滕靖
入学时间：2002年
答辩时间：2005年
论文题目：面向公交换乘枢纽的公共汽协调调度理论与方法
中文摘要
摘要
公共交通智能化调度是先进的公共交通系统（Advanced Public Transport System，APTS）的核心功能，是提升公交系统运行效率、提高服务水平的重要手段。在传统公交企业管理模式和调度体制下，公交调度权责仅限于单线路；APTS环境下，随着信息采集、处理、发布技术的完善，一种集约化的调度模式——区域调度成为公交调度的重要发展方向。这两类调度模式构成了智能公共交通调度的基础调度模式。然而公交系统的运输对象——乘客是不受调度范围限制的，客流需求分布在整个城市公交网络上，因此协调基础调度模式从而实现公交网络客流组织优化才是公交系统优化的最终目标。
本论文通过分析我国大城市公共交通运输组织方面的现状问题和发展需求，在总结国内、外公交调度理论与方法的基础上，提出了面向公交换乘枢纽的线路协调调度模式。旨在从公交网络运行层面初步建立起公共交通协调调度体系，优化公交网络客流组织。文中设计了该协调调度模式与两类基础调度模式功能上的协作关系并给出了计划调度和现场调度逻辑流程。同时还研究了支持本调度模式的关键技术——公交换乘枢纽客流OD自动推算和预测方法。
本论文从计划调度和现场调度两个层面建立面向公交换乘枢纽的线路协调调度方法。为与我国当前APTS研究进程相适应，文中设置了APTS非完全条件和完全条件两种情形。计划调度层面，根据上述两种情形重点研究行车时刻表优化，并就此建立了基本模型和协调模型。基本模型仅对行车间隔进行优化，执行其计划的现场调度的主要方法是公交车辆行车间隔控制。协调模型同时对行车间隔和离站松弛时间进行优化，执行其计划的现场调度的主要方法是公交换乘枢纽车辆驻站控制。本文分别对上述模型进行解析、给出优化算法并举例说明。文章最后，简要讨论了有待深入研究的内容。
本论文的研究成果为建立我国公共交通协调调度体系提供了系统工程层面的理论支持，并初步形成了相应的公交智能化调度决策方法，为进一步开发有我国自主知识产权的公交智能化调度系统提供了理论和方法指导。
关键词：协调调度，优化，枢纽，先进的公共交通系统
英文摘要
ABSTRACT
Public transport intelligent dispatching is the key function of Advanced Public Transport System (APTS) and an important means of improving the efficiency of public transport system and the level of service. Under the traditional management modes and dispatching systems of the public transport corporations, the dispatching responsibilities are always confined to single lines. Under APTS environment, as the technologies of information collection, processing and distribution advances, regional dispatching, an intensivism dispatching mode, is becoming one of the most important development trends of basic dispatching. The single-line dispatching and regional dispatching are the two basic dispatching modes of public transport intelligent dispatching. However, passengers, the objects of public transport, are not constrained by dispatching bounds. Passenger flows are distributed on the whole public transport network. Therefore coordinating the two dispatching modes to achieve passenger flow organization optimization of public transport network is the ultimate goal to optimize the public transport system.
This dissertation first analyzed the current conditions, existing problems and development needs of public transport dispatching in China’s big cities. After summarizing the national and international dispatching theories and methods, the multiple-line coordination dispatching mode for transit transferring centers was provided. The purpose was to set up a preliminary coordination dispatching system on the operation level of public transport network to optimize its passenger flow organization. Then the designed cooperation relationship between the coordination dispatching mode and the two basic modes and logic flows of planned dispatching and on-site dispatching were described. In addition, the automatic calculation and forecast technology of passenger OD (Origin-Destination) matrix of transit transferring centers, the key technology supporting the coordination dispatching mode, was studied in depth.
This dissertation set up the multiple-line coordination dispatching method for transit transferring centers from two aspects - planned dispatching and on-site dispatching. In order to adapt to the current development course of APTS in China, the method was developed for two situations - incomplete APTS conditions and complete APTS conditions. In aspect of planned dispatching, two models, basic model and coordination model, were set up for the above two conditions to optimize vehicle scheduling. The basic model only optimizes headway and the main corresponding on-site dispatching method is single-line vehicle headway control. The coordination model optimizes both headway and slack time and the corresponding on-site dispatching method is vehicle holding control for transit transferring centers. Detailed information, optimization algorithms and examples of the models were also given. In the end of the dissertation, the areas for future research were briefly discussed.
