【作者】 杨静；

【导师】 毛保华； 吴海燕；

【作者基本信息】 北京交通大学 ， 交通运输规划与管理， 2010， 博士

【摘要】 城市公共交通系统是重要的城市基础设施,是关系国计民生的社会公共事业。优先发展城市公共交通,不仅是缓解城市交通拥堵的有效措施,也是改善城市居住环境,促进城市可持续发展的必然要求。公交汽车作为常规公共交通方式,具有投资少、运行灵活、覆盖面广等特点,在未来相当长一段时间内仍是城市公共交通的主体。公交车站是城市公交网络中最重要的交通基础设施之一,也是最基本的交通瓶颈所在,是各种交通问题集中的地方。对公交车站的车辆行为进行深入分析,并在此基础上建立符合我国国情的车站能力计算方法体系,不仅具有非常重要的理论意义,也是十分迫切的现实需求。大城市中心区的公交车站具有停靠线路多、车辆运行准时性差、停靠时间波动大等特点。美国道路通行能力手册推荐的车站能力计算方法,没有充分考虑到公交车辆到达的随机性,对停靠的波动性处理又过于理想化；有些学者采用的随机服务系统方法与公交车辆排队进出车站的实际情况有差异；多停靠位的车站能力计算方法一般采用简单的累计有效停靠位来体现停靠位的不完全利用问题,难以体现不同车站不同运行状态的差异及车站内公交车辆的相互影响。有鉴于此,本文根据我国城市地区公交车站的实际运行情况,考虑了公交车辆到达以及停靠时间的随机分布问题,构建了基于服务水平的车站能力计算模型,并扩展到多停靠位车站。主要工作和结论包括：(1)在分析国内外公交车站能力研究现状的基础上,探讨了理论计算方法、随机服务系统方法、饱和车头时距方法、统计回归方法及仿真等方法的计算原理,总结了各种方法的优缺点和适用情形,指出现有研究存在的问题,并分析了公交车站能力“量”与“质”的双重属性。(2)从城市客流分布以及土地资源利用的角度,分析了大城市中心区车站设置的特征和发展趋势,多线路共用站已成为普遍采用的车站形式。以北京地区的公交车站为实例,研究了多线路共用站在实际应用中存在的问题,指出能力估计过高导致的车站排队是最普遍,也是最根本的问题。从车站形式及内部设计、公交车辆外形与性能、乘客乘降及候车、车站周边交通环境四方面,分析了车站能力的影响因素。(3)构建了基于服务水平、考虑公交车辆到达分布和停靠时间分布的单停靠位公交车站能力计算模型。调查数据表明,单位时间到达车站的公交车辆数量服从泊松分布,公交车辆的停靠时间服从负指数分布或k阶爱尔朗分布(k为自然数),根据k值的不同,停靠时间呈现不同的分布状态,k值越大,分布的随机性越小。选取排队概率作为服务水平的衡量指标,在到达服从泊松分布的基础上,分别建立了停靠时间服从负指数分布和k阶爱尔朗分布的单停靠位车站能力计算模型。通过算例分析,得到车站能力随设计服务水平的升高而降低,随停靠时间的随机性降低而升高的主要结论,并且k≤15,特别是k≤5的范围内,k值对车站能力的影响最为显著。(4)以单停靠位车站能力计算模型为基础,通过研究公交车辆在车站范围内的微观行为,分别建立了出站排队时间模型、空置时间模型、运营裕量模型,共同构成两停靠位车站能力的计算模型。通过算例分析了车站能力各个计算因素在不同的停靠时间分布随机程度下的变化情况,运营裕量和车站能力与排队概率的整体关系,不同的服务水平对应的运营裕量ts与车站能力Cs的具体变化情况。(5)对三停靠位公交车站的能力计算模型进行了探索性研究,根据事件发生的情形和概率,构建不同情形下的出站排队时间计算模型,进而得到平均出站排队时间的计算模型。通过算例得到3号停靠位公交车辆的出站排队时间tcp3随交通强度的增加而增加,且随平均停靠时间的升高而整体水平上升的结论。更多还原

