交通分配与信号控制组合优化研究综述  

Review of research on combined traffic assignment and signal control problem

作　　者：彭显玥 王昊[1,2,3] PENG Xian-yue;WANG Hao(School of Transportation,Southeast University,Nanjing 211189,China;Jiangsu Key Laboratory of Urban ITS,Southeast University,Nanjing 211189,China;Jiangsu Province Collaborative Innovation Center of Modern Urban Traffic Technologies,Southeast University,Nanjing 211189,China)

机构地区：[1]东南大学,交通学院,南京211189 [2]东南大学,城市智能交通江苏省重点实验室,南京211189 [3]东南大学,现代城市交通技术江苏高校协同创新中心,南京211189

出　　处：《交通运输工程与信息学报》2023年第1期1-18,共18页Journal of Transportation Engineering and Information

基　　金：国家重点研发计划项目:城市多模式交通网运行仿真系统平台开发(2019YFB1600200);江苏省科技计划项目(BE2020110)。

摘　　要：众多研究表明交通信号控制与车辆路径选择相互影响,信号配时影响车辆的路径选择行为,进而影响路网的交通流量分布,故需要根据流量情况设置信号配时方案以提升路网整体性能。本文聚焦交通分配与信号控制的组合优化研究,分析了交通分配和信号控制的关联机制以及应用于组合优化问题的经典原理和策略;归纳了组合优化问题的研究要素,并基于博弈论将组合优化问题分为迭代优化问题(纳什古诺博弈)、系统优化问题(垄断博弈)和混合优化问题(主从博弈)。迭代优化策略将交通控制与分配看作同一层次上两个互相关联的问题,迭代求解这两个问题,以期获得信号配时和交通流量分布相互一致的解决方案;系统优化策略是一种系统最优控制,交通管理者同时优化控制方案和交通流量;混合优化策略在考虑用户路径选择的前提下,以控制参数为决策变量,构建并求解组合优化模型。对三类优化问题的相关研究进展进行了归纳和综述,并分析其特征及优缺点。建议未来从以下方向进行优化:将传统的组合优化问题扩展至智能网联环境下信号控制与路径诱导的协同控制;考虑控制策略对于OD需求的影响,并扩展出行行为因素,将出行模式、离开时间等参数纳入决策变量;可依据路径选择倾向、旅行成本感知情况等标准进行车辆分类,提高控制策略的鲁棒性;采用合适的城市路网交通模型,更加详细地描述微观交通现象,精确地计算交叉口及路段延误;开发和优化求解组合问题的算法,提升算法的计算效率、适用规模以及理论特性(例如最优性、收敛性),采用并行计算、多线程计算等手段提升计算效率;最后,在交通描述精确性、信号控制方案灵活性和全面性以及算法求解效率等多个要素之间进行平衡。Numerous studies have shown that traffic signal control and drivers’ route behavior are two processes that influence each other.Signal settings impact vehicles’ routes,thus affecting the network traffic flow pattern.Signal timings are optimized according to the flow pattern to improve the system performance.By focusing on the combined traffic assignment and signal control problem,the interdependency of the two parts and classical principles and strategies applied to the combined problem was analyzed in this study.The research elements of the combined problem are summarized,and the combined problem is divided into three types based on the game theory:the iterative optimization and assignment(Nash-Cournot game),system optimization problem(Monopoly game),and hybrid optimization problem(Stackelberg game).In the iterative optimization and assignment,traffic control and assignment are solved iteratively on the same level to obtain the mutually consistent flow pattern and signal timings.The system optimization problem is a system optimal control where the traffic manager simultaneously optimizes the signal timings and traffic patterns.By considering the users’ route choice,the hybrid optimization problem takes signal timings as decision variables and formulates and solves the combined problem.Research on the three combined problems is summarized and reviewed,and their characteristics,advantages,and disadvantages are analyzed.The following directions are recommended for optimization in the future.The traditional combined problem can be extended to joint signal control and route guidance in a connected and autonomous environment.The impact of signal settings on origin-destination demand should be considered,and the travel behavior factors can be extended by adding the travel mode and departure time as decision variables.Vehicles can be classified according to criteria,such as route choice preference and travel cost perception,and the robustness of the control strategy can be improved.In addition,more appropriate urb

关 键 词：交通工程 交通分配与信号控制组合优化 综述 城市路网 

分 类 号：U491.4[交通运输工程—交通运输规划与管理]