【作者】 徐光灿；

【导师】 许茂增；

【作者基本信息】 重庆交通大学 ， 物流与供应链管理， 2021， 博士

【摘要】 交通能源供应网络是交通运输系统的重要组成部分,其为保障交通运行和城市运转提供了坚实基础。其中,成品油供给在交通能源供应网络中占主导地位,从某种程度上说,成品油供给系统的运作效率决定了交通运输系统的运行效率。当前,成品油供给系统存在资源配置不合理、运作效率低下、服务质量亟待提升等问题,政府相关部门和成品油经营企业急需找到优化成品油供给系统运作、提升成品油供给网络服务质量的方法和路径。因此,本论文的选题具有重要的理论意义和实践意义。成品油供给系统可分为成品油二次配送和成品油终端服务两个阶段,这两个阶段既有相对的独立性又有密切的关联。当前对成品油供给系统相关问题的研究多集中在单一油库向加油站配送成品油及单一加油站服务优化等问题,而缺乏从区域成品油供给资源配置和加油站服务网络优化等方面进行研究。为此本文试图从区域成品油供给系统的角度出发,研究多油库供给网络中的合作配送优化问题,多油库合作配送车辆路径资源配置优化问题,多油库合作配送的合作机制与收益分配问题和加油站服务网络优化问题,进而为区域成品油供给系统资源配置优化和管理决策提供理论和技术支撑。本文的主要研究工作和成果如下:(1)分析研究多油库成品油供给共同配送网络优化问题。首先根据聚类算法把区域内的加油站进行了重新划分,从而重新界定了每个油库服务的加油站集合;然后,以区域多油库成品油供给网络总运作成本最小为目标函数,建立优化数学模型;再次,结合GA和PSO算法各自的优点,提出GA-PSO混合启发式算法来求解所提优化数学模型;最后,以中国重庆的区域成品油供给系统为例,对比分析了优化前后的系统运作总成本、配送车辆使用数等指标,验证了上述优化模型及其算法的有效性,从而为多油库成品油合作共同配送的网络优化问题提供了一种有效的研究方法。(2)研究多油库成品油共同配送车辆路径问题的资源优化配置。同样,首先对不同油库所负责配送的加油站进行了重新的聚类,即每个油库形成了新的客户群;然后,基于多舱车辆共享、时间窗协调和车辆路径优化等机制,以系统运作总成本最低和使用配送车辆数最少为目标函数,建立了多油库成品油共同配送车辆路径的双目标混合整数规划模型;接着设计了考虑遗传变异的多目标粒子群算法来求解所提优化数学模型;最后,通过一个小规模案例和一个大型实例来验证所提优化模型及算法在解决多油库成品油共同配送车辆路径问题时的有效性和合理性。(3)研究多油库成品油供给共同配送网络中的合作机制与收益分配问题。首先介绍了常用的几种收益分配方法及严格单调路径策略;接着,提出了考虑参与合作主体业务量、投资额和风险量等因素的Shapley值修正模型来决定成品油共同配送中的收益分配;最后,以多油库成品油供给共同配送网络优化问题为基础,应用Shapley值方法和SMP策略来分析多油库成品油合作供给系统的利益分配和各油库的最佳入盟序列。成品油供给系统多主体合作机制与收益分配方法为区域合作联盟的形成和联盟的稳定性提供了理论支撑,其也是多油库成品油合作供给在实践中应用的前提和基础。(4)研究加油站网络补能车辆排队优化。首先,介绍排队理论,分析排队的基本特征及在相关领域的应用;然后,分有信息引导和无信息引导两种情况,将排队论用于分别由两个M/M/1/2、M/M/1/3、M/M/2/3组成的典型加油站布局系统的补能车辆排队分析,并从数学上严格证明了两种情况下排队系统的不同性态指标;最后,通过数值计算,直观比较有信息引导和无信息引导两种状况下相关排队指标值变化的情况,并将排队系统扩展到两个M/M/1/4、M/M/2/4组成的加油站服务网络,通过数值模拟,验证结果的一致性。上述研究一方面为政府相关部门和成品油经营企业优化交通能源供应网络运作效率,提升成品油供给服务质量提供有益参考建议,另一方面进一步丰富和发展了交通能源供应链相关理论与方法体系,具有重要的理论和实践意义。更多还原

