【作者】 孟梦；

【导师】 邵春福；

【作者基本信息】 北京交通大学 ， 交通运输规划与管理， 2013， 博士

【摘要】 随着城市化进程的加快与城市综合交通基础设施的完善,组合出行将成为居民日常出行的主要形式。组合出行模式下,出行者需要通过换乘一种或者多种交通方式完成出行。现有交通流分配研究大多针对单一出行模式的出行,或假设出行者在一次出行中只选择一种交通方式,将交通方式的选择与出行路径的选择分开考虑,不涉及交通方式的换乘,难以描述组合出行模式下的出行行为和交通系统运行情况。因此,深入研究组合出行模式下交通流分配模型与算法,能为交通管理部门制定合理的引导措施提供支持,使多种交通方式协调运营,最大限度的发挥城市综合交通网络效能,以达到缓解道路交通拥堵的目的。本论文立足于多方式交通网络中交通参与者的出行特征,研究组合出行模式下城市交通流分配模型与算法。首先,从研究思路和研究方法等方面分类分析并综合评述国内外现有相关研究,阐述本论文的研究背景条件；其次,着眼于多方式交通网络结构特征,构建多方式交通分配框架,并针对网络结构和路径判定两个关键问题进行详细分析；再次,从静态解析模型、动态解析模型和动态仿真模型三个方面建立组合出行模式下的交通流分配模型,并分别给出模型的求解算法；最后,通过算例和实际路网对所提模型和算法进行验证,分析参数变化和相关管理措施对交通流在综合交通网络分布的影响,为多方式综合交通管理与组织提供理论支持。本论文主要研究成果如下：1、结合多方式交通网络的结构特点,借助超级网络思想和扩展技术,分析了多方式交通网络中既相互独立又彼此关联的特性,构建了多方式交通超级网络。在此基础上,分析了多方式交通网络中交通参与者的出行特性及其影响因素,包括出行时间、出行费用和舒适度等,给出了量化不同影响因素的计算方法,提出了组合出行模式下的广义出行费用函数。2、引用Logit模型描述了出行者出行方式与路径选择的偏好,基于交通平衡理论和变分不等式思想,分析了多方式交通网络平衡条件,构建了组合出行模式下的随机交通平衡模型,采用相继加权平均算法进行了模型求解。通过算例验证了在求解基于Logit的随机用户平衡模型时,相继加权平均算法优于迭代加权算法；通过算例讨论了停车收费、停车站点选择、自行车租赁三种交通管理措施对交通分布的影响,反推出了合理的停车和自行车租赁的收费范围、以及确定最佳停车换乘站点的位置；进一步将模型应用于考虑出行链和降级路网条件下的组合出行分配问题中。3、建立了组合出行模式下动态交通流分配模型,研究了组合出行模式下的出发时间与路径同时选择的随机动态用户平衡问题。从路径建模的角度,引用Logit模型描述交通参与者对出行时间与出行路径选择偏好,分析了考虑出行时间选择的随机动态用户平衡条件,提出了与平衡条件等价的变分不等式模型,采用基于随机动态网络加载的算法求解模型,实现了组合出行模式下的动态交通流分配。通过算例分析参数变化对交通流分布的影响,并探讨了多类用户及突发事件下的出行行为的变化规律。4、设计了组合出行模式下动态交通流仿真运行框架,基于MesoTS中观仿真模型,建立了适于描述组合出行的中观交通仿真器；仿真器分为车辆产生模块、交通单元模块、方式-路径选择模块和车辆运动模块四部分；采用移位负指数分布描述车头时距分布,改进了C-Logit模型计算路径选择概率,基于K短路算法搜索最短路；对OD需求水平、公交车发车频率、停车收费、交通信息服务以及小汽车拥有率五种情况进行了仿真,仿真算例验证了模型的可行性与交通管理措施的有效性。更多还原

【Abstract】 With the acceleration of urbanization and the improvement of the transportation infrastructure, the combined travel mode becomes the main commuting mode in the daily life. The combined travel mode means that travelers reach their destination by using more than one traffic modals. Most of the current researches on the traffic assignment, however, focus on the single-modal travel mode, or assume travelers only use one traffic modal in a trip. The ignorance of the transfer behavior cannot fully describe the real travel behavior and the traffic operating condition. Therefore, it is imperative to further investigate into the traffic assignment model and algorithm with combined modes, in order to provide the theoretic foundation for the multi-modal coordinated operation, as well as help to alleviate traffic congestion.The dissertation focuses on the urban traffic assignment model and algorithm with combined modes, taking into account the travel characteristics in the multi-modal transportation network. Firstly, the existing relevant researches were reviewed, analyzed and summarized from the methods and ideas, which lays foundations for the further research in this dissertation. Secondly, focusing on the structure characteristics of the multi-modal transportation networks, the multi-modal super network is built based on the super network theory and the expanding technique. Thirdly, the static analysis model, dynamic analysis model and the dynamic simulation model are proposed with consideration of the combined travel mode, and the solution algorithms are discussed separately. Finally, the verifications of the proposed models and their algorithms are given through the ideal experiment network and the real transportation network. Also, the influence caused by parameters changes and the management measures are investigated.The major attained in this dissertation are listed as follows:1. Super network theory and the expanding technique are introduced to build the multi-modal transportation networks, which are mutually dependent and interacting. Based on this network, the influence factors that affect the travel choice are analyzed including travel time, travel cost and travel comfort. The generalized travel cost function is given to quantize these influences.2. The choice preferences for the travel mode and the route are modeled based on the Logit model. And the equivalent stochastic traffic assignment model is proposed by using the traffic equilibrium theory and the variational inequality theory. A solution algorithm is designed on the basis of Method of Successive Weight Average (MSWA). The model and the solution algorithm are verified by an experiment network. The results showed that:MSWA method is superior to MSA when solving the Logit-SUE model; the model and the algorithm can contribute to determine the reasonable position of park and ride site; it can also scientifically determine the reasonable range of parking fee cost in park and ride and bicycle rental cost in renting bicycles; moreover, the trip-chain-based traffic assignment and degradable network cases are discussed with corresponding experiments.3. The simultaneous combined departure time and dynamic stochastic user equilibrium assignment (DDSUE) problem is investigated with combined modes. The DDSUE condition is analyzed based on the Logit model from the path formulation perspective. The direct solution algorithm is given based on the dynamic stochastic network loading. The experiment is designed not only to illustrate the model and algorithm, but discuss the sensitivity of the model parameters. Moreover, the multi-users behavior and accident case are also undertaken in the experiment.4. A simulation-based dynamic traffic assignment model for urban multi-modal transportation network is presented with combined models. The mesoscopic simulator consists of a mesoscopic supply simulator based on MesoTS model and a time-depended demand simulator. The mesoscopic supply simulator includes four parts: vehicle generation module, traffic cell module, modal-route choice module and vehicle movement module. The mode choice is simultaneously considered with the route choice based on the improved C-Logit model. The traffic assignment procedure is implemented by a time-dependent shortest path algorithm in which travelers choose their modes and routes based on a range of choice criteria. The model is particularly suited for appraising a variety of transportation management measures, especially for the application of Intelligent Transport Systems (ITS). Five example cases including OD demand level, bus frequency, parking fee, information supply and car ownership are designed to test the proposed simulation model through a medium-scale case study in Beijing Chaoyang District in China. Computational results illustrate excellent performance and the application of the model to analysis of urban multi-modal transportation networks.更多还原