【作者】 蒋洋；

【导师】 张星臣；

【作者基本信息】 北京交通大学 ， 交通运输规划与管理， 2014， 博士

【摘要】 我国处在多种交通方式快速发展时期,综合性交通网络正在得到快速发展和完善,货物运输的选择范围正在扩大。综合交通运输体系是未来我国交通发展的重点方向,是社会经济发展与居民生活水平提高的重要基础,更是物流活动的主要功能要素。如何依托综合性交通网络建立起结构布局合理、管理科学且运转高效的多式联运系统,为物流行业的良性健康发展提供有效保障,将是交通研究者和管理者的努力方向。我国物流业在总体规模快速增长、基础设施迅速扩大的同时,暴露出资源配置与运用不科学、运行过程设计不精细导致的多式联运系统运行效益较低、综合成本较高等亟待解决的问题,运行管理机制及建设理论与方法难以满足物流行业快速发展的需求。基于综合运输体系下的现代物流需以协调决策作为管理目标,在各种有关利益主体和运输方式约束下进行运营和决策。基于此,本文以多式联运系统为研究对象,从多式联运“网络优化”建模方法研究入手,在剖析轴辐式运输网络结构的基础上,重点解决多式联运网络基础设施配置和运输服务设计的集成化问题,从“网络优化”战略规划层面和“运输服务设计”战术规划层面提出强化多式联运系统统筹性、协调性,提高系统运行效率的优化理论与方法。论文的主要研究工作包括如下四个方面的内容：(1)多式联运枢纽与网络布局优化从“网络优化”的建模方法入手,研究建立多式联运多目标枢纽及网络布局优化模型。定义包含多个利益主体的多式联运物理网络和服务网络表述方法,并针对多个利益主体和多集装箱货物品类的特点,提出了多式联运模式下混合用户均衡模型,模型中用以反映用户路径选择的负效用函数包括了运价和网络拥挤因素,并同时考虑了多种运输方式之间的中转时间、费用以及拥挤的影响。鉴于多式联运枢纽与网络布局优化问题的NP-hard特点,设计了嵌入对角化算法的交叉熵求解算法,并采用算例网络对模型和算法进行了验证。分析表明,以增加成本为代价的综合网络运输风险的降低在一定条件下并不明显；此外,多式联运网络管理者对于网络运输风险越加重视,长距离货运量将越集中分布于铁路运输。(2)多式联运网络流量路径优化从“网络优化”的建模方法入手,结合多商品流模型,构建多式联运路径选择多目标非线性优化模型。模型充分考虑了运输区段能力约束和枢纽中转服务能力约束,并通过引入O-1决策变量,使其能够直观地从优化结果中反映出货流的最佳走行路径信息。通过对非线性约束条件进行等价变换,该模型可被转化为整数规划模型,并设计交叉熵算法进行求解。(3)干线运输服务设计与多式联运网络优化综合决策建立资源约束条件下轴辐式多式联运网络优化及干线运输服务设计的混合整数规划模型,分析网络基础设施配置与干线运输服务设计间的协调优化关系,对轴辐式多式联运网络的枢纽等级配置、干线运输服务设计、货物托运人和收货人对于中转枢纽的选择、货物托运人和收货人对干线运输服务的选择以及线路区段改造方案等决策进行综合优化。分析表明,公铁联运中转枢纽节点的能力制约着多式联运系统的服务能力,瓶颈区段的识别并改进以及合理的干线快慢运输服务组合能够大幅度降低网络运营成本。(4)支线运输服务设计与多式联运网络优化综合决策以一体化物流管理为目标,建立网络优化与支线运输服务设计二者综合优化的Ⅱ阶段决策模型。模型第Ⅰ阶段表达为枢纽布局、网络设计及货流分配的0-1整数规划问题,第Ⅱ阶段则在第Ⅰ阶段输出结果的基础上求解带时间窗约束的支线运输服务设计问题；针对模型的Ⅱ阶段结构特点,以两个阶段相互影响相互反馈为求解思路,设计交叉熵为主体的启发式算法。分析表明,模型及其求解算法的操作过程合理有效,模型的实用性得到了验证。最后,论文将该Ⅱ阶段决策模型适当扩展,应用于“单轴”轴辐式大规模算例网络,为运力配置决策提供依据。更多还原