【Abstract】 The traffic energy supply network is an important part of the traffic system,which provides a solid foundation for the traffic and city operations.Among them,the supply of refined oil plays a dominant role in the traffic energy supply system.To some extent,the operating efficiency of the refined oil supply system determines the operating efficiency of the transportation system.At present,there are some problems in the refined oil supply system,such as unreasonable resource allocation,low operation efficiency and urgent need to improve the service quality.The relevant government departments and refined oil enterprises are badly in need of finding ways and paths to optimize the operation of refined oil supply system and improve the service quality of the supply network.Therefore,the topic selection of this paper has important theoretical and practical significance.The refined oil supply system can be divided into two stages,the secondary distribution and the terminal service of refined oil,which have relative independence and close correlation.At present,the research on the related problems of refined oil supply system mainly focuses on the distribution from a single depot to gas stations and the optimization of service of a single gas station,but lacks the research on the allocation of regional supply resources and the optimization of service network of gas stations.Therefore,this dissertation aims to investigate the cooperative distribution design optimization with multi depots,the resource allocation optimization of vehicle routing for cooperative distribution of multi-depot,the cooperative mechanism and profit distribution of multi-depot cooperative distribution and the optimization of service network of gas stations from the perspective of regional refined oil supply system.Furthermore,it provides theoretical and technical support for resource allocation optimization and management decision of regional refined oil supply system.All achievements of the dissertation are summarized as follows:(1)The optimization of the joint distribution network for the supply of refined oil from multiple oil depots is analyzed.Firstly,the gas stations in the region are redivided according to the clustering algorithm,so as to redefine the set of gas stations served by each oil depot.Then,an optimized mathematical programming model was established with the objective function of minimizing the total operating cost of the regional multi-depot refined oil supply network.Thirdly,combining the advantages of Genetic Algorithm(GA)and Particle Swarm Optimization(PSO)algorithms,a GA-PSO hybrid heuristic algorithm is proposed to solve the proposed optimization mathematical model.Finally,taking the regional refined oil supply system in Chongqing,China as an example,the total operating cost of the system and the number of distribution vehicles before and after optimization are compared and analyzed,and some practical problems are illustrated,thus it provides an effective research method for the network optimization of the cooperative distribution of refined oil products in multiple oil depots.(2)The optimal resource allocation of vehicle routing problem for refined oil distribution in multiple depots is studied.Similarly,the gas stations distributed by different oil depots are reclustered,that is,each oil depot formed a new customer group.Then,based on the mechanisms of multi-compartment vehicle sharing,time window coordination and vehicle routing optimization,a dual-objective mixed integer programming model for vehicle routing of multi-depot refined oil distribution was established,with the lowest total cost of system operation and the least number of vehicles used for distribution as objective functions.Secondly,a multi-objective PSO algorithm considering genetic variation is designed to solve the proposed mathematical model.Finally,through a small case and a large example,the effectiveness and rationality of the proposed optimization model and algorithm in solving the problem of vehicle routing for the joint distribution of refined oil in multiple depots are verified.(3)The cooperation mechanism and profit allocation in the joint distribution network of refined oil supply in multiple oil depots are studied.Firstly,several commonly used profit distribution methods and strict monotone path(SMP)strategy are introduced.Then,a modified Shapley value model considering the business volume,investment amount and risk amount of the participants is proposed to determine the profit distribution in the joint distribution of refined oil.Finally,the Shapley value method and SMP strategy are used to analyze the profit distribution of the multi-depot refined oil supply system and the optimal entry sequence of each depot based on the optimization problem of the multi-depot refined oil supply joint distribution network.The multi-agent cooperation mechanism and profit distribution method of refined oil supply system provide theoretical support for the formation and stability of regional cooperation alliance,which is also the premise and foundation for the application of multi-oil depot refined oil cooperation supply in practice.(4)This paper studies the queuing optimization of fuel replenishing vehicles in gas station network.Firstly,the theory of queuing is introduced,and the basic characteristics of queuing and its application in related fields are analyzed.Then,the queuing theory is applied to the queuing analysis of a typical gas station layout system composed of two M/M/1/2,two M/M/1/3 and two M/M/2/3,and the different state indicators of the queuing system under the two conditions are proved strictly mathemically.Finally,through numerical calculation,the changes of relevant queuing index values under two conditions of information guidance and no information guidance are compared intuitively,and the queuing system is extended to two gas station service networks composed of two M/M/1/4 and two M/M/2/4.Through numerical simulation,the consistency of the results is verified.The above research provides useful suggestions for relevant government departments and enterprises to enhance the supply efficiency of transportation energy network and improve the service quality of refined oil supply,and it further enriches and develops the related theories and methods system of traffic energy supply chain,which has important theoretical and practical significance.更多还原