【摘要】 为了利用交通流固有的周期性特征改善城市路网的交通状况,提出了一种基于迭代学习的城市道路交叉口信号控制方法。通过对交叉口信号的迭代学习控制,使其能够适应不断变化的交通状况,从而提高交通流在路网中的运行效率;考虑实际交通系统会受到各种外部因素的干扰,分析了当系统状态和输出存在扰动时迭代学习控制律的鲁棒收敛性。研究结果表明:有界的状态扰动和输出扰动使得系统的跟踪误差收敛到一个界内,且该界仅由系统不确定的外界扰动决定;当扰动为0时,系统输出能够完全跟踪期望输出;仿真分析验证了所提出方法的有效性,该方法可以避开交通流复杂的建模和辨识过程,直接通过迭代控制律计算交叉口各相位的绿灯时间,计算量小,设计简单,易于实现。更多还原

【Abstract】 In order to improve the traffic condition of urban road network by using the inherent periodic characteristics of traffic flow,an iterative learning based signal control method for urban road intersections was proposed. The intersection signals controlled by iterative learning approach can adapt to the changing traffic conditions and improve the efficiency of traffic flow running in the network.Considering the existence of various external disturbances in real traffic system,the robust convergence of the iterative learning control law was analyzed when system state and output were under the disturbance.The results show that the bounded state and output disturbances make the system tracking error convergence in a given bound,which only depends on the uncertain external disturbances of the system.The system outputs will completely track the desired outputs when there are no disturbances.The effectiveness of the proposed method is verified by simulation analysis.The proposed method can avoid the complex modeling and identification process of traffic flow and directly calculate the green times of each phase at intersections by the iterative control law.It is easy to implement with less computation and simple design.更多还原