【作者】 张靖；

【导师】 徐铖铖； 王树盛；

【作者基本信息】 东南大学 ， 交通运输规划与管理（专业学位）， 2022， 硕士

【摘要】 随着我国社会经济的不断发展,大中型城市的出行需求持续增长,快速路的建设成为提升城市通行效率的主要手段。快速路通常与城市道路立体交汇,常见的交汇手段有高架道路和地下隧道两种。其中,高架道路是较为传统的快速路建设方式,但近年来隧道快速路也逐渐被广泛应用,且通常由多个连续隧道组成隧道群形式。隧道群快速路的出现,在为车流提供连续流通行条件的同时,也带来了较为严重的拥堵问题,且多发于交织区路段。为了解决这一问题,需要实施特定的交通控制策略,通过预防交通瓶颈的生成或加速交通瓶颈的消散,达到缓解交通拥堵、提升通行效率的目的。但是,目前对隧道群快速路及其周边路网实施控制的相关研究还较少。因此,面向城市隧道群快速路及其周边路网的拥堵问题,为了提出科学、有效的控制策略和控制技术,本文依次开展了如下研究:首先,为了研究城市隧道群路网拥堵成因,解析通行能力下降关键机理,本文利用无人机群拍摄的典型路网在高峰时段的视频数据,对进出流量和车辆驶出-汇入比例对隧道群交织区通行能力的影响展开定性和定量分析。其中,利用二元Logistic回归建立交织区通行能力下降辨识模型,对多个候选变量及其衍生变量与通行能力下降的相关性进行建模,结果表明:匝道驶入流率、主路驶出流率、主路驶出比例及匝道驶入比例的升高都是造成交织区通行能力下降的显著因素。之后,作为对隧道群路网实施交通控制的基础,本文开展了城市隧道群路网状态感知与态势预测两个模块的研究。二者的功能为利用路网内交通检测器提供的交通流参数,判断路网当前状态、预测路网演化态势。其中,在路网状态感知模块中,本文利用断面速度差算法判定隧道群主路瓶颈位置,利用占有率-排队车辆数关系判定匝道和交叉口进口道的排队长度。路网态势预测模块的功能是基于元胞传输模型实现的,其中后者针对隧道群路网进行了二次开发。对状态感知模块和态势预测模块展开的仿真测试表明,二者均能较为理想地实现预期功能。为了实现在拥堵状态下对城市隧道群路网的调控,本文进行了城市隧道群路网协同控制技术体系的开发。考虑到隧道群路网的道路组织结构,本文采用了三级控制策略协同控制的方式,依次利用了入口匝道、上游交叉口、出口匝道三级缓冲区,对车流进行排队控制。同时,设计了三级策略级联架构,通过设定好的路网状态阈值来控制各级策略在优先级框架内的有序启动。另外,为了解决单个缓冲区域的控制问题,本文又设计了三个子控制器——匝道控制器、信号控制器和路径诱导控制器,分别对各个缓冲区域实施局部控制。为了验证本文提出的城市隧道群路网协同控制方法的有效性,本文基于上述级联式协同控制方法与三类子控制器设计,同时融合路网状态感知与预测模块,在SUMO微观仿真环境内搭建了城市隧道群路网协同控制原型系统。在原型系统中实施的仿真试验表明,在模拟高峰时段的交通需求下,本文所开发的协同控制策略,将拥堵向上游的扩散范围减少了1.3km(-44.83%),拥堵持续时间减少了10min(-16.39%)。同时,全仿真过程中快速路主路车辆的车均通行时间降低了26.98%,路网内全部车辆的车均通行时间则降低了7.46%。更多还原

【Abstract】 With the continuous development of China’s social economy,the travel demand of large and medium-sized cities continues to grow.The construction of expressways has become the main means to improve urban traffic efficiency.Expressways usually intersect with urban roads in a threedimensional way.The common means of intersection include elevated roads and underpass tunnels.Among them,elevated road is a more traditional expressway construction method,but in recent years,tunnel expressway has gradually been widely used,and usually consists of multiple continuous tunnels.The emergence of tunnel group expressways not only provides continuous flow conditions for traffic flow,but also brings serious congestion problems,which mostly occur in the weaving sections.In order to solve this problem,it is necessary to implement specific traffic control strategies to alleviate traffic congestion and improve traffic efficiency by preventing the formation of traffic bottlenecks or accelerating the dissipation of traffic bottlenecks.However,at present,there are few studies on the control of tunnel group expressways and their surrounding road networks.Therefore,in order to put forward scientific and effective control strategies and technologies,this paper has carried out the following research in order to solve the congestion problem of urban tunnel group Expressway and its surrounding road network:Firstly,in order to study the causes of road network congestion in urban tunnel groups and analyze the key mechanism of capacity decline,this paper uses the video data of typical road networks captured by drones during peak hours to carry out qualitative and quantitative analysis on the impact of in and out flow and vehicle out / in ratio on the capacity of the weaving area of tunnel groups.Among them,binary logistic regression is used to establish the identification model of capacity decline in the weaving area,and the correlation between multiple candidate variables and their derived variables and capacity decline is modeled.The results show that the increase of ramp in flow rate,main road out flow rate,main road out proportion and ramp in proportion are all significant factors causing capacity decline in the weaving area.Then,as the basis of traffic control for the tunnel group road network,this paper carried out the research on the two modules of urban tunnel group road network state awareness and situation prediction.Their functions are to judge the current state of the road network and predict the evolution trend of the road network by using the traffic flow parameters provided by the traffic detectors in the road network.In the road network state awareness module,this paper uses the section speed difference algorithm to determine the bottleneck position of the main road of the tunnel group,and uses the occupancy queue number relationship to determine the queue length of the ramp and the intersection entrance.The function of the road network situation prediction module is realized based on the cellular transmission model,in which the latter has carried out secondary development for the tunnel group road network.The simulation tests of the state awareness module and the situation prediction module show that both of them can achieve the expected functions satisfactorily.In order to realize the regulation of urban tunnel group road network under congestion,this paper develops the coordinated control technology system of urban tunnel group road network.Considering the road organization structure of the tunnel group network,this paper adopts a three-level control strategy of coordinated control,and successively uses the three-level buffer zones of on ramp,upstream intersection and off ramp to control the traffic flow.At the same time,a three-level strategy cascade architecture is designed to control the orderly start of strategies at all levels within the priority framework by setting the road network state threshold.In addition,in order to solve the control problem of a single buffer zone,three sub controllers,ramp controller,signal controller and path guidance controller,are designed to control each buffer zone locally.In order to verify the effectiveness of the collaborative control method of urban tunnel group road network proposed in this paper,based on the above cascade collaborative control method and the design of three types of sub controllers,and integrating the road network state perception and prediction module,a prototype system of urban tunnel group road network collaborative control is built in sumo micro simulation environment.The simulation test implemented in the prototype system shows that,under the simulated traffic demand in peak hours,the collaborative control strategy developed in this paper reduces the upstream diffusion range of congestion by 1.3km(-44.83%),and the duration of congestion by 10min(-16.39%).Meanwhile,in the whole simulation process,the average vehicle travel time of vehicles on the main road of the expressway is reduced by 26.98%,and that of all vehicles in the road network is reduced by 7.46%.更多还原