制科学与工程]