The research results of the dissertation provided theoretical supports on the system engineering level for setting up Transit Coordination Dispatching Systems in China. It also basically formed the corresponding public transport intelligent dispatching decision-making methodologies, providing theoretical and methodological guidance for the future development of China’s own public transport intelligent dispatching systems.
Key Words: Coordination Dispatching，Optimization，Transit Transferring Center，Advanced Public Transport System
张海军
入学时间：2002年
答辩时间：2005年
论文题目：城市高速道路入口匝道动态协调控制方法研究
中文摘要
摘要
城市高速道路是适应大城市汽车交通需求发展和土地利用特征而出现的，并通过其较大的服务容量和较高的服务水平对城市汽车交通出行做出了积极的贡献。但随着汽车交通需求的不断增长，各高速道路先后出现了交通拥挤，且拥挤程度日益加剧。根据国外经验，入口匝道控制是应用最广泛、也最有效的一种缓解高速道路拥挤的交通控制形式。
针对城市高速道路中的常发性交通拥挤，并考虑到目前实际可用的交通信息采集和处理技术条件，本论文选取一种曾得到成功应用的局部区域入口匝道动态协调控制方法——Bottleneck（瓶颈），在交通状态分析和仿真实验的基础上，对其进行研究和改进，从而提出更适合于城市高速道路的入口匝道动态协调控制方法。
Bottleneck方法将单点与系统协调两个层面的控制相结合，本论文首先从单点控制入手进行研究。选取需求-容量差额控制（Demand-Capacity Strategy，DC）和线性状态调节（Asservissement Linéaire d'Entrée Auotroutière，ALINEA）两种典型的入口匝道单点动态控制方法进行研究；对其中的排队约束模型进行细致的参数解析，提出一种更适合于城市高速道路的入口匝道到达交通需求预测方法，并在排队约束模型中加入对高速道路主线服务水平的保障机制。然后，对Bottleneck方法进行研究，在不纳入预测机制的前提下，针对协调层面调节率（Bottleneck Metering Rate，BMR）振荡、流量数据随机波动引起的主线拥挤状态漏检以及对上游入口匝道调整过度等问题分别进行研究，并提出相应的改进模型和方法。
本研究成果适用于目前交通信息采集和处理技术条件下的城市高速道路交通控制，并且能够随着相关技术的发展而得到进一步完善。
关键词：城市高速道路，入口匝道动态协调控制，入口匝道单点动态控制，入口匝道排队约束，主线拥挤控制，交通状态分析，交通仿真
英文摘要
Abstract
Urban freeway, designed to cope with the features of automobile traffic demand and land use in metropolis,contributes positively to urbanautomobiletrips with its large capacity and high LOS (level of service). With the continuousincreasing of traffic demand, many urban freeways became congested and the situations are getting even worse. According to the experiences in foreign countries, on-ramp metering is one of the most frequently used and the most effective freeway control means intended to reduce congestion.
Aiming at recurrent congestion in urban freeways, considering the practically existing technologies of traffic information collection and processing, a successfully used coordinated responsive metering methodology for neighboring on-ramps, namely Botttleneck, is chosen. Then it’s studied and improved on the basis of traffic simulation experiments and traffic status analysis. Finally, a coordinated responsive on-ramp metering methodology, which is fitter for urban freeway, is brought up.
In Bottleneck methodology, local metering and coordinated metering are combined. Two typical local responsive on-ramp metering methodologies, namely DC (Demand-Capacity Strategy) and ALINEA (Asservissement Linéaire d'Entrée Auotroutière), are firstly studied. Their on-ramp queue restriction model is analyzed and improved. Then Bottleneck methodology is discussed in detail. Under the precondition of no predicting mechanism, the methodology is adjusted against the fluctuation ofBMR (Bottleneck Metering Rate), the misdetection ofmainline congestioncaused byfluctuation of flow data andthe overreduction to metering rates of the upstream on-ramps.
The study results are fit for urban freeway control under the condition of existing technologies of traffic information collection and processing, and can be improved with the progress of relevant technologies.
