陆莹

入学时间：2005级

答辩时间：2008年

论文题目：轨道线路流体化及其应用研究

中文摘要

摘要

轨道线路流体化的基本思想，是通过让轨道交通车辆提前一定时间有目的减速，使得车辆在通过下一个轨道交通信号灯时，其信号刚好由红色“停车”信号转变为绿色“通行”信号，从而避免车辆刹车停止再重新启动的过程。与传统的轨道交通运行方式不同的是，线路流体化方式能够有效地提高高密度区间的轨道交通准点率，并且增加其通行能力，是城市轨道交通研究领域借鉴城际铁路技术的一个新尝试。

本论文首先分析了线路流体化这一技术从城际铁路领域转移到城市轨道交通领域的主要原因，以及技术转移的可能性。特别阐述了线路流体化应用于城市轨道交通的研究价值。

论文介绍了利用专业数据库与一般统计工具相结合进行数据分析的方法，并对全欧洲轨道交通运行密度最高的区段——巴黎北站站和夏特来站之间的连接隧道进行数据抽取，重点分析了通行时间与延误、以及延误变化等参数。

论文探讨了高密度轨道交通网络的建模方法，介绍了这一交通网络局部模型建立与检验、整体模型融合等一系列过程，并通过理论阐述与实践应用相结合，借助专业建模工具Sisyfe对研究隧道及周围环境进行系统建模。

论文同时提出了基于线性规划的流体化方案优化方法。系统阐述了流体化控制模型的优化机理，与建立在辅助工具——轨道交通软件Lipari基础上的实施方法。

论文还提出了流体化与非流体化方案的模拟对比方法。通过Sisyfe和Lipari两大软件就三套模拟方案加以系统实现，并将同一套模拟方案下的流体化与非流体化模拟结果进行定量对比分析，重点评估线路流体化在提高准点率和减少对其它车辆延误等方面的积极效应。

最后，本论文就整个研究进行了总结，并对潜在的问题和进一步研究方向进行了简要讨论。

关键词：线路流体化，轨道交通信号，建模，准点率，线性规划，会聚冲突，运行路线时刻图。

英文摘要

ABSTRACT

The fundamental thought of line fluidization in urban rail transportation is to slow a train down before it reaches a rail communication signal, providing a sufficient time gap for the signal to change from red “stop” to green “pass”, and thus avoiding an abrupt process of braking to stop and restarting. Compared with the traditional ways of operation, the method of line fluidization improves the punctuality of every train in areas of high traffic density and at the same time increases the overall operating capacity. It is hence considered a new trend of technology to transfer from the research area of inter-city railway to the other research area of urban rail transportation.

First of all, this thesis analyzes the main reasons for introducing the technology of line fluidization to the research area of urban rail transportation from the original research area of inter-city railway, and the technological feasibility of such a transfer. It particularly expounds the value of this research.

In its main body, this thesis introduces the method of data analysis based on a professional database combined with the classic statistical tools. Examining in particular the highest density section of Europe-wide urban rail transportation, the tunnel connecting the station of Paris North Station and the station of Châtelet, this thesis uses selected data for its analysis with the emphasis on certain parameters such as the travel time, the delay, and the evolution of delay.

Also, this thesis proposes a modeling method of high-density urban rail transportation network. It introduces a series of procedures about building and testing local models, as well as their integration for the whole model. Through the approach of methodical explanation combined with practical application, it also sets up a system model for this tunnel and the surrounding environment, with an assistance of the professional tool Sisyfe.

Moreover, this thesis proposes an optimization method for fluidization scheme with linear programming. It explains systematically the principle of optimization for fluidization controlling model and the implementation in accordance with this principle that bases on the assistant urban rail transportation software, Lipari.

In addition, this thesis proposes a method of comparison between fluidization scheme and non-fluidization scheme by simulations. With the use of two important softwares Sisyfe and Lipari, three sets of simulation schemes have been realized systematically. Besides, it analyzes the simulation results of fluidization and non-fluidization schemes in the same set, by the quantitative comparative method, focusing on the evaluation of line fluidization’s advantages of improving the punctuality of one train and reducing the delays in others trains.

Finally, this thesis summarizes the whole study, and briefly discusses the potential problems and interesting topics for further study.

Key Words:Line fluidization, Signal in urban rail transportation, Traffic model, Punctuality, Linear programming, Convergent conflict, Space-time graph.

