城市交通信号的迭代学习控制及其对路网宏观基本图的影响

发布时间：2019-05-20 05:11

【摘要】：针对在城市交通信号控制中存在对交通流难以精确建模的问题,首先利用交通流的重复性特点,提出了一种基于迭代学习的城市交通信号控制方法,并证明了在不确定初态下迭代学习控制算法的收敛性.其次,结合路网宏观基本图的特性分析了基于迭代学习的交通信号控制策略对路网交通态势的影响.结果表明,当迭代的初始状态在期望初态值的小范围内波动时,系统的跟踪误差仍能收敛到一个界内;通过对交通信号的迭代学习控制,路段的实际占有率能够逐步逼近期望占有率,从而使路网内的车辆密度分布更加均匀,确保交通流在更优的宏观基本图下运行,防止因车辆密度分布不均引起的通行效率下降及交通拥堵的发生.最后,通过仿真实验对所提方法的有效性进行了验证.
[Abstract]:In order to solve the problem that it is difficult to model traffic flow accurately in urban traffic signal control, an urban traffic signal control method based on iterative learning is proposed by using the repetitive characteristics of traffic flow. The convergence of iterative learning control algorithm in uncertain initial state is proved. Secondly, combined with the characteristics of the macroscopic basic diagram of the road network, the influence of the traffic signal control strategy based on iterative learning on the traffic situation of the road network is analyzed. The results show that when the initial state of the iteration fluctuates in a small range of the expected initial state, the tracking error of the system can still converge to a boundary. Through the iterative learning control of the traffic signal, the actual occupation rate of the road section can gradually approach the expected occupancy rate, so that the vehicle density distribution in the road network is more uniform and the traffic flow runs under the better macro basic map. Prevent traffic efficiency decline and traffic congestion caused by uneven distribution of vehicle density. Finally, the effectiveness of the proposed method is verified by simulation experiments.
【作者单位】：西北工业大学自动化学院;
【基金】：国家自然科学基金重点项目(61134004)资助~~
【分类号】：U491.54

【参考文献】