Keywords: UrbanFreeway, Coordinated Responsive On-ramp Metering, Local Responsive On-ramp Metering, On-ramp Queue Restriction, Mainline Congestion Control, Traffic Status Analysis, Traffic Simulation.
严海
入学时间：2002年
答辩时间：2005年
论文题目：停车交通选择行为理论研究
中文摘要
摘要
城市社会进步和经济的发展，推动着城市的机动化。由于小汽车拥有和出行的增加，导致了城市停车交通产生了供需的矛盾，本文以停车交通需求的本源——停车交通的选择行为作为研究的对象，解释现象，探讨机理，对其本质进行解析，寻求进行停车需求管理的切入点和方法。为停车交通的规划、建设和管理提供理论指导和对策参考。
在本论文中，首先对停车交通选择行为研究结构体系进行研究。从停车交通的供需角度出发，分析停车交通选择行为对系统平衡的影响，针对与停车选择行为相关的停车交通问题，层进提出停车交通选择行为的研究需求体系、方法体系、内在机理等。
其次，对停车交通选择行为的调查和数据分析方法进行了研究，着重解决了三个问题：一是调查指标设计，二是抽样方法和最小样本量确定，三是针对SP调查，分析了不同的调查表达方法以及其可能存在的问题，并对调查误差产生的机理和调查数据的处理进行了研究。
第三，分析了停车交通政策对于选择行为的影响，主要包括三方面内容，一是停车供应时间弹性和效用,二是停车费用弹性和效用，三是时间和停车费用的替代关系。在此分析的基础上，提出停车政策的建议。
随后，对停车导行信息下的停车选择行为进行了两方面的研究，一是导行信息对行为的影响与约束，二是停车选择行为对于不同信息条件下的反应。并通过建立模糊推理的框架模型，解析停车选择行为的反应机理。
最后，针对特殊活动的停车选择行为与常态下的不同，分析行为的成因和后果，通过建立效用模型，给出解析特殊活动的停车选择行为的方法，进而提出管理的方法和思路。
关键词：停车交通 选择行为 供需 系统平衡 停车政策 停车导行信息 弹性
效用Logit模型 模糊推理 特殊活动
英文摘要
ABSTRACT
Social and economic development pushes urban mobility forward. Because of the increase in holding and using cars, the conflict between parking supply and demand appears in many cities. The main focus of this study is to study the parking choice behavior, which is the root of parking demand. Through explaining phenomena, discussing mechanism and analysis the essential of parking demand, This paper sought the cut-in point and methods to manage parking demand. And it will be useful to parking planning, constructing and management sectors.
The first part studied the system structure of parking choice behavior research. From the viewpoint of parking system structure, it is pointed out that the main influence of parking choice behavior is system equilibrium. Aiming at the problems related with behavior, the study demand system, method system, and internal mechanism are presented.
The second part studied the research and data analysis methods. Survey index design, spot check method and the minimal sample number are resolved. In the studied of data analysis method, aiming at SP survey, some survey expression and involved errors are analysized, the origin of survey errors are studied.
The third part of this paper studied the influence of parking policy to behavior. Three aspects are involved, the first one is time elasticity and utility, the second one is cost elasticity and utility, the third one is replaceable relationship between time and cost. And on the basis of analysis, policy advices are presented.
After these aspects, parking choice behavior under parking guidance information (PGI) is studied. One is the influence and restriction of PGI to behavior; another is behavior reaction under different PGI condition. And frame model based on fuzzy set theory is established, it is used to resolving behavior.
At last, aiming to the choice behavior dissimilarity under special and normal circumstance, the cause and solution of behavior is analysized, and utility model is established to explain choice behavior under special event. Further more, managing methods and ideas are advanced.
