【作者】 杨信丰；

【导师】 李引珍；

【作者基本信息】 兰州交通大学 ， 交通信息工程及控制， 2010， 博士

【摘要】 城市化进程的加快,城市人口、机动车数量的急剧增加,使交通需求逐渐增大,城市道路交通的压力越来越大,由此带来的交通拥堵势必成为我们必须要面对和解决的难题。城市道路交通网络是交通运输的基础物理设施,其服务水平关系到社会可持续发展、居民生活质量及城市的形象,因而,深入研究城市道路交通网络总体容量对发展我国交通事业,构建和谐畅通的城市交通无疑具有十分重大的意义。本文对已有相关原理与方法进行分析总结的基础上,从影响交通网络容量的主要因素出发,针对交通网络容量概念不一致的现象,结合论文的研究目的提出了基于服务水平的交通网络容量的定义,设定了判定交通网络服务水平的指标及划分等级标准,为交通网络容量的度量奠定了基础。依据交通网络中交通个体的路径选择行为对交通网络容量的影响及各种交通分配模型的特点,论文首先假设每个道路使用者均可准确掌握各个路段及交叉口的实时交通信息,考虑交叉口对交通网络容量的影响,建立了基于用户均衡的交通网络容量模型。通过对已有算法及模型特点的分析,利用Kuhn-Tucker定理,得到了模型的一阶最优必要条件,依据此条件提出容量计算的收敛定理,设计了求解用户均衡交通网络容量模型的算法。通过算例分析证实了此容量算法比已有容量算法缩减了计算过程,减少了最短路径的计算次数,提高了算法的计算效率。当交通动态信息系统不健全时,道路使用者只能根据自己对路径的估计判断进行路径选择,因此,本文建立了基于随机用户均衡的交通网络容量模型。通过对随机用户均衡分配方法的分析,指出了两类常用随机用户均衡分配算法的缺陷；根据一条路径对其它路径所承担交通量的影响程度提出路径影响度概念,依据路径影响度概念重新定义了有效路径集合,并设计了有效路径集合的路径影响度参数标定方法。为增加初始阶段有效路径集合中路径数量,减少计算过程,论文设计了求解初始有效路径集合的遗传算法,并以此设计了基于有效路径集合的求解随机用户均衡交通网络容量模型的算法。论文通过算例对有效路径集合的影响度参数进行了分析,并对基于用户均衡的交通网络容量及基于随机用户均衡的交通网络容量进行了对比分析。实际交通网络中的路段及交叉口的通行能力不会保持不变,而是在一定的范围内波动,于是在路段及交叉口的通行能力不确定情况下,对交通网络容量的波动规律及波动范围的研究具有重要实际应用价值。论文根据道路使用者的长期路径使用情况,利用不确定环境下的均衡状态表达式作为不确定环境下的路径选择原则；依据此原则分别建立了不确定环境下的用户均衡交通网络容量模型及随机用户均衡交通网络容量模型。根据不确定变量不同的处理方法,分别设计了基于等价转化的模型求解算法和基于模拟的模型求解算法；并对不确定环境下的交通网络容量和确定环境下的交通网络容量进行了分析对比。对于已建成的城市来说,随着城市人口、机动车数量的增加,城市的交通需求会不断增大,由此出现交通网络的容量不能适应经济和社会发展的情况：为了使交通网络容量满足交通需求,需要对交通网络中路段或者交叉口进行能力扩建。因此,论文在提出路段及交叉口贡献度概念的基础上,建立了一定费用下基于贡献度的交通网络容量扩张模型,并设计了该模型的求解算法。在日常的城市交通活动中,经常会出现由于短时间交通管制或偶发事故等原因导致的路段或交叉口中断,这给城市交通带来较大的影响；而不同的路段或交叉口中断造成的网络容量损失是不同的。针对此问题,论文提出了交通网络关键路段及交叉口的定义,分别建立了交通网络容量的关键路段及关键交叉口求解模型,并设计了相应的求解算法。依据对算例的计算分析得出的结论,可以为城市交通部门制定适宜的政策引导居民出行,缓解交通状况,决定投资方向及制定紧急决策等提供理论依据。更多还原