马芳

入学时间：2005级

答辩时间：2008年

论文题目：智能公交系统中GPS信息质量问题解析研究

中文摘要

摘要

智能公交系统是智能交通系统的重要研究内容。用GPS定位技术改进公交系统的指挥调度和信息服务是提高公交服务水平的重要手段，是智能公交系统的主要研究领域之一。

本论文首先对GPS技术在智能公交系统中的应用进行探讨，系统地分析了公交定位系统的信息需求、信息属性和信息质量要求，全面解析GPS信息在采集、处理、传输和应用四个环节中可能产生的误差，针对各种误差分别提出改进方法。文中采用模糊评判算法获得定位系统的实时定位状态的可信度值，对实际应用中的GPS信息质量做出科学、合理的评价，为后边部分提出信息质量改善技术提供科学依据和数据支撑。

为改善GPS信息质量，用Kalman滤波对GPS信息进行数据预处理和信息融合，并给出了处理静态、动态定位数据的Kalman滤波模型的参数选取方法。通过Kalman滤波，去除原始GPS数据中的噪音数据和不合理数据，为下一步的地图匹配提供可靠的数据基础。Matlab仿真试验表明，采用数据融合算法后GPS定位精度明显优于GPS初始定位精度。

利用电子地图的地理数据对经过数据融合的公交车辆定位数据进行配准纠正即地图匹配，来协调GPS定位信息和电子地图的道路信息所造成的显示误差，最后，通过实验对地图匹配算法的性能进行了验证。

关键词：智能公交系统，GPS，数据融合，卡尔曼滤波，地图匹配

英文摘要

ABSTRACT

Advanced Public Transportation Systems (APTS) is an important part of Intelligent Transport System (ITS). The core method of improving the public service level is to use Global Position System to enhancing PT dispatch system and information service.

Firstly, the article discussed the application of GPS location technique to public traffic system. Through analyzing the information need, the information attribute and the information quality requirement of public location system, a comprehensive analytical of GPS information error can result from the parts of collection, processing, transmission and application. In view of the kinds of error, the article brings up improvement method. The confidence of position state can be obtained by using the rank blur actuality assessment model. The scientical and rational evaluation of GPS information quality will provide scientific basis and data supporting to the data fusion and map-matching.

In order to improve the quality of GPS information, the paper gives the data processing model—Kalman filtering model and the selection of parameters used in static situation or in dynamic situstion in the model. The noise data and the unreasonable data can be removed from the original data through Kalman filtering. Matlab simulation shows that after using data fusion algorithm, the GPS location accuracy is better than the initial GPS location accuracy.

Finally, the article use the geographic data of the digital map to registrate the bus location data, that called map-matching, to coordinate the error that caused by the GPS location information and the road infoumation on the electronic map. The map-matching algorithm performance is verified through the experiment.

Key words:APTS, GPS, Data fusion, Kalman fitering, Map-matching

宋光华

入学时间：2005级

答辩时间：2008年

论文题目：城市平面道路连接交通影响分析方法研究

张海雷

入学时间：2005级

答辩时间：2008年

论文题目：信号控制交叉口转向交通设计适应性研究

中文摘要

摘要

城市道路交叉口具有转向、连接的功能，是实施交通可达性的重要保障，交叉口不仅是交通事故的多发地，也是道路网络中的瓶颈，制约整个道路系统交通功能的发挥。

本文首先分析了信号控制交叉口运行特征，针对交叉口潜在冲突点位置和数量，分析在时空资源设计后的变化情况，指出转向交通承担着交叉口转向的功能，为交叉口的主要研究重点和难点。在分析交叉口转向交通的运行特征和影响因素后，结合研究综述，确定本文研究重点为掉头交通、右转交通和公交专用进口道与转向交通协调设计。

在掉头交通研究中，论文结合掉头交通设计现状和运行特性，总结其主要影响因素，归纳总结出三种设计模式和时空设计方法，选择安全性和通行能力进行适应性分析。在通行能力分析中，建立右转交通干扰条件下的掉头交通通行能力模型，并通过实际调查标定相关参数。

在右转交通设计研究中，论文运用定量与定性结合的评价方法，重点对空间设计研究：对右转交通设计因素中的车道宽度与转弯半径、渠化岛设计进行详细分析，并给出其设计适应性。

最后，针对现有公交专用道在交叉口与转向交通运行交织的现象，提炼出公交专用进口道与转向交通协调设计的问题，指出通过交通组织和车道功能置换来解决，重点分析路边式公交专用道与右转交通的交织设计模式，并通过延误建模分析得到设计模式的适应性。