【Abstract】 China is experiencing the period of the rapid developments of its multiple transportation modes. The integrated transportation network in China is rapidly developed and improved, and the alternative freight transport modes are increased. Establishment and improvement of an efficient integrated transport system will be our effort direction for the development multi-mode transport in our country. It is also the important foundation of the sustainable development of the social economy and the improvement of the living standard of the people. Furthermore, it is one of the main functional elements of logistics activities. An integrated transportation network with rational layout, scientific management, and efficient operation is strategically important to promote the development of logistics industry in China.The scale of logistics activities is rapidly developing and the level of infrastructure is improving. Meanwhile, it is accompanied by a series of problems, such as low efficiency and high cost of intermodal transportation system. The problems such as unreasonable resources allocation and plain operation design must be solved. The operation management method and construction mechanism does not fit the requirements of rapid development of logistics activities. Coordinate optimization is the management goal of the modern logistics activities based on integrated transport system. Considering the goals and requirements of intermodal transportation system, the study focuses on network optimization and transportation service design. A comprehensive optimization method is proposed from the perspective of strategic planning and tactical planning. The contents of the study are organized as follows.(1) Intermodal transportation hub location and network design The study is proceeding from the perspective of strategic planning. The paper proposes a bi-objective programming formulation for the intermodal transportation network design with multiple stakeholders (network manager, carriers, hub operators and intermodal users). The study represents a given intermodal network as the physical and operational networks. The model incorporates a parametric variational inequality (Ⅵ) that formulates the mixed equilibrium (UE&SO) behavior of intermodal users in route choice for any given network design decision. The disutility of a carrier link or transfer perceived by an intermodal user is composed of two portions:the actual rate charged by the carrier on the carrier link or hub operator providing the transfer service and the monetary value of transportation or transfer time that equals the actual handling time multiplied with value of time(VOT). In reality, the second portion reflects the impact of congestion on the intermodal operator’s route choice. A Cross-Entropy (CE) method embedded with diagonalization algorithm is employed to solve the model. Finally, the developed model is applied to a case study to demonstrate its applicability and gain managerial insights to some extent.(2) Intermodal transportation routing optimization The paper studies the intermodal route choice model and algorithm. The study proposes a multimodal hazmat model that simultaneously optimizes the locations of transfer yards and transportation routes. The capacities of links and transfer yards are embedded into the model. A0-1decision variable is defined which could make the optimization results reflect flow routing. However, it is nonlinear and also contains an absolute term, which makes it very difficult to solve. Several new constraints are introduced to replace the nonlinear and the absolute terms. A Cross-Entropy (CE) method is employed to solve the model.(3) Comprehensive optimization between intermodal transportation network optimization and main line transport service design The study presents an optimization framework for planning intermodal shipments based on hub-and-spoke network. The aim of the research is to minimize the total transport cost from the perspective of network optimization and main line transport service design. The decisions are including hub capacity planning, main line transport services designing, the access to alternate intermodal terminals and main line transport service for shippers and receivers and segments reconstruction. A Cross-Entropy based solution methodology is developed. The optimization framework is applied to problem instances to gain managerial insights. Our analysis indicates main line transport services accounts for a significant portion of transport cost.(4) Comprehensive optimization between intermodal transportation network optimization and feeder line transport service design A two-phase programming formulation is proposed for the simultaneous decision of network optimization and feeder line transport service. The hub location and spoke allocation problem for phase I and the feeder line transport service design problem for phase II. The problem in phase I is formulated as a0-1integer programming model. The problem in phase II is formulated as a vehicle routing problem with time window. A Cross-Entropy based solution method is proposed to solve it. Numerical experiment based on a simple network is carried out to be able to provide detailed model and algorithm outputs to better illustrate how the developed model and algorithm work. In the end, the model is applied to desing hub-and-spoke transpoation network with liner shipping for large-scale example.更多还原