Key Words:parking, choice behavior, supply and demand, system equilibrium, parking policy, parking guidance information (PGI), elasticity, utility, Logit model, fuzzy set theory, special event
褚浩然
入学时间：2002年
答辩时间：2006年
论文题目：基于活动分析的交通行为特征和方式选择模型研究
中文摘要
摘要
随着城市规模的扩大和机动化社会的来临，日益突出的城市交通问题已经给城市方方面面的发展带来了负作用，交通运输规划与管理理论及实践的进一步研究成为目前学术界和工程界关注的热点。众所周知，城市交通是居民出行的派生现象，交通是一种源于活动的需求，居民日常交通行为如居民出发时间、活动目的、位置选择、交通方式选择和路径选择等，都与城市各类交通现象及问题的产生有直接或间接的联系，所以从时间、空间、活动和方式四个维度上研究交通系统是必要的。传统模型的建模一般利用集计的统计数据和出行特征参数，模型理论基础处于理想状态下，假定条件过多，与出行决策的实际情况有一定的出入，无法满足中、微观交通规划和交通管理策略分析的要求。
通过详细的居民活动日志调查或者对传统调查数据的整理和分析，可以准确地把握居民出行规律，将居民的一次出行看作是一种链的结构，提炼出行链特征参数，明确其物理意义，更加有助于理解交通问题和找到解决问题的办法。通过对影响交通选择的因素进行分析并量化影响因子，最终建立合理的交通方式选择模型，为交通方式分担预测奠定基础。
通过开展交通行为的研究，对居民出行方式选择的行为偏好进行调查，应用先进的社会学统计分析方法——结构方程模型，建立公交出行选择模型，作为交通政策分析和辅助决策的有力工具。
本文基于活动分析理论，从交通的源头出发，将出行目的和同质行为人群作为分类标准，根据实绩和偏好数据深入分析居民的交通行为特征，细化交通方式选择模型，研究活动分析方法与交通需求管理策略之间的关系，为中微观规划方案、交通需求管理策略的制定和决策分析提供切实的依据。
主要研究内容及成果如下：
第一章主要阐述了论文的研究背景及意义，说明了该研究在交通规划和交通管理研究领域的重要性，对研究现状进行了总结和分析，提出了基于活动分析方法在交通行为特征分析和方式选择模型研究方面应用的优越性，确定了本论文的主要研究内容，并给出了论文的逻辑结构框图。
第二章介绍和比较了基于出行的交通需求分析模型和基于活动的交通行为模型，从模型方法和实现功能两个方面说明了基于活动的交通行为模型应用的优越性。概述了交通方式选择模型和交通需求管理策略的国内外研究，对目前的应用现状和存在问题进行了评述，说明了活动分析方法在交通规划和交通需求管理上应用的必要性。
第三章主要研究内容及成果包括两个方面：其一提出了出行链相关概念、定义和出行链特征指标，分析了居民出行活动一般特征，并说明一般特征指标的局限性，证明出行链指标的物理意义，认为出行链特征指标用来评价城市出行强度更加适合，建议采用平均出行链长度作为衡量出行频率和长度的综合指标。其二从居民空间特征、居民属性和出行属性的角度，分同质行为人群和分活动目的分析了各类属性参数与出行链特征之间的相关性。
第四章的主要内容是围绕交通方式选择模型建模前的准备工作和正式建模进行，包括建模影响参数的解释、建模路线，确定要选择的参数，对影响参数进行定性和定量的分析。根据前面章节的分析，不同人群和不同活动目的的交通行为有显著不同的理论基础出发，建立了五类人群四种活动目的的交通方式选择模型，并与传统交通方式选择模型进行了比较，建议方式选择模型采用概念明晰、对影响因子进行精确分类的活动模型。
第五章的研究目的是建立一种有效的政策分析方法，具体步骤包括：使用SP调查方法和结构方程模型(SEM)对两个城市居民公交出行选择的偏好情况和影响因素进行调查和建模，设计了八项评价指标和三类潜变量，量化了各类影响指标和潜变量在公交方式选择重要性中所占的权重，并结合实际情况和公交改善策略对计算结果进行了分析，比较了慈溪和厦门两座城市居民对现状公交出行选择不同的态度，有针对性地分别制订了公交优先策略。
第六章从时间使用、行为聚类人群和目的分类、居民属性和出行属性分析三个主要方面论述了活动分析方法与交通需求管理策略的关系，说明活动分析方法是交通需求管理策略可行性和有效性的理论基础。并进行了实例研究，从行业特征、通勤时间分布、群众满意度、道路基础设施和公共交通客流量的时间分布特征五个方面论证了交通需求管理策略之一——“错峰出行”策略的可行性。
最后，对本文研究结论和成果进行了整理，从延续本研究的角度，提出了进一步的研究方向。
关键词：出行链；活动分析方法；方式选择模型；结构方程模型；交通需求管理策略；
英文摘要
ABSTRACT
With the expansion of city size and the coming motorized society, increasingly more and more transportation problems restrain development of city in many aspects. How to implement transportation planning and management according to realism is becoming attention focus by researchers and engineers. All is known, transportation is a activity-derived demand, there is direct or indirect touch between producing all kinds of transportation problems and daily travel behavior of residents, such as go-off, activity purpose,mode choice,and route choice. It is necessary of studying transportation system from time,space,activity and mode. Traditional model is calculated according to aggregated statistic data and trip characteristic parameters, and basic theory of model is at ideal state, which is different with actual trip decision. Current travel analysis theory is on the basis of aggregated data and don’t meet requirement of microcosmic transportation planning and management strategy analysis.