论文选择转向交通中掉头交通和右转交通重点分析，通过其相互运行关系分析和时空资源优化设计，结合安全性和通行能力评价指标进行适应性分析。成果可以丰富信号控制交叉口相关设计的理论，可以为我国交叉口的设计与改造提供技术与方法，同时也为实现城市道路公交优先提供一定的设计建议。

关键词：信号控制交叉口，掉头交通，右转交通，交通设计，公交专用道，适应性

英文摘要

ABSTRACT

Intersection has the function of turning and link, and ensures to carry out the accessibility of traffic. Intersection is not only the place where traffic accidents occur frequently, but bottleneck that restricts capacity of the whole road network.

The thesis analyzed the operation features of the signalized intersection firstly. According to analyses the position and number of the potential conflict points which have the difference after the design of space and time, it showed that turning traffic took on the function of turning which was the difficulty in the intersection. By analysesing the operation features and theinfluencing factorsof turning traffic, and considering research summarization, it ascertained that the research were the traffic design of u-turn、right turn and coordination bus-only approaches with turning traffic.

In the research of traffic design of right turn, the thesis combined the present situation and operation characteristics of u-turn, and summarized the maininfluencing factors, and summed up three kinds design patterns including their design of space and time. Then it analyzed the applicability by comparison of the security and capacity. In the analysis of capacity, this study proposed the capacity models of u-turn under the right turn, and demarcated the parameters with the survey.

This study combined quantitative and qualitative evaluation in the research of traffic design of right turn, and focused on the geometric design and applicability of the design factors of right turn: the width of lane with the turning radius、the design of island ofdiversion channel.

Finally, this thesis abstracted the problem in coordination bus-only approaches with turning traffic by their phenomenon, and indicated that the solutions were traffic organization and replacement of the lane function. This study analyzed the design patterns of bus lane in roadside with right turn, and proposed the delay model to gives appropriate solution of the design patterns.

This thesis focused on the u-turn and right turn in the turning movements, analyzed their operation system and optimization design of the space and time, and gave the gives appropriate solution by the evaluation of the security and capacity. Its achievement can enrich signalized intersection theory, provide the technology and method for design and reforming in intersections ofChina, and gives the suggestions for the design of bus priority in intersections.

Key words：Signalized intersection，U-turn，Right turn，Traffic design，Bus lane，Applicability

劳云腾

入学时间：2005级

答辩时间：2008年

论文题目：信号控制交叉口交通阻塞动态预测方法研究

中文摘要

摘要

交通拥挤和交通安全问题是世界上两大交通问题，研究城市信号交叉口交通阻塞预测，对预防和缓解城市道路交通阻塞问题有着重要的理论意义和实践价值。

论文首先分析了城市道路信号交叉口的交通阻塞机理，总结出三大阻塞发生原因：需求过大、供给不足和供需间的协调不够；并归纳和提出了城市道路的两种交通瓶颈：瓶颈断面和瓶颈连线，以及由交通瓶颈引发的两种类型的交通阻塞：瓶颈断面阻塞和瓶颈连线阻塞。在交通阻塞机理分析的基础上，论文随后在阻塞量化、阻塞预测、阻塞预测模型分析和阻塞预测方法运用等四个主要部分展开深入研究。

在阻塞量化研究中，论文提出两种基本交通阻塞量化模型：符合描述瓶颈断面阻塞的需求—容量法和符合描述瓶颈连线阻塞的连线承载率模型，以及适合分析城市平面道路阻塞问题的两种改进模型：考虑到达率的需求—容量法和考虑转向率的连线承载率模型；并将考虑转向率的连线承载率模型在三种基本渠化路段中进行运用分析；同时本文还对不同模型的优缺点和适用性进行了深入的分析。

在阻塞预测研究中，文章建立了城市信号交叉口交通阻塞动态预测模型。论文提出了瓶颈断面的判断方法，归纳了两类城市道路交通阻塞的发生条件，并在此基础上提出两类阻塞问题的动态预测模型；鉴于交通不稳定的特性，本文还提出预测阻塞发生的概率模型，从更为合理的角度描述阻塞发生的状态。在研究道路阻塞问题发生概率的同时也定义了连线保持畅通的可靠度和路网保持畅通的可靠度，可以为研究路网可靠性提供依据。

在交通阻塞预测的基础上，论文提出了描述城市道路交通状态的六种交通状态空间，其中“空间E”是瓶颈断面阻塞空间，“空间F”是瓶颈连线阻塞空间。同时论文也对交通状态空间的演化规律和空间特性进行了详细分析，这为交通管理和控制策略的制定提供了基本的依据。

