生产系统的柔性设计：供应链与生产车间

发布时间：2018-03-22 17:57

本文选题：过程柔性　切入点：供应链柔性　出处：《清华大学》2016年博士论文　论文类型：学位论文

【摘要】：随着经济的发展,消费者对产品或者服务的要求更加个性化和快速化,市场产生了更多不确定因素。柔性是应对这些不确定的有力工具。本文的研究从规划和运营两个层面,分别对供应链和车间层级的柔性设计和性能分析进行了研究。其中,在规划层面,本文从更贴近实际的单级、单周期,到更符合现实复杂度的多级、多周期,对供应链柔性设计问题进行了研究,分别提出了二阶段随机规划模型和动态规划模型。在生产车间运营操作层面,本文考虑发生干扰后的柔性重调度问题,并引入了人类行为观点作为优化目标,提出了混合整数规划模型。单级、单周期供应链的过程柔性设计问题广泛存在于现实中。与现有文献中各种特定假设不同,本文的假设不要求系统是否平衡、是否对称、成本参数是否同质等。这些一般化的假设更加贴近实际环境,因此本文研究具有一定的实用意义。本文使用二阶段随机规划方法来求解过程柔性设计问题,设计了基于Benders分解的L型算法,并且讨论了几种有效的算法改进措施。对比实验结果显示,相对于文献中的启发式设计方法,本文提出的方法适用范围更广。本文的方法同样适用于同时考虑产能投资和柔性设计的问题。针对多级、多周期供应链环境的柔性设计问题,本文建立了动态规划模型。通过分析模型性质,证明了该模型可以基于级和产品分解成简单子问题。该结果扩展了带产能约束的供应链库存管理相关研究。基于这些简单子问题,本文进一步证明了下级节点的最优补货策略为基本库存策略,从而简化了复杂模型的求解。在此基础上,本文分析研究了多级、多周期等供应链特性对长链条柔性性能的影响,发现长链条柔性在多级、多周期供应链环境下的性能仍然非常接近全柔性。在生产车间运营层面,本文考虑了生产过程出现中断情况下,如何通过重调度保证工厂生产柔性的问题。本文考虑人类行为观点,优化因重调度带来的损失给管理者引起的损失感觉。本文的优化目标与传统重调度研究的优化目标不同,通过对比两种模型在两种不同重调度策略(尽早策略和尽晚策略)的结果,发现本文的模型更加倾向于尽早处理干扰,这有利于尽快消除干扰对供应链柔性产生的可能影响。本文上述研究,不仅为供应链柔性设计和分析问题的文献做出了有益补充,也对实际供应链管理提供了一定的启示。
[Abstract]:With the development of economy, consumers' demand for products or services becomes more personalized and rapid, and the market produces more uncertainties. Flexibility is a powerful tool to deal with these uncertainties. The flexibility design and performance analysis of supply chain and shop floor are studied respectively. In this paper, the flexible design of supply chain is studied, and two stage stochastic programming model and dynamic programming model are put forward respectively. In the production shop operation level, the flexible rescheduling problem after interference is considered in this paper. By introducing the viewpoint of human behavior as the optimization goal, a mixed integer programming model is proposed. The process flexibility design of single-stage, single-cycle supply chain is widely used in reality. The assumptions in this paper do not require whether the system is balanced, symmetric or not, whether the cost parameters are homogeneous, and so on. Therefore, the study of this paper has some practical significance. In this paper, two-stage stochastic programming method is used to solve the flexible process design problem, and an L-type algorithm based on Benders decomposition is designed. In addition, several effective algorithms are discussed. Compared with the heuristic design method in literature, the experimental results show that, The method proposed in this paper is also applicable to the problem of both capacity investment and flexible design, aiming at the flexible design of multi-stage and multi-cycle supply chain environment. In this paper, a dynamic programming model is established. It is proved that the model can be decomposed into simple subproblems based on order and product. The results extend the research of supply chain inventory management with capacity constraints. This paper further proves that the optimal replenishment strategy of lower level nodes is the basic inventory strategy, which simplifies the solution of complex model. On the basis of this, this paper analyzes and studies the effects of supply chain characteristics such as multistage and multi-period on the flexible performance of long chain. It is found that the performance of long chain flexibility in multi-stage and multi-cycle supply chain environment is still very close to full flexibility. In this paper, we consider the viewpoint of human behavior and optimize the sense of loss caused by the loss caused by rescheduling. The optimization goal of this paper is different from that of traditional rescheduling research. By comparing the results of the two models in two different rescheduling policies (early policy and late policy), it is found that the model in this paper is more inclined to deal with interference as early as possible. This will help to eliminate the possible impact of interference on supply chain flexibility as soon as possible. This study not only provides a useful supplement to the literature on the design and analysis of supply chain flexibility, but also provides some enlightenment to the actual supply chain management.
【学位授予单位】：清华大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：F274

【相似文献】