将车辆起讫点引入四向 BML(Biham-Middleton-Levine) 模型, 在周期边界下研究城市交通流的动力学演化过程. 驾驶人选择到终点的最短路径时不仅要考虑相对位置, 还要考虑道路排列的行 (列) 的奇偶性. 采用静态路径选择策略时, 由于路径在初始时刻已经确定, 在很低密度下就发生局部死锁, 并不能有效提高临界密度. 采用基于全局交通信息的动态路径选择策略, 驾驶人根据速度阈值, 在满足转向条件时可以改变运动方向来避开拥堵区域. 模拟结果表明, 采用较小的阈值就可以有效减少交通拥堵, 且不会显著增加行程距离. 如果采取分区的局部控制策略, 还可以进一步减小平均速度的脉动.

The origin-destination pair of vehicles was introduced into the Biham-MiddletonLevine (BML) model with four-directional traffic to investigate the dynamical evolution process of urban traffic flow under a periodic boundary. The driver selects the shortest path to their destination considering not only the relative position but also the parity of rows (columns). When a static route choice strategy is used, local gridlocks occur at very low density because the path is already determined at the initial moment, and it is ineffective to improve the critical density. Using the dynamic route choice strategy based on global traffic information, the driver can change the direction of motion to avoid congested areas when the steering condition is satisfied according to the speed threshold. Simulation results show that using a smaller threshold value can effectively reduce traffic congestion without significantly increasing travel distance. The fluctuations of mean speed can be further reduced if a local control strategy in a subdomain is adopted.