一种分布式交通信号控制方法及仿真实现

在基于多智能体的分布式道路交通控制概念模型和单路口交通信号自学习控制方法的基础上，提出了一种基于多智能体的分布式交通信号区域控制方法。通过相邻路口信号控制智能体的信息交互和协调，在确保路口绿灯时间利用率较高的前提下，尽量使相邻路口驶来的车队不停车地通过路口。编制交通控制微观仿真软件，在一个由8个路口组成的交通网络中对多种信号控制方式进行仿真实验，实验结果表明这种新控制方法的控制效果明显优于传统的定时控制和感应控制方式。

A Distributed Traffic Signal Control Approach and Simulation

On the basis of distributed multi-agent-based traffic control conceptual model and the self-learning traffic signal control approach, a distributed multi-agent-based regional traffic signal control method is proposed. Through exchanging messages and cooperating between signal control agents in adjacent intersections, the vehicle fleet from adjacent junction passed the local intersection without stop as fast as possible while the utility efficiency of the time of green light in the local junction was at relatively high level. After programming a microscopic simulation program of traffic control, several different signal control methods in a traffic network was simulated consisting of eight junctions. The result of simulating illustrates that the effect of the new signal control approach is obviously better than that of the traditional fixed-time control approach and actuated control approach.