【作者】 刘昕；

【导师】 王殿海；

【作者基本信息】 吉林大学 ， 交通信息工程及控制， 2011， 博士

【摘要】 随着经济发展的高速增长,汽车数量急剧增加,交通拥堵问题导致的“城市病”愈演愈烈。交通是城市经济活动的命脉,直接关系着国民生计,因此解决交通拥堵问题成为全球关注的焦点。近年来,ITS从人-车-路的动态关系出发,以已有出行者信息为基础,提出一系列提高道路利用率和交通流量,降低交通拥挤程度和交通事故发生率,减少因交通拥堵和事故等造成的延误,以及科学诱导等方法,在一定程度上缓解了交通拥堵问题。交通信息是ITS中交通管理、交通控制、交通诱导、交通指挥及交通信息服务等功能的重要信息来源,是进行科学交通管理与规划的依据。因此,交通信息采集的全面性、可靠性、精确性和实时性直接关系着道路交通系统的管理效果和控制效果。现有的基于各种磁频、波频、视频和GPRS/CDMA等交通信息采集技术,获得了部分实时交通参数,解决了一些交通问题。但是上述方法受采集技术或环境限制,只能获得断面的交通信息,不能够及时全面地反映宏观交通状态或实时描绘微观出行者的需求。本文选择了逐渐替代数字广电、电信和网络业务的IPv6网络作为采集环境。IPv6具有大地址空间、安全性好、移动性支持优异和高效QoS等优点,契合了对全部出行者进行实时采集的需求。本文以国家重点基础研究发展计划(973计划)项目“大城市交通拥堵瓶颈的基础科学问题研究”课题“城市交通系统的组织优化与控制”和吉林省科技发展计划项目“移动IPv6运营基础架构研究”为依托,集中研究了基于IPv6的智能交通信息采集与处理方法,对未来智能交通网络建设与管理观念进行探索性研究。论文的研究建立在两个假设条件之上：第一,以全面的静态交通信息关联库为支撑,包括地理信息系统(GIS)、车辆自然信息库和出行人基本信息库等,实现人民生活全面数字化；第二,具有完善的交通通信网络,且车辆上均配置信息采集和移动通信的车载单元。本文首先提出了基于IPv6的人-车-路-环境协同的概念体系,建立了以信息流为主线的逻辑体系,搭建由移动接入层、公共传输层和控制中心组成的基于IPv6的ITS框架体系。构建一个集信息采集、通信、存储为一体的IPv6智能交通信息系统(简称IPv6信息系统)。车载信息采集单元完成前端信息采集,采集内容包括：GPS定位信息和其它传感器读取的车辆仪表信息,主要包括运行方向和速度等参数和出行人持卡识别信息。采集信息封装到IPv6数据包中,并通过高速传输网络报送至控制中心。采用光纤传输和远距离无线传输协议802.16e构建骨干传输网络。为避免车辆密集情况下,并发报送造成网络拥塞,设计了根据车辆的车况、通讯和信用状态信息组建相对稳定和快速的车载自组网络协议,降低了对无线信道的需求,减轻无线信道的拥塞,保证采集信息的实时稳定传输。控制中心进行信息提取获得智能交通参数,用于实时交通管理、控制与诱导,分析交通安全、出行分布和城市土地利用等问题。控制中心还完成交通网络AAA核心管理和信息存储与发布功能。出行者在移动过程中,需要不断地进行网络域间切换。不当切换方式将造成网络和应用的中断。为此,引入AAA机制进行交通网络的运营和管理,提出邻域快速认证、授权策略和全程记账,并建立基于消费信息进行实时收费的道路交通计费模型。通过认证、授权和记账的实时作用,实现基于AAA的智能交通网络动态管理方案,由此来保证车辆在进行域间切换时的网络连续和网络与交通资源的合理配置。在交通信息网络实时和连续的前提下,进行了交通流的描述和特征提取。以边标号法和矩阵论为基础,对宏观路网的自然属性、路网中路段的自然属性和节点的社会属性进行描述,并实现了矩阵、真实路网和抽象路网的相互转化,将真实的路网抽象成数学模型。在宏观路网和路段特性分析基础上,提出了微观的路段网格(Block)表达法。通过对Block的空间属性、邻居属性、车辆存在属性、时空通行限制属性的分析,建立了出行人、车辆、相位、微观Block、路段、节点和环境之间的时空协同关系模型。在该模型基础上,根据Block的车辆存在性,通过定位在Block内的车辆唯一标识UniqueID与车辆自然信息和位置报送信息的关联关系,采集到海量智能交通信息。进而提取出交通流三要素流量、速度和密度的特征信息数据——特别是其他方法无法准确获得的区间速度、密度等重要参数；提取出车辆OD矩阵,获得路径选择概率、单车转弯概率和公交乘客量等交通参数。仿真实验证明信息采集网络有效避免了网络拥塞,保证了系统传输效率；AAA机制可以实现对车辆的实时动态管理。车辆定位模型表达了人-车-路-环境的协同关系,既可以实现微观车辆实时信息采集与跟踪,又能够获得中观的流量、密度、速度等信息,还能够呈现宏观路网的实时状态和了解交通流参数的分布规律,可以为智能交通控制、诱导和安全提供依据。更多还原

