实时交通信息下基于抽象路网结构的交通均衡

Traffic Equilibrium with Real-Time Traffic Information Based on Abstract Road Networks

由于现实世界的复杂性和人为因素的不可控性,交通系统的最优交通配流难以实现.多数交通系统运行在用户最优原则下.通过基于多智能体交通微观仿真技术在3种不同抽象路网中的仿真研究,得出如下结论:不论在交通系统中是否植入实时交通信息系统(RTTIS),系统都能够建立近似交通用户均衡.在这些均衡状态中,在平行路网中RTTIS的植入,恶化了交通系统的运行,车辆平均速度降低50%;在网格状和环状路网中其植入却使交通效率得到了优化,车辆平均速度分别提高33%和71%.RTTIS能够将交通系统从某一用户均衡状态转换到新的用户均衡状态,在这个过程中系统效率可能得到提高也可能变坏,这依赖不同的交通网络结构.

Considering the cost and complexity of realizing SO（system optimum） principle in traffic assignment,most traffic systems with and without real-time traffic information system（RTTIS） run in UO（User Optimum） principle.It is necessary to examine whether UE（user equilibrium） can be built and used to improve the traffic efficiency in the systems with and without RTTIS.The results of our experiment indicate that approximate UE states can be established in three types of networks with and without RTTIS,that in these UE states the traffic efficiency is improved in grid networks（the average speed is increased by 33%） and ring networks（the average speed is increased by 71%） with RTTIS,and that in parallel networks the new UE achieved by RTTIS decreases traffic efficiency（the average speed declines by about 50%）.So the UE situations should not be taken as an indicator of traffic system efficiency.