在阻塞预测模型分析研究中，论文分析了预测模型参数的影响。分析了三个主要参数对阻塞预测的灵敏度：预测时刻、连线输入和连线输出，以及连线输入和输出关系（包括流量关系和信号联动控制关系）对阻塞的影响，在此基础上文章提出了缓解策略的基本的效用量化模型。效用量化模型的提出为交通阻塞策略的优选提供了依据。

在阻塞预测方法运用研究中，论文在交通阻塞问题分析和预测的基础上提出了城市道路交通阻塞的预防和缓解的基本流程以及提出阻塞的输入输出管理控制策略的优选方法；并总结了五种阻塞改善策略：规划层面优化、政策层面引导、交通管理层面优化、交通控制层面优化和交通设计层面优化。以往的研究中并没有重视控制和设计层面的“降需策略”，往往追求局部的最优而导致了路网整体效益的下降，本文通过仿真进一步说明了这个问题。论文还提出了预防和缓解交通阻塞的“四步模型”，这个四步模型对缓解城市交通阻塞问题，提高城市网络的可靠性和居民出行效率等方面都有着重要的意义。

最后论文总结了本研究的主要研究成果以及不足，本文的研究主要可以应用于城市信号交叉口自适应主动交通控制、城市道路主动交通管理以及城市道路网络交通分析及状态预测等方面。不过论文也存在一些不足，对于不同情况下一些应用上细节问题还需要进一步研究。另外在本论文的基础上，主动交通工程研究、交通预报研究和多来源数据的融合等方面研究也是在本论文的后续研究的主要方向。

关键词：信号交叉口，交通阻塞，动态预测，交通状态空间

英文摘要

ABSTRACT

The traffic congestion and traffic accident are the two major problems for traffic engineering. Doing more research on the traffic congestion forecast is a crucial element in preventing and mitigating urban traffic congestion problems theoretically and practically.

First of all, the mechanism of urban traffic congestion at intersection is analyzed, and three types of factors for traffic congestion are summed up: too large traffic demand, not enough traffic supply and inharmony between these two factors. In addition, the research puts forward two kinds of traffic bottleneck: bottleneck point and bottleneck link, and two kinds of traffic congestion: bottleneck point congestion and bottleneck link congestion. Based on this analysis, four main issues are carried out in the following researches: traffic congestion quantification, traffic congestion forecast, analysis for forecast model, and application for forecast model.

In the research of congestion quantification, this thesis puts forward two basic kinds of quantification models: demand-supply model for bottleneck point congestion and link loading-rate model for bottleneck link congestion, and two improvement models for ground traffic congestion problem: demand-supply with arrival rate model and link loading-rate with turning rate model. And then link loading-rate with turning rate model is applied and analyzed in three basic types of channeling sections. The advantage and applicability for these four models are also compared in detail.

In the research of traffic congestion forecast, dynamic forecast models for traffic congestion at signalized intersection are proposed. This paper presents the method of determining the bottleneck section, concludes two types of congestion occurring situstion, then, bring forward the dynamic forecast models for traffic congestion. Moreover, for the unstable character in traffic system, probability model, which can be more reasonable to describe the congestion state, is put forward for traffic congestion forecast. In addition, the link or network reliability, which is useful for the network reliability research on transportation, is defined with the research of probability model.

Based on the traffic congestion forecast, six types of traffic state spaces are put forward to describe the urban road traffic condition, with “space E” is bottleneck point congestion space and “space F” is bottleneck link congestion space. Meanwhile, the evolved law and space character are analyzed in detail, which can provide some basic rules for the determination of traffic management and traffic control strategy.

In the research of analysis for forecast model, the effect of parameters in forecast model is analyzed. Firstly, the sensitivity of three main parameters—forecast time, link input/output and effect of relationship (flow relationship and control signal relationship) between input and output—is analyzed. And then based on the analysis, this thesis proposed a basic quantificational utility model which can provide some basic idea for the choice of traffic congestion mitigation strategy.

In the research of application for forecast model, based on the analysis and forecast, this thesis provides a basic flow process for preventing and mitigating urban traffic congestion, and optimized method for the traffic management and control strategies. Moreover, five types of traffic congestion mitigation strategies are concluded: optimization in traffic plan level, guidance in traffic policy level, optimization in traffic management level, optimization in traffic control level and optimization in traffic design level. In previous research, the reduction of traffic demand strategy in control and design level do not been considered seriously and the pursuit of local optimization may cause the benefit damage of the whole network, this problem is demoed with the simulation. In addition, this thesis brings forward a “Four-Step Model” for preventing and mitigating traffic congestion which is significant for the improvement of network reliability and travel efficiency.