【Abstract】 Rapid economic growth, vehicle possession increase, and traffic congestion result in serious "city problems". City traffic is vital in urban economic activities, and become a focus of solving congestion worldwide. In recent years, ITS, in light with user-vehicle-road dynamic relation, based on travelers’data, to some extent releases the problem of traffic congestion by improving road utility and traffic flow, lowering the rate of traffic congestion and accidents, reducing the delays caused by congestion and accidents, as well as guiding scientifically. Traffic information is the key source for functions of traffic management, traffic control, traffic guidance, and traffic information service. It provides foundation for scientific traffic management and planning. Therefore, comprehensive, reliable, accurate and timely traffic information collection directly affects the results of management and control in road transportation system.Based on current traffic information collection techniques of magnetic frequency, wave frequency, video frequency and GPRS/CDMA, some of the traffic problems can be solved when real traffic parameters are obtained. While restricted by surroundings and approached of information collection, only part of the entire information is available. Demand can not be met if macro transport situation or real micro individual travelers are required to describe. IPv6 network, which gradually replaces digital radiocasting, telecommunications and network, serves as data collection condition. It satisfies all the requirements for real data collection of all vehicles, with the merits of mega-spatial, guaranteed, mobile, and highly efficient QoS features.The paper is based on the Plans for National Key Rudimental Research Developments (Project 973)--"Optimal organization and control in city traffic system" in the Program of "Research on the fundamental scientific issue of urban traffic bottleneck", and the Plans for Jilin Province Scientific and Technological Development-"Framework research by mobile IPv6 orientation". Tentative research is made on the network construction and management concept of future intelligent transport. The research of the thesis is on two presuppositions. First, it is established on the foundation of overall static transport information database, including GIS, vehicle information database and pedestrian information database. City residents’life is entirely digitalized. Second, transport communications network is consummated and vehicles are installed vehicle-carried units for information collection and mobile communication.The system framework of comprehensive transport information collection and procession was first raised in light of IPv6 in the thesis. It is composed of three layers for collecting, transmitting and storing ITS information, i.e., mobile input interface, public transmission, and control center. Vehicle-carried data collection units collect information of GPS data, and vehicle meter data read by sensors. These data are about the running direction and speed, and travelers’ ID as well. The data are packed in IPv6 data package, and reported to the control center through high speed transmission network, which adopts light-fiber transmission and long distance wireless transmission 802.16e protocol. To avoid network jam in rush hour of vehicle data collection and report, we design the comparatively stable and rapid vehicle-carried network in accordance with vehicle conditions, communications, and accreditation. The adoption of 802.11g protocol makes multi-way transmission to reduce the requirement and congestion of wireless data channel. Real and stable transmission of collected data is guaranteed. Control center processes the collected data to make real traffic management and control, and to analyze issues of transport safety, trip distribution and urban land allocation. It is also accountable for the core functions of management, data archive and broadcast for transport network AAA.When vehicles run, continuous contacts are conducted within the domains in the network. Improper switches will block the contacts. The introduction of AAA system into transport network running and management may fasten the speed of ascertain and authority in adjacent domains, and formulate transport fee-levy models based on real time consuming data. Through the real time functions of ascertain, authority and record, dynamic management approach of intelligent transport network is realized base on AAA. Network continuity and rational allocation of transport resources are guaranteed when vehicles run past different domains.At the premise of real and continuous transport information network, the features of traffic flow are defined.Based on marginal tag method and matrices theory, the natural features of macro road network and blocks, as well as the social features of nodes are defined. Real road network is abstracted into mathematic models to actualize the transformation of matrix, real road network, and abstract network. After analysis on macro road network and blocks, the idea of block is stressed on to analyze its spatial, adjacent, vehicle existing, and temporal-spatial restriction features. Temporal-spatial coordination models are formulated among pedestrians, vehicles, phases, micro block, route, nodes and environment.When the UniqueID is joined with reports of vehicles’ natural data and their positions, numerous data are collected, considering the properties of vehicles’existence in the block. Flow, speed and density, three key parameters of traffic flow, are obtained, especially variable speed and density that can hardly gathered by other means. OD matrices are formulated to get the probabilities of route selection, single vehicle turning, and volume of public transport passengers.Simulation tests prove that data collection network effectively tackles network congestion, guarantees system transmission. AAA system makes real time dynamic management on vehicles. Coordination of pedestrian-vehicle-road-environment is expressed by vehicle positioning models. Three levels are achieved to lay foundation for intelligent transport control, guide and safety that micro vehicle real time data are collected and tracked; flow, density and speed are obtained; real time data and traffic flow parameters in macro road network are found.更多还原