城市交通网络信号控制系统的实时演算模型

Modeling Urban Traffic Control Systems from the Perspective of Real Time Calculus

基于实时演算(real-time calculus,简称RTC)理论,为单/双行道两类城市交通网络的定时和自适应两类信号控制系统建立了统一的形式化模型.首先,将车流和交叉路口分别建模为RTC的到达曲线和资源曲线;然后,根据不同信号控制策略,将紧邻路口间的曲线进行综合计算,得到整个交通网络的RTC模型.应用最小加代数方法,RTC模型能够计算车辆在路口的最长等待时间D和路口拥堵车队的最大长度B.基于RTC模型,应用MATLAB对8组不同规模的城市交通网格进行仿真,实验结果表明:(1)与双行道网络相比,单行道网络更能有效处理较稀疏的交通流.以定时控制为例,在车流频率u≤1/2时,单行道网络能够将交通拥堵指标D和B分别降低至少2.66倍和3倍;(2)双行道网络中,车流频率u存在一个临界区域,在临界域内,拥堵指标随车流频率递增变化,一旦u低于或超出临界域,拥堵指标则分别保持稳定不变或不可控;(3)自适应策略优于定时控制策略,例如在双行道网络中,自适应控制策略对应的拥塞指标D和B比定时控制策略分别降低1.68倍和1.26倍.

This work presents a new framework for urban traffic flow control based on the real time calculus（RTC） method. The queuing behavior of the traffic flow is transformed into an arrival curve, and the capacity of the intersection is characterized by a service curve. According to different signal control strategies, the service and arrival curves at an intersection are used to calculate the outgoing arrival curve. This result curve at each intersection is further integrated with the curves at the adjacent intersections, which finally exhibits the RTC model of the whole traffic network. The presented model can evaluate the bounds of the delay D of a vehicle and the backlog B of an intersection. The experiments are settled on the urban girds, and reveal the changing trend of the congestion factors D and B that are under fixed-time and adapted control strategies respectively. This is followed by a discussion of how this modeling method helps to estimate the effect of different signal control strategies.