【作者】 罗端高；

【导师】 史峰；

【作者基本信息】 中南大学 ， 交通运输规划与管理， 2009， 博士

【摘要】 城市交通网络是由自驾车、出租车和公交车等交通模式组成的混合交通网络。出租车在城市道路车流中占有一定比例,其运营行为既区别于大运量的公共交通的运营行为,也区别于自驾车出行模式。因此,出租车规模和运营行为对城市交通具有很大影响。但是,现有的交通网络均衡分配方法较少考虑城市混合交通网络中出租车运营对城市交通的影响。本文把出租车的运营影响纳入到一般的交通网络均衡分析方法中来,开展全面系统的研究,有助于进一步丰富和发展交通分配理论,为制定合理的交通规划方案及交通管理措施提供理论基础和科学依据。首先研究了把出租车作为多种交通方式之一的交通网络均衡配流问题。在公交优先策略实施的前提下,假定拥有自用车的出行者选择的交通方式有自驾车、出租车及公交,无车的出行者选择的交通方式有出租车及公交。分析了停车收费措施对路网中出行者的交通方式和路径选择行为的影响,运用路网均衡建模理论研究了在对自驾车停车收费、出租车不收费下的混合交通网络均衡配流问题,建立了考虑公交优先的多用户多方式交通网络均衡变分模型。探讨了不考虑公交优先情形下(公交车流与其它车流之间也相互影响),具有多类时间价值的出行者的交通方式和路径选择行为。基于BPR公式构造了混合交通网络的路段旅行时间函数,运用路网均衡建模理论研究了拥挤路网中混合交通网络均衡配流问题,建立了不考虑公交优先的多用户多方式交通网络均衡变分模型。并设计了对角化技术与MSA方法相结合的算法求解该模型。然后,根据出行者对路网不确定性的风险态度,将出行者分为三类:保守型,冒险型和中庸型。分析了降级路网中出租车多类用户的交通方式和路径选择行为,建立了与均衡条件等价的变分不等式模型,设计了求解模型的有效算法。算例表明不同风险态度类型出行者在降级路网中的交通方式和路径选择存在较大差异,而且交通管理措施对不同风险态度出行者效用的影响也不同。接着,考虑停车收费对出租车方式和非出租车方式的差异影响,研究了停车收费影响下的城市出租车网络运营平衡问题,建立了同时考虑两个平衡子问题的组合数学规划模型。其中一个平衡子问题表示为一个组合网络平衡模型,路网中不同方式车流相互影响的需求弹性的载客出租车与非出租车方式车流的路径选择行为及空驶出租车流的平衡分布特征;另一个子问题描述为一组线性和非线性方程,这些方程确保出租车等待时间和乘客等待时间关系,旅客出行需求及出租车供给平衡关系,出租车服务时间守恒约束得到满足。采用基于灵敏度分析的线性搜索算法来求解该模型。模型得出了一定出租车运营规模和停车收费水平下各交通区域的乘客等待时间,为合理确定城市出租车运营规模提供理论依据。最后,基于出租车空驶搜索行为由到目标区域的出行时间和出行需求分布特征共同决定,引入搜索吸引半径的概念,分析了固定需求条件下出租车运营网络的供需平衡关系,建立了需求分布影响下的城市出租车网络平衡模型,并针对模型结构特征设计了模型求解的迭代求解算法。通过实例,计算了不同出租车运营规模下各交通区域的乘客等待时间。该模型可以为城市出租车网络运营规划与管理提供科学的指导和理论参考。更多还原

【Abstract】 Urban transportation network is a mixed traffic network, composed of cars, taxies, buses and so on. The taxi flow takes up a large percentage of the urban road traffic, and its service is quite different from the large-capacity public transportation and the car, so the fleet size and service behavior of taxi have a large influence on urban transportation. However, the existing network equilibrium assignment approach has less considered the influence of the taxi service on urban transportation in the mixed urban traffic network. The paper considers the influence of taxi service into the general equilibrium analysis approach for a comprehensive system research. It is expected to broaden and develop the traffic assignment theory and offer theoretical and scientific basis for establishing reasonable transportation planning and traffic management strategies.Firstly, a traffic network equilibrium assignment problem regarding taxi as a single mode of various traffic modes is studied. With the premise of the public transportation preference strategy implementation, we assume the traveler who has a car can travel by car, taxi or bus, and the other who has no car can travel by taxi or bus. The influence of the parking charge measure to traveler’s traffic mode and route choice behavior is analyzed. A mixed traffic network equilibrium assignment problem is studied under the condition that parking charge measure is carried out for cars but not for taxi, using road network equilibrium model fundamental. A multi-user and multi-mode mixed traffic network equilibrium variational inequality model is established considering the parking charge measure influence. On the condition that public transportation preference strategy (the public transportation vehicles and other vehicles have an interaction) is not considered, the traveler’s choice behaviors of traffic modes and routes under different values of time are discussed. Based on the BPR formula, the road travel time function of the mixed traffic network is constructed. The mixed traffic equilibrium assignment problem in the congestion network is studied using road network equilibrium model fundamental. A multi-user and multi-mode mixed traffic network equilibrium variational inequality model without considering the public transportation preference strategy is established and a hybrid solution algorithm which is based on diagonalization and MSA is designed for solving the model.Secondly, according to the traveler’s uncertainty risk manner to the road network, the traveler can be divided into three species: risk-averse travelers, risk-prone travelers, and risk-neutral. In the degradation road network, the taxi multiclass user traffic mode and the route choice behavior are analyzed, and the variational inequality model equaling with the equilibrium condition is established, then an efficient algorithm is designed for solving the model. The numerical experiment indicates there is a large difference between different risk manner travelers for their choices of the traffic mode and the route in the degradation road network. Moreover, the traffic management measure is also different to the different risk manner traveler effectiveness’s influence.Thirdly, considering the different influence of the parking charge on the taxi manner and the normal manner, the urban taxi network service equilibrium problem under the parking charge influence is studied, and a combinatorial mathematics planning model is established with two equilibrium sub-problems considered simultaneously. One equilibrium sub-problem is described as a combined network equilibrium model, which is used for describing the route choice behavior of occupied taxi and normal traffic and the equilibrium assignment distribution characteristics of the traffic flow of vacant taxi with the interaction of various traffic modes and elastic demands in the road network. The other is described as a set of linear and nonlinear equations, which guaranty that the relation between taxi waiting time and the passenger waiting time, the balance between customer demand and taxi supply, and the taxi service time conservation equation are satisfied. The waiting time of each traffic area with a certain taxi service size and a parking charge level can be obtained by the model. It can offer theoretical basis for ensuring reasonable taxi service size.Finally, based on that empty taxi driving search behavior is decided by the travel time arriving at the target area and the customer demand distribution characteristic together, and introducing the concept of searching attraction radius, supply and demand balance relations of the taxi service network under condition of the fixed customer demand are analyzed, and the urban taxi network equilibrium model with the demand distribution influence is established, and the iterative solution algorithm is designed according to the model structure characteristic. Using numerical experiment, the passenger waiting time in each traffic area with different taxi service size is calculated. The model can provide scientific guidance and academic reference for urban taxi network service planning and management.更多还原