Through developing study of travel behavior, transportation planning models can be optimized further, contact between all kinds of models is set up and the model pool become organic whole, is not seriation. Avtivity-based model help to improve trip analysis approaches and provide detailed basis for microcosmic transportation planning project, transportation policy and decision analysis. Particular residents activity log survey and data processing are favor of holding residents trip rule. A trip is understood as chaining structure and trip chainings parameters are abstracted, which help to find approaches of solving transportation problems. Through analyzing influence factors of transportation choice and quantifying influence factors, reasonable travel mode choice model is presented as basis of mode split forecasting.
Surveying stated preference of residents mode choice and applying advanced society statistics method——structural equation model, public transportation choice model is set up, as a tool of policy analysis and assistant decision.
Base on activity analysis theory and transportation headstream, residents are divided into five behaviorally homogeneous groups. According to revealed and stated preference, travel behavior characteristics are analyzed further and transportation mode choice model is sorted. Relation of activity analysis theory and transportation demand management is studied, that provided actual evidence for microscopic planning project，transportation demand management strategy and decision analysis.
Main study contents and production are as follows:
In Chapter One, study background and significance of dissertation are set forth and importance of study on transportation planning and management is shown. Study status quo is sum up and analyzed and based on activity analysis theory, transportation behavior characteristic analysis and mode choice model study are applied, which is proved to be excellent. Accoring to it, study contents are made sure and logic structure flow chart is presented.
In Chapter Two, transportation demand model based on trip and transportation behavoir model based on activity are introduced and compared. From two aspects of model means and realized function, excellence of the latter is obvious. according to domestic and overseas study result，several aspects are summarized, including study on travel demand analysis model based on trip, study on travel behavior model based on activity, study on travel mode choice model, and travel demand management strategy. Problems and applied status quo are reviewed and activity analysis theory applied to transportation planning and management is necessary.
In Chapter Three, main study contents and production include two aspects. Firstly, trip chaining concept，definition and characteristic indice are presented, and residents ordinary trip characteristic is analyzed. Trip frequency is proved to be limited and physics meaning of trip chaining characteristic indice is clear. It appraised city trip intension to be suited and suggested average trip chaining length to be comprehensive index of scaling trip frequency and length. Secondly, from three aspects of residents space characteristic, family and trip attribute, relativity between all kinds of attribure indice and trip chaining characteristic is analyzed based on activity purpose and behaviorally homogeneous groups.
In Chapter Four, main contents enclose preparation and constitution of transportation mode choice model and include explanation of influence indice, course of model construction, indice choice and so on, and analysis of determining the nature and quantification. According to fore analysis, based on the theory that different groups and activity purpose have different behavior characteristics, mode choice model of five residents and four activity purpose is presented and compared with traditional model。The result is that activity-based model is excellent.
In Chapter Five, motive of study is setting up a effective method of policy analysis. Main contents are as follows. Firstly, according to Stated Preference Survey and Structural Equation Model, stated things and influence factors of public transportation choice are surveyed and based on surveyed data, model is set up. In the model eight evaluation indices and three latent variables are designed and weight of all kinds of indices and latent variables in public transportation choice are quantified. Calculation results are analyzed by the fact and improvement strategy of public transportation. Two cities, Cixi and Xiamen, are compared by attitude of improving public transportation and stratety of publice transportation priority is seperately drawed.
