基于双层协同架构的高速公路车辆合流决策与规划方法  

Control and Decision-Making Method of Expressway Vehicle Merging Based on Two-Layer Collaborative Architecture

作　　者：姜圣 丁闪闪 彭峰[3] JIANG Sheng;DING Shanshan;PENG Feng(China Academy of Information and Communications Technology,Nanjing 210024,China;China Design Group Co.,Ltd.,Nanjing 210014,China;School of Transportation and Logistics Engineering,Wuhan University of Technology,Wuhan 430070,China)

机构地区：[1]中国信息通信研究院,江苏南京210024 [2]华设设计集团股份有限公司,江苏南京210014 [3]武汉理工大学交通与物流工程学院,湖北武汉430070

出　　处：《大连交通大学学报》2023年第5期8-14,22,共8页Journal of Dalian Jiaotong University

基　　金：江苏省交通运输科技与成果转化项目(KY2019067)。

摘　　要：以自动驾驶车辆群体为研究对象,针对高速公路合流区易产生交通流紊乱这一问题开展研究。首先,提出了公路合流决策与规划的双层协同架构,架构上层采取基于分组的协作式策略,即将车头时距小于设定阈值的车辆组成车辆队列,将其看作一辆特殊的自动驾驶车辆,再进行最优通行次序求解。其次,结合车辆合流实例验证决策的有效性,并利用数值分析对基于枚举法和基于FIFO的通行次序决策方法进行对比分析。架构下层考虑车车通信延迟的多智能体一致性理论,对合流后的多车系统进行间距和速度的一致性控制。实例验证表明,分布式一致性控制算法可以实现车辆队列系统的协同控制。最后,利用真实合流区车辆轨迹数据,利用组队协同控制算法,进行仿真验证试验。试验结果表明,双层协同架构下的决策与规划方法不仅可以实现近似全局最优,同时对提升自动驾驶车辆合流的安全性、提高交通流效率具有显著的作用。Taking autonomous vehicle group as the research object,the problem of traffic flow disorder in highway confluence area is studied,and a two-layer collaborative architecture of highway confluence decision and planning is proposed.The upper layer of the architecture adopts group-based cooperative strategy,that is,the vehicles whose time headway is less than the set threshold are formed into a vehicle queue,which is regarded as a special autonomous vehicle.Then,the optimal traffic order is solved,and the effectiveness of the decision is verified by combining the example of vehicle conjunctions.Numerical analysis is used to compare and analyze the traffic order decision methods based on enumeration method and FIFO.Based on the multi-agent consistency theory of vehicle-vehicle communication delay,the distance and speed consistency control of the multi-vehicle system after convergence is carried out.The example shows that the distributed consistency control algorithm can realize the cooperative control of the vehicle queuing system.Finally,a simulation verification test is carried out by using the real vehicle trajectory data in the merging area and the teamwork cooperative control algorithm.The test results show that the decision-making and planning method under the two-layer cooperative architecture can not only achieve the approximate global optimal,but also play a significant role in improving the safety and efficiency of the autonomous vehicle merging.

关 键 词：公路运输 协同控制 合流区 队列一致性 

分 类 号：U491[交通运输工程—交通运输规划与管理]