基于MPC的可变相序交通信号优化控制方法

A Variable Phases Traffic Signal Optimal Control Method Based on MPC

传统基于固定相序机制的道路交叉口信号控制算法的局限性较大,而使用可变相序信号控制方法则具有更多的优化空间。提出了一种基于混合Petri网的可变相序信号控制模型,该模型充分体现了Petri的特点,结构简单紧凑、直观易懂。基于该模型,交叉口交通信号的优化问题可以直接转化为模型中冲突解决策略的选择问题。使用基于模型预测控制(model predictive control,MPC)的算法框架,采用滚动优化策略,基于当前系统状态以及整个信号周期内的预测状态设计优化目标函数,算法具有更好的动态适应性,从而达到交通信号的相序及绿灯通行时间的深度优化。以单交叉口4相位信号控制模型为例,使用基于MPC的可变相序优化控制方法,路口平均排队长度缩短了12%。

The traditional road intersection signal control algorithm based on fixed phase sequence mechanism has great limitations,while there is a larger optimization space for variable-phases optimization algorithm. We propose a variable-phases traffic signal control model based on hybrid Petri nets,which fully embodies the Petri,and is compact and intuitive with simple structure. Based on this model,the optimization of intersection traffic signals can be directly translated into the choice of conflict resolution strategy. We use the algorithm framework based on MPC （model predictive control）,with the rolling optimization strategy,and design the optimization objective function based on the current state of the system and the signal cycle prediction state. The algorithm has better dynamic adaptability,so as to achieve the deep optimization of traffic signal phase order and the green passage time. Taking the 4-phase signal control model of single intersection as an example,the average queue length of intersections is shortened by 12% using MPC-based variable phase order optimization control method.