【Abstract】 Public transport system is the basic infrastructure in city, and also the social commonweal concerning the national economy and the people’s livelihood. Transit priority policies are not only the effective measures that release traffic jam, but also the demand of city’s sustainable development. The traditional transit mode (in fact, bus) is occupying the main part of transport in a long time for the advantage of low energy and land cost, flexible operation and large coverage of network. As the nodes of transit network, bus stops are the basic facilities of bus service, also the bottleneck of traffic where a lot of traffic problems happen. Therefore, it has great theoretical significance and real world demand to built capacity calculation method basing on the real situation in China after an in-depth analysis on the vehicle behavior around and inside bus stops.Bus stops in urban area of big cities have several vivid characteristics, which include large number of bus lines, poor punctuality, and strong fluctuation of dwell time. The classic bus stop capacity calculation formula presented by HCM (Highway Capacity Manual, HCM) didn’t fully considering the bus arriving and dwelling random, for this, some researcher used the stochastic service method to calculate bus stop capacity. However, the actual bus in-and-out rule can not be followed in stochastic service method. Moreover, for multi-berth bus stop, the existing models only introduce an effective berth number to show the effect of incomplete use of the berths in the capacity calculation formula, but the real situation is complicated and varied among different bus stops and different periods in the same stop, so the effective berth number is too simple and inflexible to describe the variation of capacity. Considering the disadvantage of existing methods, this dissertation develops a novel method for bus stop capacity calculation that based on the recognition and discrimination of random bus arriving time, random dwelling time and in-and-out-of-berth rule which are the facts in the operation of bus stop in urban area of big cities. Main works of this dissertation are summarized as below:(1) Introduction on the previous studies of the bus stop capacity calculation methods and discussion on the advantage and disadvantage of these methods, which include theoretical calculation method, stochastic service method, saturated headway method, statistical regression method, and simulation method. As a production, the unsuitability of these methods on the capacity calculation of bus stops in urban area is summarized and the quantity and quality of bus stop capacity are analyzed.(2) Analysis on the characteristics and trends of bus stops setting in urban area of big cities, from aspects of passenger flow distribution and land source using. As a conclusion of the analysis, the multi-lines design (there are more than one line that dwelling on the bus stop) is widely used and will be the trend in future. The practical problems happening in multi-lines stop are figured out from the real cases in the city of Beijing, and the overestimation on capacity is found to be the most fundamental and common problem. Moreover, this dissertation is analyzing the influencing factors of capacity from four aspects, including bus stop setting and interior design, bus properties and operation, passenger loading-and-unloading and arriving characteristics, and traffic surrounding.(3) Development of a novel method for capacity analysis and estimation, based on the bus arrival and dwell time distribution, as well as the designed service level. As known from the observed data, bus arrival in unit time follows Poisson distribution, and dwell time follows either negative exponential distribution or k-order Erlang distribution. The order k presents the random level of dwell time, and the bigger is k, the lower is the random level. While queue rate is selected as the measure index, and arrival distribution is set to be Poisson, the calculation models of bus stop capacity is developed separately for dwell time following negative exponential distribution and k-order Erlang distribution. As the conclusion of case study, the capacity is decreasing with the elevation of designed service level and dwell time random, furthermore when in the range of k≦ 15 especially k≦ 5, k has great effect on the capacity.(4) Development of a novel method on the two-berth bus stop capacity calculation. Deriving from the method proposed in anterior chapter, three sub-models, including queuing-to-out time sub-model, idle time sub-model and operating margin sub-model, have been established to constitute the capacity calculation model, basing on analysis about the microscopic behaviors of bus in the bus stop. From the results of case study, the effects of dwell time random level on the capacity calculation factors, the general relation between operating margin as well as capacity and queue rate, and the trend of operating margin and capacity for different service level, are analyzed.(5) Exploratory research on the three-berth bus stop capacity calculation. After situation and possibility analysis, this chapter models the queue-to-out time under different condition, then gets the average queue time model. As the conclusion of case study, the average queuing time is increasing with the elevation of traffic density and the average dwell time.更多还原