In Chapter Six, from three aspects of time use, behaviorally homogeneous groups and activity purpose sort, residents family and trip characteristic, relativity of activity analysis theory and transportation demand management is discussed, the result proved that activity analysis theory is theory and practice basis of transportation demand management feasibility and availability, and factual example is presented. The feasibility of Shanghai staggered shift is reasoned from five aspects of industry characteristic, commuter time distribution, mass satisfaction, time distribution characteristic of road infrastructures and passenger volume of public transportation
Finally, study conclusion and production is cleared up and further study direction is presented from visual angle of continuing the dissertation.
Keywords: trip chaining; activity analysis approaches; mode choice model; structural equation model; transportation demand management;
范海燕
入学时间：2002年
答辩时间：2006年
论文题目：公交线网调整决策支持系统理论与方法研究
中文摘要
摘要
在国家大力发展公共交通和提倡公交优先的背景下，研究了公共汽车网络（以下简称公交线网）调整决策支持系统的基础理论和方法。公交网络布局的合理化不仅可以提高公交系统的运输效率、提高服务水平，还可以降低运输成本，更好地支撑城市交通的可持续发展，同时可为发展先进的公共交通系统（Advanced Public Transportation System，APTS）奠定基础。
随着城市经济和社会的快速发展，土地使用性质不断变化，原有公交网络的布局呈现出不适应性，有调整之必要。然而公交网络的确定广泛地涉及到诸多影响因素，其调整决策工作量和难度皆较大，怎样的公交网络是最佳的？怎样调整公交网络一直是制约公交发展的一大难题。建立公交线网调整决策支持系统（Bus Network Adjustment Decision Support System, BNADSS）将为线网的合理调整提供可能，也是现代公交网络发展及运营管理之重大需求。
论文综述国内、外公交网络优化方法，并进行了理论和实际应用上的比较，同时深入地研究决策支持系统（Decision Support System, DSS）的基础理论、方法和技术，在分析了公交线网调整决策的屏障和需求的基础上，构建了BNADSS的框架结构，剖析了决策管理者的决策框架，并特别在模型技术上作了如下研究：运用层次分析方法建立了公交线网调整的决策模型，为决策方案的选择提供了技术支持；对BNADSS中模型进行了分类，运用网络拓扑结构的理论建立了路网可靠性模型；运用灰色系统理论与神经网络相结合的方法建立了公交客流预测模型，并与指数回归预测方法进行了对比，验证了模型较优；以可靠性作为一项重要评价指标，建立了公交网络评价指标体系，对厦门公交网络进行评价得出了线网的优化调整方案；在详细分析公交运行时间影响因素的基础上，建立了公交线路运行时间可靠性模型，得出了车辆准时到达每一站点和从某站到达任意站点的运行时间可靠的概率；最后，运用EMME/2软件对厦门市公交线网进行了调整，为管理者提供了决策支持。
本论文的研究成果理论层面为建立BNADSS提供了技术支持，应用层面为公交线网的实际调整作出了决策指导，为进一步开发BNADSS奠定了基础。公交运行时间的可靠性研究是未来公交领域研究的重要方向，值得关注。
关键词：公交线网，调整，决策，决策支持系统，优化，评价，可靠性
英文摘要
ABSTRACT
Based on the background of develop public transportation and give it the highest priority in china, researched the basal theory and method of bus network adjustment decision support system (BNADSS). Reasonable arrangement not only can improve transportation efficiency and service level ,but also can reduce transportation cost, and established foundation for transportation sustainable development and advanced public transportation system(APTS) .
With the quick development of society and urban economic, the property of land has changed continually, customary arrangement of bus network has not adapt to this, so to adjust bus network is necessary. But found a bus network will involve in many factors, both workload and difficulty are great in this process. Which bus network is optimal? How to adjust the bus network is one difficult problem, which will restrict public transportation system to develop. Founding BNADSS will provide possibility for rationally arrange bus network, and it is one important demand of the development and management of modern bus network.
The paper summarized the optimal method of bus network overseas and internal, compared the difference between the theory and practice, at the same time researched base theory, method and technology of DSS, based on analyzing the barrier and demand of bus network adjustment decision, founded framework structure of BNADSS. Analyzed decision framework of management, specially on the model researched the technology as follows: founded decision model of bus network adjustment by the AHP method; provided technical support for the selecting of decision plan; classified the model for the BNADSS, founded reliability model of road network with network topological structure; founded forecast model of public transportation passenger volume by combine gray systemic theory and naval network, and compared it with the exponential recursive method, validated the excellent of this model; founded evaluate index system of bus network on the base of taking reliability as a important index, evaluated and get a optimal adjustment plan for the XiaMen city; by analyzing the influence factor of bus travel time, founded reliability model of bus travel time, got the probability of reliability for bus to reach every stop and from one stop to another stop on time; finally adjusted the bus network of Xiamen city with emme/2 software, provided decision support for the manager.