In the finality, this thesis sums up the main conclusions, disadvantages and applications: initiative traffic control, initiative traffic management, and network traffic state forecast and analyzed. However, more researches are still needed, such as some application detail in different case, initiative traffic engineering, traffic information predict and data fusion for different kinds of traffic data.

Key Words:signalized intersection, traffic jam, dynamic forecast, traffic state space

夏胜国

入学时间：2005级

答辩时间：2008年

论文题目：公交枢纽客流特征分析及行人路径交通分配研究

中文摘要

摘要

大力发展公共交通，建立城市综合交通网络体系是解决城市交通出行难问题的最有力措施之一。公共交通换乘枢纽是综合交通网络中各种交通方式的衔接点，是锚固城市公共交通网络体系的基础，也正是乘客出行过程的“瓶颈”，其布局规模、功能定位、设施配置、换乘布局模式以及交通组织与信息化管理水平等因素将对整个网络的运输效率产生决定性的影响。以往对枢纽的换乘研究更多的从工程的角度出发，缺乏从行人在枢纽中的行走特性与在枢纽内部行人通道系统中的分配来考虑枢纽的布局、规模、换乘衔接等问题。

本文主要针对枢纽中运动的行人特征与分配原理进行研究，通过对枢纽客流的预测以及客流分配情况的研究，进而对枢纽的换乘衔接布局、设施规模等提出反馈与意见。

首先，论文对枢纽内部的行人运动特性进行了研究，分别从宏观和微观的角度，对行人的交通流特性进行了阐述与探讨。此外，还对枢纽的客流特征以及慢行交通语言对枢纽行人行为的影响进行了研究。

其次，对枢纽的类型及其所对应的客流进行了分类，并建立了一套枢纽客流预测方法。引入了枢纽内部行人OD的概念，描述了将预测得出的枢纽各交通方式之间的换乘客流矩阵转换为枢纽内部行人OD的基本步骤。

最后，论文对比了道路交通流分配与枢纽内部行人分配之间的异同点，对枢纽内部行人路径选择的影响因素以及行人流在枢纽内部的分配原理进行了研究，建立了路径选择模型以及枢纽分配基本流程框架系统，并对行人设施与客流大小之间的反馈关系进行了研究。

论文较为全面的阐述了枢纽客流预测方法与枢纽内部客流分配的基本原理与模型框架，并研究了枢纽行人设施与客流量大小之间的反馈关系，对枢纽的设计与建设具有一定的指导和借鉴作用，也为枢纽服务能力的评价提供基础数据。

关键词：公共交通枢纽，枢纽客流特征，客流预测，行人流分配

英文摘要

Abstract

Developing public transportation and establishing the urban transit network system is one of the most important measures which are solved the urban traffic problem. Public Transit transfers are the fundamental facilities to anchor the urban transit network system and to connect various passenger transportation modes. They are also the bottleneck of passengers’ trips. Their layout and scale, function allocation, facilities location, transfer layout modes and traffic organization etc. facts have a decisive effect. The past study about transfer transit was always thought about the problem from the angle of project. There are lack that the way of thinking about the problem of transit’s layout and scale, transfer from passenger flow characteristics and assignment.

The research discusses the passenger flowcharacteristics and thetheorem aboutpassenger flowassignment in the transit passenger system. Through the research we can bring up the feedback and suggestion about transfer transit’s layout, facilities and scale.

At first, the research discusses the passenger motioncharacteristics in the transfer transit. From themacroscopic and microcosmicangle, thethesis discusses thepassenger flow characteristics.Besides, the thesis discusses the effect which the passenger transportation language influence the passenger actions.

Secondly, thethesis discusses theclassification of transit category and its passengers flow. And the thesis also builds up a set method which forecast the passenger flow within transit. The thesis expands on the concept about passenger OD of transit inside, and discusses the basic steps which change the traffic mode transfer matrix to transit inside passenger OD matrix.

Finally, the thesis discusses the factors which effect passenger route choice inside of transit and the basictheorem with passenger assignment in transit, and builds up the route choice modes and the framework of passenger flow assignment procedures.

Key Words：Passenger flow characteristics, Passenger flow forecast,