一种基于分布式网络多智能体的匝道协同控制方法

摘要

In order to improve the growing serious traffic congestion on the junction of freeway and expressway (junction of road networks ), considering the capacity and conversion of heterogeneous road network, on basis of multi-agent and local induction control, the control mode of multi-ramp metering is discussed. Then a ramp CPSO-RBF-PID control model, a consensus coordination model of ramps on the same road, and a competitive coordination model of ramps on different road are designed. With the topology of the junction of road network, a multi-agent ramp control model based on distribution network is presented. The appropriate network is selected for the model validation. The results show that the model is effective. The method can stabilize the density of the junction of road network, and relief traffic congestion in the area.%针对高速公路与城市快速路交汇区域(路网结合部)拥堵日益严重的现状,综合考虑异构交通网络的通行能力与转换需求,基于多智能体和局部感应控制思想,对路网结合部匝道控制模式进行分析,建立单点CPSO-RBF-PID匝道控制模型、关联主线多匝道一致性协同方法和不同主线间匝道竞争性协同方法.在此基础上,将路网结合部进行网络拓扑,建立基于分布式网络多智能体的匝道协同控制方法.并选取相应路网结合部对所建模型进行验证,结果证明,该模型具有良好的控制效果,能够起到有效地稳定路网结合部交通流密度,缓解结合部区域交通拥堵的作用.