【Abstract】 Along with rapid developing of urban growth and increasing of urban population and vehicle population, traffic demand and traffic pressure are increasing gradually. Therefore, the traffic jam has become a problem to be faced and solved. Road network is urban physical infrastructure, its service quality is crucial for social sustainable development, quality of life and image of city. So it is very important to research the capacity of urban road network for developing our traffic and constructing harmonious and smooth traffic.On the basis of analyzing and summarizing the existing related principles and methods of road network, according to the inconsistent definition of urban road network capacity, considering the main factors and combining research purpose, the new conception of road network capacity is defined based on the service level. Moreover, the index and division standard are set up for judging the service level of urban road network. This work lays the foundation for measuring urban road network capacity.Assume that each traveler can acquire real-time traffic information accurately. In accordance with the influence of traveler’s route choice behavior and characters of various traffic assignment models, the model of urban road network capacity is established based on user equilibrium (UE) traffic assignment considered the influence of intersection. Furthermore, the first order necessary condition for optimal this model is obtained by use of Kuhn-Tucker theorem and the convergence theorem is put forward which is used for calculating road network capacity in this paper. Simultaneously, the numerical example verifies that this road network capacity algorithm has less calculation process than existing algorithm and reduces calculation times of shortest path. Consequently, the efficiency of road network capacity algorithm is improved greatly.While the dynamic traveler information system is imperfect, traveler chooses route only according to his own estimation. Thereby, the urban road network capacity model based on stochastic user equilibrium (SUE) traffic assignment is established to deal with this situation. By analyzing the major methods of SUE traffic assignment, some disadvantages that affect them to practical application are fount. So, the conception of efficient path set is redefined based on path effect degree which is proposed according to the degree of one path affecting traffic volumes on other paths, and calibration algorithm of path effect degree parameter for efficient path set is design. In order to increase the quantity of elements in efficient path set at initial stage and decrease the calculation process, a genetic algorithm (GA) is designed to find the initial efficient path set. Furthermore, the algorithm of this road network capacity model is put forward based on efficient path set. Besides, effect degree parameter of efficient path set is analyzed and the road network capacity models based on UE traffic assignment and SUE traffic assignment are contrasted and analyzed through a numerical example. In real traffic network, the capacity of link or intersection is not constant, but rather fluctuation within some bounds. Consequently, under this condition, the fluctuation regulation and fluctuating range of road network capacity are apparently necessary to be well-studied. Accordingly, the equilibrium condition expression is brought forward as the principle of path choice under uncertain environment. According to this principle, the road network capacity models based on UE traffic assignment and SUE traffic assignment under uncertain environment are established respectively. Considering different disposal methods of uncertainty variable, two algorithms are designed based on crisp equivalents and computer simulation. In addition, the road network capacity models based on UE traffic assignment and SUE traffic assignment under uncertain environment are contrasted and analyzed.For current city, the increase of urban population and vehicle population results in the increase of traffic demand and traffic pressure. Accordingly, current road network capacity may not adapt the demand of economy and social development. In order to meet the traffic demand, link or intersection in traffic network needs to be extended. So, on the base of putting forward the contribution degree conceptions of link and intersection, the contribution degree based capacity expansion model is set up with fixed cost and its algorithm is also designed.In daily urban traffic activities, traffic broken on link or intersection is common because of short time traffic control or contingency and influences urban traffic significantly. However, different link or intersection broken may cause different loss of road network capacity. Therefore, the definitions of vital link and vital intersection are given to describe the capacity loss. Furthermore, the respective road network capacity models and algorithms of vital link and vital intersection are provided to analyze bottleneck of road network.By extensive analysis of the numerical example, some valuable conclusions are acquired which can provide the theoretical basis for traffic control departments to set suitable policy measures, ease the traffic pressure, guide resident trips, decision investment and draw up emergency decisions, etc.更多还原