【作者】 戢晓峰；

【导师】 刘澜；

【作者基本信息】 西南交通大学 ， 交通运输规划与管理， 2009， 博士

【摘要】 随着社会经济的快速发展与城市化进程的加剧,城市交通拥挤正成为一个普遍的社会现象,尤其是在大中城市,交通拥挤已经成为影响城市功能正常发挥和可持续发展的一个全局性问题。如何有效的进行交通拥挤管理,一直是世界各国交通研究与管理人员关注的热点问题。随着先进的信息技术、控制技术等引入交通领域,在此基础上发展起来的智能交通系统(intelligent transport systems, ITS)被公认为是能够全面有效地解决交通拥挤问题的最佳途径。由于ITS的信息具有海量、多源、异构等典型特征,对交通信息的处理、检索和提取,已成为ITS发展和运用的关键。高效的交通信息提取方法,是基于ITS进行城市区域路网拥挤管理的前提。本文在城市交通拥挤管理与交通信息提取相关研究的基础上,深入研究交通状态信息、交通流特征信息的提取方法,并依据交通管理者与出行者的实际需求,实施有效的区域路网拥挤管理方案。本文的主要研究内容包括以下几个方面：回顾城市交通拥挤管理政策与相关技术的研究进展,分析了区域路网拥挤管理的信息需求。在描述交通信息特征的基础上,完善了交通信息提取的概念与理论体系,并建立了基于交通信息提取的区域路网拥挤管理框架。将粒度计算理论运用于交通信息提取,定义了交通信息颗粒及其粒度,并基于vague集理论提出了交通状态信息颗粒的构造方法。提出了区域路网的交通状态信息提取方法,以实现路段交通状态判别与区域路网瓶颈路段的信息提取。为了科学分析区域路网的整体交通运行状态,基于模糊推理提出了动态路网交通状态的判别方法,为宏观路网的交通拥挤管理提供更为丰富的交通状态信息。建立了浮动车、感应线圈与人工报告作为信息源的模糊信息融合模型,能够实现路段与区域路网的交通状态判别。确定了路网瓶颈路段识别的主要影响因子,提出了基于粗糙集的路网瓶颈路段识别方法,通过规则提取以识别路网瓶颈路段。将交通状态的判别结果作为决策属性,综合考虑道路等级、时间及交通参数,构造区域路网的交通流特征信息表,运用粗糙集方法提取路网交通流特征。利用图论对城市路网进行描述,建立了基于信息熵的路段交通流特征信息量评估模型,并结合实际路网的拓扑结构特性,设计了可变信息板布局方案的初步生成方法。针对驾驶员的VMS信息响应特性,设计了面向拥挤管理的VMS诱导策略,即在关联路段均处于拥挤状态时基于vague集对交通状态进行排序,以判别交通状态相对最优的关联路段进行发布。通过分析交通诱导系统用户的路径规划需求,判定了理想路径的影响因素及个性化路径需求的属性与特征,分别针对距离、时间、服务水平及综合属性分析了相应的路径规划方法,并设计了个性化路径集成规划框架。在研究出行信息的内容、特征与传递效用的基础上,建立了面向多用户的出行信息传递模型与搜寻模型,详细分析了多用户的出行信息需求。依据我国出行信息传递的实际情况,提出了出行信息有效性的评价指标体系和评价方法,并建立了出行信息传递指数模型,能够分别从出行者与交通管理者的角度,评价出行信息接受水平与传递质量,为区域路网的出行信息传递模式设计提供理论支持。更多还原

【Abstract】 With the development of society and economy and urbanization process, the traffic congestion is prevalent in the world, especially be in large and middle-sized cities, urban traffic congestion has become an important issue that affects normal urban function exertion and urban continually development. How effective to carry on the traffic congestion management, has been the hot topic by researches and engineers in various countries.Along with the advanced information technology, control technology and other technologies were introduced into traffic field, ITS has developed on the basis of these technologies, and traffic congestion was considered could be resolved effectively and completely based on ITS. Because ITS information has multi-source, heterogeneous and so on typical characteristic, travel information processing, indexing and extraction have become the key of development and application of ITS.The highly-efficient traffic information extraction method, is the premise of congestion management in regional road networkt based on ITS. On the basis of congestion management and traffic information extraction and relevant study, this dissertation makes an in-depth study of traffic state information and traffic flow feature information extraction methods, then implements the effective congestion management plan in region road network based on the practicality demand of traffic managers and travelers.The main contents of this dissertation are as follows:The research progress of congestion management policy and the correlation technologies was reviewed, and the information demand of congestion management in region road network was analyzed. The concept and theoretical system of traffic information extraction was proposed based on traffic information characteristic description. Then a methodology that provides a framework of congestion management was presented based on traffic information extraction. Granular computing theory was applied in traffic information extraction, and traffic information granule and its granularity were defined. A method was proposed for traffic state information granule construction based on vague sets. The traffic state information extraction method of regional road network traffic state was proposed for road section travel state and bottleneck identification.In order to analyze the traffic state of regional road network accurately, travel state analysis methods were proposed for regional road network based on the fuzzy inference theory; it could provide the richer travel state information for macroscopic congestion management. A fuzzy information fusion model was proposed for multi-source information from floating car, loop and artificial report; it can identify the travel state of links and regional network. The major influencing factors of bottleneck links identification were determined. The bottleneck links identification method was proposed through rules extraction based on rough set theory. The travel state was acted as decision attribute, and then the traffic flow feature information table was constructed by road grade, time and traffic parameters. The traffic flow features of regional road network were extracted based on rough set. The urban road network was described based on graph theory, and also the information quantity evaluation model of traffic flow feature was proposed based on information entropy. The preliminary generation method of VMS layout scheme was design combined with the topological structure characteristics of practical road network. According to drivers’response properties of VMS information, the VMS guidance strategies for congestion management were designed. When the correlation links were all in congestion state, the relative optimal traffic state link discrimination method was proposed based on vague sets sorting. The drivers’route-planning demand of traffic guidance system was analysed, the influence factors of ideal-route are distinguished, and then the attributes and characters of individuation route demand are analyzed in detail. The route-planning methods are proposed based on distance, time, level of service and integration attributes. The integrated individuation route-planning frame is proposed.The content, characters and transmission utility of travel information were studied intensely, and the travel information transmission model and travelers’ information searching model for multi-user were proposed, then the travel information of multi-user demand were analyzed in detail. According to the actual situation of travel information transmission in China, the index system and assessment method were proposed for effectiveness assessment of travel information. Then a travel information transmission index model was proposed. These methods could be applied to evaluate the acceptance level of travel information and transmission quality based on the travelers’and traffic managers’viewpoints, in the hope to lay a theoretic platform for travel information transmission mode programming in regional road network.更多还原