On the theory, the achievement of this paper provided technology support for the BNADSS, and on the appliance, it can guide decision for adjusting actual bus network, and established foundation for the BNADSS exploitation. The research on the reliability of bus travel time is important direction on the field of public transportation system in future, it is worth give attention to it.
Key Words:Bus Network，Adjustment, Decision , Decision Support System(DSS)，Optimization, Evaluation，Reliability
杨天军
入学时间：2002年
答辩时间：2006年
论文题目：城市道路交通安全管理决策支持系统关键技术研究
中文摘要
摘要
道路交通伤害已经成为全球性公共卫生问题，道路交通安全对人类健康和社会经济影响巨大。目前中国道路交通事故死亡人数居世界第一，而且交通事故呈上升趋势，全国道路交通安全形势严峻。
在道路交通事故中，城市道路的交通事故比例高，虽然死亡事故不高，但是由于事故数量大，总的经济损失比例很高，因此如何进行有效的城市交通安全控制一直是国内外的交通工程领域研究的重点和热点，其次随着经济的发展城市化公路的交通安全状况也令人担忧，急需提高道路交通安全管理水平。
针对道路交通安全问题的定性多，定量少，道路交通安全管理者难以科学决策等问题。结合安全专家的经验和知识，引入决策支持的有关理论和技术，对道路交通安全信息进行采集、管理、分析、决策和反馈，以提高道路交通安全管理的效率和水平。本文探讨了城市道路交通安全管理决策支持系统的理论和方法，对系统管理中的事故成因机理分析与预测、安全评价方法、安全对策与决策等关键技术进行深入研究，还讨论了决策系统中相应各子系统的数据库和模型库等，并进行了详细的分析和设计。主要研究内容和成果如下：
1．论文从分析我国的道路交通安全状况入手，阐述研究城市道路交通安全管理决策支持系统的必要性和紧迫性，针对城市道路交通安全管理决策的紧迫性和决策所需信息的不完全性及不确定性等特点，按照管理决策支持系统建立的一般过程，对系统的功能和构成进行研究，在此基础上确定了系统各个主要组成部分和构建方法。重点从六个方面即道路交通安全综合信息管理、事故成因分析子系统、事故预测子系统、道路交通事件管理及预警系统、安全评价子系统、安全对策子系统、经济分析及决策子系统对道路交通安全管理决策支持系统研究设计。
2．结合实际调查数据，运用统计、数据挖掘等方法对我国城市道路交通事故和城市化公路交通事故进行成因分析，并对道路交通安全形势及事故预测方法进行汇总，提出利用不同车型进行万车事故率的预测方法和交通规划方案的事故预测评价方法。设计了事故成因分析和预测各子系统的理论框架和所需的模型库等。
3．针对国内外目前的安全评价方法的不足，结合交通管理部门的实际需求，从发生事故的潜在性（Exposure）、发生事故的机率（Risk）和事故后果(Consequence)三个方面提出一条路、一个区（城市的行政分区）及一个城市的道路交通安全评价指标和评价方法，为新的道路交通安全评价指标体系奠定基础，并给出了实例。利用交通冲突技术对交叉口进行安全评价研究，提出了道路交通安全评价决策支持子系统的双层模型结构。
4．在道路交通安全对策与决策方面，结合交叉口的安全特点，从法律角度明确了路权的概念，对交叉口安全设计进行研究，根据交叉口的事故分析提出安全对策，结合实例并对安全管理措施进行经济效益评价，对安全管理方案进行经济和决策分析。对城市交通干道的安全风险、对策及经济效益进行研究。同时建立道路交通安全对策子系统和安全管理方案决策的流程结构。
5．从系统构成、数据的管理和实现、模型的实现以及人机界面等方面对系统的实现进行分析设计和研究。并结合程序开发流程，进行实例应用研究。
本研究所提出的安全管理决策支持系统，通过安全管理和决策的相关理论和技术，以及对驾驶员和行人等的道路交通安全调查，对交通事故进行定性和定量的分析后，能够及时发现道路交通安全隐患，提出改善措施，减少事故的损失，对安全管理方案进行评价，辅助安全管理者决策，提高安全管理的效率。
关键词：道路交通安全，道路交通安全管理，决策支持系统，事故成因分析，事故预测，道路交通安全评价，交通冲突技术，道路交通安全对策，决策分析
英文摘要
ABSTRACT
Road traffic damage has been world commonality sanitation problem, road traffic safety influence human health and social economy greatly. The death number of traffic accident of our homeland is the first in the world, the number of traffic accident is increasing, road traffic safety condition of our country is very seriously.
