【作者】 张敏捷；

【导师】 徐建闽；

【作者基本信息】 华南理工大学 ， 交通信息工程与控制， 2013， 博士

【摘要】 从整体战略角度出发，将路网内所有交叉口作为研究对象，对交叉口与交叉口、交叉口与交叉口群、交叉口群与群之间进行有效的协调控制设计，实现整个控制区域内交通运行效果最优化，解决城市交通容量不足、交通拥堵和交通污染等问题，已成为城市交通控制系统发展的新要求。本文以群体动力学和交通协调控制理论为基础，研究交叉口群体间的关联关系，从双交叉口协调控制、干道协调控制、过饱和路网协调控制的角度探讨和研究群体动力学应用于交通协调控制的原理和方法。完成的主要科研工作与研究成果概括如下：1.对交通协调控制的基本模式进行了探讨，分析和总结了交叉口-交叉口、交叉口-交叉口群协调控制的基本方式；提出了基于群体动力学的交通协调控制理论需要解决的基本问题；最后，通过分析交通协调控制基本问题的特点，提出了基于群体动力学的交通协调控制研究的基本方法体系。2.提出类相位差的概念，通过分析车流的到达模式和延误特性，确定了在双交叉口协调控制中，交叉口所能采取的控制策略集；引入群体演化博弈方法，对每个控制决策局势进行收益分析，计算局势的平衡点和系统收益；从而确定系统博弈的最佳策略。在此基础上，提出完善的协调控制流程和模型方法，并设计蚁群算法进行模型求解。为双交叉口协调控制提供了新的方法，也为群体动力学在交通协调控制领域的深度应用建立了理论基础。3.提出了以宏观传输模型为基本框架的快速交通仿真模型和以交通最优运行为协调控制目标的双层规划模型，从而构建起以快速交通仿真为基础的优化控制方法。最后将该方法应用于虚拟路网的交通协调控制案例，并与Transyt-7F进行了对比，结果表明了该方法的有效性。4.对高交通需求下的干道协调控制策略进行了设计。采用递阶协调的方式，通过动态优化交叉口的周期和绿信比，实现干道主向交通流的理想相位差协调控制、同时对向交通流的优化协调控制策略。建立了交叉口绿信比优化调整的动力学算子，并对其稳定性做出了分析和验证；对主向交通流理想相位差协调方式的实现进行了探讨，建立了相邻交叉口的周期联动公式，给出干道交叉口周期调整算法，同时针对对向交通流引起的交叉口游离问题，结合宏观模型，给出相应的交叉口绿信比微调算法；综上这些，给出了基于群体动力学的干道交通协调控制方法的总流程。在此基础上，建立优化控制模型，并设计有效的求解算法对模型进行求解。最后对该方法的动态性能进行了详细分析。5.以第二章的交叉口协调方式分析为基础，提出了基于群体演化的过饱和交通协调控制模型与方法。方法首先对路网中的关键交叉口进行判别，形成以该关键交叉口（或群）为中心的多级边界交叉口群；依托这些边界交叉口群，对关键交叉口的交通需求进行管理，合理地限制过剩交通流流入关键区域，避免出现过饱和拥堵；由此生成基本的过饱和控制方案，以该方案为基础，对其进一步优化各项参数，最终得到优化的过饱和协调控制方案。通过具体案例将本章方法与TRANSYT-7F优化方法进行对比，证明了该方法的有效性。最后对方法的性能和特征做了深入考察。仿真结果表明，由于最终的协调控制方案由初始方案经过群体演化改造而来，其是否合理对最终方案的协调控制效果影响很大，在生成初始方案时应该考虑到交叉口群的实际交通需求；另外，由于车道组最大允许的饱和度上限关系到边界群的层深度，也对协调效果具有重要的影响。更多还原

【Abstract】 Effective design on coordination control of intersection&intersection, intersection&group of intersections, and group&group from an overall strategy view, is the newestdemand of urban traffic control systems, for achieving the objectives that traffic optimalrunning, dealing with insufficient urban traffic capacity, traffic congestion and pollution. Inthis paper, using theries of group dynamics to research on two intersetions coordinationcontrol, artery coordination control and coordination control at oversaturated conditions arethe core objectives. And the main work and achieves are summarized as follows:1. The basic model of traffic coordination control are discussed and summarized, and soas the methods for intersetion&intersection coordination control, intersetion&group ofintersections. After that, some essential issues are presented for the promising trafficcoordination control theory based on group dynamics. And finally, through the analysis ofthe characteristics of these issues, the basic method systems of the theory are presented.2. The conception of mimetic phase difference is presented. We analysed thecharacterestics of the arrival model and delay model of traffic flows, and confirmed thecontrol strategies set of each intersection in two intersection coordination control. Then themethod of group evolutionary game is introduced. Each situation will be investigated. Thebalance point (the best point) and its benifit in the situations are calculated, and so the beststrategy in the game is confirmed. On this basis, traffic coordination control model andmethod based on group dynamics are suggested, and so does the corresponding solvingalgorithm. This research will be the newest part in traffic coordination control theory, and italso increases the depth of application of group dynamics.3. In order to seek an effective method of traffic coordination control for road networks,the traffic behaviors at intersections are analyzed based on the macroscopic transmissionmodel via a gridding procedure, and a new macroscopic simulation model is constructed tofast evaluate control schemes. Then, by taking the optimal traffic running as the objective, abi-level programming model for traffic coordination control is proposed. Moreover, bycombining the macroscopic simulation model with the proposed bi-level model, a method oftraffic coordination control based on the improved macroscopic traffic model is presented. The method is finally tested on a virtual network with three different traffic demands, and theresults are compared with those of the Transyt-7F method. It is found that the proposedmethod is feasible and is superior to the Transyt-7F method especially in the high demandscenario.4. Artery coordination control strategy is designed for high traffic demand situations.With the method of hierarchical coordination, dynamically adjusting and optimizing cycleand green ratio of intersetions, the main flow on artery will be controlled under ideal phasedifference, and at the same time opposite flow under optimized offset coordination. Groupdynamics is adopted to establish adjusting operator for green ratio. Then the stability of theoperator is verified. The main ideal method of phase difference coordination to traffic floware discussed, and the linkage formula of cycle and offsets of neighbor-intersections areestablished, and so does the adjustment algorithm for the cycle of each intersection. Anotherinching green ratio algorithm is presented, for the congestion dilemma of opposite trafficflow. And based on these, the optimization model is established. The solving algorithm isgiven subsequently. Finally, the dynamic performance of the method is analyzed in detail.5. The coordination model and method for oversaturated network is presented, based ongroup evolutionary and the discussion in Chapter2. In this method, the key intersection(s) inthe network must be firstly identified, and so one or more intersection group(s) can beformed, which are all key-interseciton-centered. These groups can accurately control thetraffic volumes that come into the key area, and so to manage the demand of the key area orintersection. This will be a valid method for avoiding key area congestion. Thus the basicscheme for oversaturated network control is produced. And all to do is to optimize the otherparameters of intersections further. On this basis, the coordination control method isestablished. Some simulations are used to test this method, and the results show theeffectiveness of the model and the method. Finally, some important factors that will affectthe performance of the method are further discussed, and some conclusions are found, thatmaybe effective approaches to promote the performace of the method.更多还原