刘蔚鹏
随着城市机动车保有量的快速增长，城市道路交通基础设施建设难以跟上机动车的增长速度，交通供需矛盾日益加剧。在高峰期，城市路网普遍处于饱和甚至过饱和的状态，区域性交通拥堵频发，被动地调节供给已经难以解决问题，主动控制交通需求成为了更好的应对方案。路网宏观基本图揭示了路网饱和程度与路网运行效率之间的普遍规律，单峰型抛物线的曲线为区域交通总量控制找到了控制目标，通过截流和卸载，也即限制区域的驶入速率、提高区域的驶出速率，能够有效控制区域的车辆总数，提升路网的整体运行水平。本文从车辆轨迹宏、微观分析出发，以边界交叉口作为重点研究对象，研究控制通道选择及信号配时调整的方法，进而提出针对城市中心区的边界控制策略。主要研究内容包括以下几个方面：
（1）以青岛市南区部分区域为例，通过轨迹数据的处理、分析，计算相关宏观指标，分析区域的高峰、拥挤特征，结合宏观基本图的相关理论，进而提出通过轨迹数据获取宏观基本图的方法，并基于宏观基本图进行区域交通流状态划分，为后续研究打下基础。
（2）结合冲击波理论与边界通道的轨迹特性，提出边界受控路段选择模型，并采用实际案例与其他已有模型进行对比分析。通过车流守恒方程推导分析了边界控制的影响机理，在阀门控制模型基础上，提出边界交叉口信号配时调整方法，明确了配时参数调整、相位补偿及信号协调的模型及方案。为后续边界控制策略的制定提供依据。
（3）提出针对城市中心区的边界控制策略。基于反馈阀门控制的理念，构建结合反馈控制模型、驶入流量限制及分配模型、绿信比调整模型的边界控制模型。。该控制模型以排队长度及最长、最短绿灯时间为约束条件，以维持区域稳定流交通状态作为主要控制目标。
（4）基于实际路网建立仿真模型，验证策略有效性。使用VISSIM软件建立仿真路网，以青岛市南区的实际路网、实际交通状况进行模型标定，获取初始宏观基本图并实施控制策略。选取路网内部车辆数、平均车速、平均延误、平均停车次数作为评价指标，对于实施控制与未实施控制的情形进行对比，区域整体运行效率在控制后得到了提高，验证了该控制策略的有效性。
关键词：交通总量控制、宏观基本图、边界控制、反馈阀门控制、信号配
Abstract
As the urban vehicle inventory increasing rapidly, the urban road traffic infrastructure construction is difficult to keep up with the growth rate of vehicle, therefore worsen the imbalance between supply and demand of traffic. During the peak hours, the urban network falls into saturated or even over-saturated condition, frequently causing regional traffic congestion problem. Controlling demand initiatively rather than adjusting supply passively becomes a better solution to the problem. Network macroscopic fundamental diagram(MFD) reveals the general rule between network saturation degree and its operating efficiency, and unimodal type parabola finds the controlling target for regional traffic total control. By the method of cut-off and discharge, in other words, restricting the in-coming rate and increasing the leaving rate, the number of vehicles in the region can be controlled, therefore promoting the operation efficiency of network. This paper starts at the analyze of vehicle trajectory in macroscopic or microscopic aspect and regards boundary intersection as important research object. After researching the method for selecting controlling passage and adjusting signal timing plan, the perimeter control strategy is put forward, aiming for relieving congestion in the urban center. Main research contents include the following aspects:
(1) Take the South District of Qingdao City as an example. By handling and analyzing the trajectory data, the peak feature and congestion feature are analyzed. The method for acquiring MFD via trajectory data is put forward by combing existing theory of MFD. After that, division of regional traffic flow state is finished based on MFD, providing ground for follow-up study.
(2) Selection model for boundary cotrolled link is raised based on trajectory feature analysis at boundary passage, and it’s compared with other existing models via real case. With the help of the conservation equation of traffic flow, the impact mechanism of boundary control is analyzed. Based on the valve control model, the signal timing plan adjustment method is put forward, providing basis for the boundary control strategy followed by.
(3) Put forward boundary control strategy for city center area. Based on the concept of feedback valve control, the boundary control model is built, which includes feedback control model, in-coming flow restriction and allocation model and split adjustment model. This control model is under the constraint of queuing length and maximum or minimum green time, aiming for maintaining the steady traffic state.
(4) Verify the validity of strategy by construct simulation model based on real network. Use Vissim to build simulation model and calibrate it with the real network and real traffic state of the South District of Qingdao City. The initial MFD is acquired and the strategy is put into effect. Four index including network internal vehicle number, average vehicle speed, average delay and average stops are choosed for performance assessment, and the uncontrolled and controlled situation are compared. The overall operation performance is improved by implement the strategy, therefore verifying the effect of controlling strategy.
Key Words：Total Traffic control, macroscopic fundamental diagram(MFD), boundary control, feedback vavle control, signal timing plan