In road traffic accidents, the proportion of city road is higher, although the fatal accident is low, the number of accidents is great, total economic loss proportion is high. The emphasis and hotspot problem is how to control the city road traffic safety effectively in civil and abroad traffic engineering, and with the development of economy, the high development of urbanization of highway, which has bad traffic safety status, is needed to improve the level of road traffic safety management urgently.
There are many quantitative problems in road traffic safety, however the problems have less qualitative nature, it is difficult to decide in science for road traffic manager. Integrating with experience and knowledge of expert, theories and technologies of decision support system (DSS) are added in traditional traffic safety management system, road traffic safety information is collected and managed and analyzed and decided and fed back in this paper, in order to improve the level and efficiency of road traffic safety management. Theories and methods of DSS on city road traffic safety management are discussed, some key technologies of this management system, for example: analysis and prediction of traffic accidents, safety evaluation method, safety measures and safety decision-making, also the database and modal database of some subsystem is discussed and designed in detail, main research result is described below:
Fist, basing on analyzing the road traffic safety status of our nation, it is necessary and urgency to research on the key technologies of DSS on city road traffic safety management, aiming at the urgency to make decision and lack of abundant and decided information, according to the procession of establishing DSS for management, system functions and structure are analyzed in detail, then main composing and construction method of management system are confirmed. The DSS of road traffic safety management is mainly designed in six parts, which includes: road traffic safety comprehensive information management, subsystem of accident causation analysis, traffic accident prediction subsystem, traffic incident management and alarming in advance, safety evaluation subsystem, safety countermeasures subsystem, economic analysis and decision subsystem.
Second, with the survey data, the cause of city road traffic accident and development of urbanization highway accident is analyzed in statistical and data mining methods, reference on the former method of road traffic position analysis and accident prediction, a new prediction method using traffic modes accident rate is put forward to predict the accident number of ten thousand vehicles, and the evaluation modal of traffic planning projects accident prediction is advanced. Theory frame, modal database and other database of subsystem of accident causation analysis and accident prediction are designed.
Third, comparison the limitation of current evaluation method, the evaluation index system and method about one road and administration sub-area of a city and one city are advanced from the exposure, risk, consequence of accident with the needing of traffic management branch, which is base of new road traffic safety evaluation index. At last, evaluate the cross safety using traffic conflict technology, Double level structural modal of road traffic safety evaluation subsystem is advanced.
Fourth, According to traffic safety countermeasures and decision-making, with the characteristics of crossing safety, might of road user is asserted in law angle, safety design of crossing is analyzed, basing on crossing traffic safety analysis, the safety countermeasures are put forward, Combing the application example, safety management countermeasures are evaluated in economic, safety management projects are analyzed in economic and decision-making. Basing on the analysis of traffic arterial road risk, safety countermeasures and economic benefit are analyzed. Then road traffic safety countermeasures subsystem and safety project decision flow chart are put forward.
Fifth, the realization of this system is studied and designed in system structure, management and realization of data, realization of model, man-machine interface and etc. Combining the process of programming flow, some applications are studied.
the DSS on safety management advanced in this paper, basing on theories and technologies of safety management and decision making, safety survey of driver and pedestrian, can analyze the traffic accident in qualitative and quantitative, detect the hidden road traffic safety, help to put forward amending measures and decrease accident loss, evaluate the safety management projects, support the safety manager to make decision, improve the safety management efficiency in fact.
Key Words:Road traffic safety, road traffic safety management, DSS, Cause of accident analysis, Accident prediction, Evaluation of road traffic safety, Technology of traffic confliction, Road traffic safety countermeasures, analysis of decision-making
