一类交通信息物理系统的动态路径引导

Dynamic Route Guidance Strategy in Transportation Cyber-physical Systems

针对车辆采取相同路径后可能会导致潜在拥堵的问题,提出了基于实时信息的交通信息物理系统框架,将Q-learning作为动态路径引导策略,针对动态路径引导中的频率和方式展开研究。仿真结果表明,结合实时交通信息的动态引导能有效提升道路通行能力,多次引导能在一定程度上缓解一次性引导中出现的潜在的拥堵,导致的博弈强度因引导方式和频率而异。

Aiming at the potential congestion caused by vehicles taking the same path,a transportation cyber-physical system framework with real-time information is proposed.Q-learning is used as a dynamic route guidance strategy,and the frequency and modes of dynamic route guidance are studied.Simulation results show that the dynamic guidance combined with real-time traffic information can effectively improve the road capacity,and multiple guidance can alleviate the potential congestion in one-shot guidance to a certain extent.The game caused by dynamic guidance varies on the frequency and modes of the guidance.