交通干线递阶控制策略及其仿真研究

Traffic Trunk Lines Hierarchical Control Strategy and Its Simulation

交通流量信息在不同的时间段内存在很强的非线性.传统的控制算法中,一旦车流在瞬间增大,各种调控指标会随之发生非线性变化,造成不收敛,很难形成均匀平衡,造成调度效率下降。在分析城市干线交通控制研究现状的基础上,提出一种无线传感器和模糊神经网络的干线多路口交通信号递阶控制策略。采用两层WSN结构,收集单个路口车辆信息,经数据融合后传递给第二层,并采用模糊神经控制方法在线调整路口各方向的绿信比,接收第一层检测的各交叉路口的交通数据,确定干线上的信号周期,完成调控。仿真结果表明,改进方法真实地反映了交叉口的实时信号控制过程,证明采用平均等待车辆长度作为性能评价指标,控制效果优于普通模糊控制。

Based on the analysis of the current research situation of the urban trunk line traffic control, the trunk line multi- intersection traffic signal hierarchical control strategy was proposed based on wireless sensor and fuzzy neural network. The method was used to adopt two - layer structure of the WSN and collect the traffic information of single intersection. After the data fusion, the information was then transmitted to the second layer, and the ratio of green signal in respective direction of the intersection was adjusted online using the fuzzy neural control method. The intersection traffic data of the first layer were received, and signal period on the trunk was determined to complete the regulation. Simulation results show that the improved method truly reflect the real - time signal control process of the intersection, and it is proved that taking the average length of the waiting vehicles as a performance evaluation index, the control effect is better than the ordinary fuzzy control.