【作者】 张毅；

【导师】 郭晓汾；

【作者基本信息】 长安大学 ， 载运工具运用工程， 2008， 博士

【摘要】 面对日益严峻的灾害形势，做好救灾物资各个环节的工作至关重要。因此，系统研究救灾物资的决策理论与方法是十分必要的。阐明了救灾物流网络的构成要素、基本特征、规划的概念与意义、分类、依据、目标与原则及内容与步骤，基于轴辐式网络结构，提出了救灾物流网络节点规划的新思路，构建了多枢纽站联合分派的救灾应急物流网络结构。根据救灾物资的需求特征，基于事例推理理论，运用人工智能技术，提出了救灾物资需求即时估测的新方法。通过权衡多目标之间的关系，构造了多品种救灾物资应急采购的优化模型，从而可以合理确定供应商组合，使得质量、价格、交货准时性等指标均达到一定水平的满意程度。运用自然灾害救助区划研究的成果，以县域为统计单元，计算分析了陆地交通通行能力和通达时间两个应急响应能力指标，探讨了国家救灾物资储备库的宏观布局问题。基于轴辅式网络结构，建立了无容量约束的救灾物流网络枢纽点定位模型，采用专家选择法和启发式算法相结合的方法，对模型进行了算法设计和求解，满足了枢纽定位问题的时效性和经济性兼顾的要求。基于灰色白化权函数聚类理论，提出了救灾物资库存分类的新方法。基于时效性目标，研究了无限供给率和有限供给率的确定性库存及随机性需求的确定性和随机性订货提前期的库存问题，给出了最佳订货量、再订货点、最大库存量和盘存周期等库存控制参数计算公式，以确定不同类型救灾物资的库存策略。基于轴辐式救灾物流网络结构，提出了救灾物资配置决策新方法，建立了配送车辆路线决策优化模型，并对模型进行了遗传算法设计和求解，得到了既满足受灾点时限要求、又能使配送成本最低的车辆路线优化方案。基于时效性、安全性和经济性属性，建立了评价救灾物资车辆备选线路的数学模型，确立了选择具有最高期望效用方案的方法。基于灾后因道路遭受破坏而影响救灾物资配送的考虑，利用时空网络流动技巧，建立了多目标混合整数规划模型，采用分解启发式求解方法，得到了实际救灾工作中可以接受并提高了求解效率的计算结果。更多还原

【Abstract】 It is important to deal well with every aspect of the disaster relief materials with theaustere disaster situation.So it is very necessary to study systematacially the decision-makingtheory and method of disaster relief materials.The composing elements,characters,planning concepts and meanings,classification,basis,goals and principles,contents and steps of the disaster relief logistic network wereclarified;based on hub-and-spoke network topology,a new idea of the disaster relief logisticnetwork node program was put forward;and the disaster relief logistic network structure ofthe multi-hub combining allocation was constructed in the paper.According to the demand characteristic of disaster relief materials,based on deducingexample,and using the artificial intelligence technology,a new method to estimate instantdemand of disaster relief materials was put forward;and an optimal model of emergencypurchage of many disaster relief materials was constructed through balancing the relation ofmulti-objective,which could confirm reasonable vendor combination and make the indexessuch as quality,price,consignment time be satisfied.Using the study result of natural disasters relief regions and making a county as astatistic unit,two emergency response indexes,ground traffic capacity and arriving time werecalcaluted and analyzed;and the overall layout of the national disaster relief material hub wasdiscussed.Based on the hub-and-spoke network structure,an uncapacitated hub locationmodel was founded,which was designed and calculated with the method of expert choice andheuristics algorithm.The result indicated the the model could satisfy the request of both thetime-effective and economical objective of hub location question.Based on grey whitenization weight function clustering,a new method about disasterrelief materials inventory sorting was proposed.Directing against the time-effective objective,the certain inventory problem of the infinitude supply ratio and the finitude supply ratio andthe random demand inventory problem of the determined and random ordering lead-timeswere studied;and a calculation formula of inventory control parameters,such as the bestordering quantity,reorder point,maximal inventory quantity,etc,was given to confirm theinventory strategies of different kinds of disaster relief materials. Based on hub-and-spoke network,a new decision making method about allocatingdisaster relief materials was put forward;a vehicle routing optimization model wasestablished;and through disigning and calculating the model with Genetic Algorithms,anoptimal scheme was obtained,which could both satisfy time limit request of the disasterdepots and optimize vehicle routing of the minimum delivery cost.Directing against thetimely,security and economy requests,a math model was constructed,which could evaluatethe prepared routing of the disaster relief material vehicles;and a method to choose thehighest expectation utility scheme was determined.Considering the influence of road destroyafter disaster on the materials distribution,a multi-objective mixed integer programmingmodel was founded by using flowing technique of time-space network;and the acceptedcalculation results in actual relief work were obtained with the decomposition heuristicalgorithm.更多还原