【作者】 汤中明；

【导师】 刘志学；

【作者基本信息】 华中科技大学 ， 管理科学与工程， 2010， 博士

【摘要】 当今世界,全球化经济竞争不断加剧,供应链管理理论不断发展,使得VMI越来越受到企业和学者的关注与重视。从理论研究和实际应用来看,VMI通常能够给下游企业带来较多好处,但对于供应商而言,由于受到其物流能力的制约,VMI的作用得不到充分发挥。VMI与TPL的有效集成能够充分发挥供应商与TPL各自的优势,有效改善整个供应链的物流服务水平。鉴于此,本文在探讨VMI & TPL供应链管理模式的基础上,对该模式下的库存运输集成优化问题进行深入研究,达到了预期研究目标。首先,阐述了VMI和TPL国内外研究现状,对VMI与TPL集成供应链进行了研究。在系统分析VMI与TPL集成推动力的基础上,提出了分散协调型和集中控制型两种不同集成程度的VMI & TPL供应链运作模式,并重点研究了VMI与TPL集成程度对供应链收益的影响。结果表明：VMI与TPL集成程度对供应链整体及TPL的收益没有绝对的影响,而对供应商和零售商的收益有正向影响。其次,从系统论角度界定了库存运输集成的内涵,构建了VMI & TPL模式下库存运输集成过程与战略决策模型。在界定库存运输集成内涵的基础上,对VMI & TPL模式下库存运输集成过程进行分析,提出了VMI与TPL集成要解决的核心问题,构建了分散协调型和集中控制型这两种VMI & TPL模式下库存运输集成的战略决策模型,并对其战略决策结果进行比较研究。结果表明：VMI与TPL集成程度越高,集成运作所需要的库存与运输资源越少。再次,建立了分散协调型VMI & TPL模式下供应商补货决策模型。论文以“多供应商、单TPL和单制造商”供应链物流网络为基础,引入供应链管理环境下的信息共享机制,同时考虑供应商与TPL之间的运输费用契约,从供应商角度构建了分散协调型VMI & TPL模式下供应商补货决策模型,并设计了有效求解该模型的模拟退火遗传算法。最后,构建了集中控制型VMI & TPL模式下库存运输集成仇化模型。论文以“多供应商、单TPL和单制造商”供应链物流网络为基础,从TPL角度出发探讨了集中控制型VMI & TPL模式下的库存运输集成问题,建立了基于滚动计划期的库存运输集成优化模型,并证明了该模型问题是一个NPC问题。论文对模型进行了分解,并提出了求解该模型的基于Lingo9.0和模拟退火遗传算法的启发式算法。更多还原

【Abstract】 With the increasingly intensified competition in the emerging global economy and the development of supply chain management theory, vendor-managed inventory (VMI) is concerned more and more by enterprises and scholars. According to the theory research and application, VMI can always bring benefits to down-stream enterprises. But for vendors, the effect of VMI can’t be ensured due to the constraint of logistic capability. By integrating VMI with third-party logistics (TPL), the superiority of both vendors and TPL can be realized which should improve the supply chain logistics service level. Therefore, this dissertation is aimed to research the integrated optimization of inventory and transportation in VMI&TPL, which is based on analyzing the integrated supply chain of VMI and TPL.Firstly, researching status of VMI and TPL is expounded and integrated supply chain of VMI and TPL is analyzed. Based on analyzing the driving forces of integrating VMI with TPL, two kinds of integrated models, which are decentralized coordinating and centralized controlling supply chain, are proposed and analyzed comparatively. At the same time, the influence of integrated extent on benefits of VMI&TPL supply chain is analyzed in detail. As it turned out, integrated extent of VMI and TPL has no absolute influence on the benefits of supply chain and TPL. However, it has positive influence on the benefits of vendors and retailers.Secondly, the connotation of integrating inventory with transportation is proposed from systemic perspective. The integrated course and the key questions to be resolved in the course are discussed. Strategy decision problem, the first of the key questions, is also researched. The strategy decision models of integrating inventory with transportation under decentralized coordinating and centralized controlling situation are built separately. The influence of integrated extent on strategy decision is analyzed comparatively. The conclusion indicates that integrated extent has positive impact on strategy decision.Thirdly, the vendor’s replenishment-decision model under decentralized coordinating VMI&TPL supply chain is built. The model considers supply chain logistics network which contains many vendors, single TPL and single manufacturer. Information sharing mechanism in supply chain managent and freight constract between vendors and TPL are inducted into the model. According to the models characters, correspondent simulated annealing and genetic algorithm is designed, which is proved to be effective by comparative analysis with basic genetic algorithm.At last, the optimization model of integrating inventory with transportation under centralized controlling VMI&TPL supply chain is built, which considers the same network as the model under decentralized coordinating VMI&TPL supply chain. Because of strongly computational complexity of the problem, the dissertation proves it to be a NPC problem and proposes a heuristic algorithm to solve it which is based on Lingo9.0 and simulated annealing and genetic algorithm.更多还原