【作者】 王雅楠；

【导师】 胡祥培；

【作者基本信息】 大连理工大学 ， 管理科学与工程， 2008， 硕士

【摘要】 物流配送是企业对顾客服务的最后一个环节,如何为客户提供及时而满意的配送服务,是物流配送运营商必须解决的重要问题。在实际的物流配送过程中,经常会受到某些随机事件的干扰,如客户需求量变动、客户时间窗变动、车辆故障等,使已有的配送计划变得不可行,无法满足客户的需求以致客户满意度大幅下降。如何有效地处理干扰事件,使得对整个物流配送系统产生的扰动最小,已成为物流配送管理中的难点问题。干扰管理(Disruption Management)作为一种实时处理干扰事件的方法论,主要针对经常性干扰事件的处理和管理。它根据各种实际问题和干扰事件的性质,建立相应的优化模型和有效的求解算法,及时生成最优的调整方案。该调整方案既要满足新目标与新约束,又要达到多方满意,使其对物流配送系统的扰动最小。本文运用干扰管理的方法,对物流配送中发生频率极高的客户需求量变动问题展开了研究,主要研究工作如下:(1)在分析物流配送需求量变动的干扰问题的基础上,提出了扰动辨识的方法;从客户,车辆驾驶员和物流配送运营商三个主体的角度研究了干扰事件的扰动因素,建立了相关因素的扰动度量函数;根据不同物流运营商的经营目标,给出了不同的系统客户不满意度的衡量准则。(2)以带有时间窗的可拆分的车辆路径问题(Split Delivery Vehicle RoutingProblem with Time Windows,SDVRPTW)为研究背景,构建了物流配送需求量变动的干扰管理模型。(3)研究并设计了基于禁忌搜索算法(Tabu Search Algorithm)和插入算法(Inserting Algorithm)的混合启发式算法,用于求解物流配送需求量变动的干扰管理模型,实现快速实时生成物流配送需求量变动问题的干扰管理调整方案。(4)用算例验证了物流配送需求量变动的干扰管理模型以及求解算法的有效性。本文构建的需求量变动的干扰管理模型,可以克服传统的优化方法事前性和全局优化的缺陷,有效降低干扰事件给物流配送系统带来的扰动,能够大大减小了客户不满意度和成本的损失,为物流配送系统干扰管理提供新的思路。更多还原

【Abstract】 Distribution is the last task when a company services customers. How to deliver the right products to customers at right time is the most important problem for distribution companies which focus on Milling customer satisfaction. In practice, there are numerous uncertainties such as changes in customer demand, changes in service time windows, vehicle breakdowns and so on which disable the initial distribution plan, leading to decreasing customer satisfactions. The way how to deal with disruption events effectively, which minimize disturbance resulting from disruption happening in distribution system, is the difficult problem in urban distribution management.Disruption management is a methodology that resolves this kind of problems in real-time. It is a multi-objective decision-making method that a new distribution plan should fulfill new goals and new constraints, as well as minimize the deviations from the original plan and achieve multi-party satisfaction when disruption events appear. This paper applies disruption management to demand changing problem which happens frequently in distribution system. The main researches are as follows:(1) On the basis of analyzing customer demand changing problem, the paper proposed the method of disturbance identification, researching on disturbance factors from the three actors of customers, drivers and distribution operators and constructed disturbance measurement function. Different criteria to measure systematic dissatisfactions are investigated according to operational goals in different distribution companies.(2)Based on Split Delivery Vehicle Routing Problem with Time Windows, using the methods of measuring disturbance, a disruption management model for demand changing in distribution system is constructed.(3)Hybrid heuristic combined Tabu Search algorithm and Inserting algorithm to solve model is designed. When a disruption event appears, the initial distribution plan is optimized in real-time.(4)Applying the model and algorithm to a case, the result shows that the proposed model and algorithm are feasible and effective.The presented disruption management model for demand changing in distribution system is proved to be efficient in minimizing the impacts of disruption events. It reduces customer dissatisfactions and the rise of costs.更多还原