随机生产能力和需求环境下跨国公司生产分销计划研究

发布时间：2018-03-07 17:51

本文选题：随机环境　切入点：跨国公司　出处：《西南交通大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着市场全球化和经济一体化趋势的增强,很多企业纷纷在世界各地设立分支机构,从事生产销售活动,通过内部贸易促进产品、原材料、技术服务在子公司间的流动,形成了以跨国公司为主体的国际生产体系。因此,如何进行全球范围内合理的生产、运输及库存安排,从而达到供应链税后利润最大化或成本最低化成为跨国公司的重要议题之一。首先,论文阐述了跨国公司供应链生产分销计划的研究背景、意义,总结了文中使用的方法和思路,并提炼出本文的创新点。其次,介绍了跨国供应链的基本理论,包括内部贸易、转移价格、供应链计划、关键问题以及决策模式,并对现有的文献进行了简单的回顾和总结。随后,本文以不同决策模式下的供应链生产分销计划为主线展开了研究,分别建立了集中决策下的生产分销模型和分散决策模式下的生产分销计划模型。这两种模型拥有相似的结构,包括一个母公司和多个子公司的跨国公司,属于随机生产能力和随机需求下的多计划期、多公司、多产品的二级供应链模型。但这些复杂模型的处理方法却不尽相同。相同之处是：对模型中含有不确定因素——随机生产能力和随机需求的约束的处理方法是将利用机会约束规划原理将这些约束转换为有明确意义的约束；通过变量替换把非线性模型转换为线性模型。不同之处是：分散决策模型通过拉格朗日松弛算法松弛掉企业之间的耦合关联约束,将整个供应链系统的集中决策问题分解为每个企业内部独立的生产分销计划决策问题,由母公司作为协调中心,借助拉格朗日松弛算法不断迭代,最终求出较优解。最后,两个模型都给出了算例分析。在集中决策下,算例分析验证了本文的两个观点：一、在生产能力有限的情况下,为了达到全局利润最大化,优先满足可以带来较大利润回报的市场而使得另外的市场需求得不到满足。二、在特定情况下,牺牲分公司的利益可以使全球利润达到最大化；在分散决策下,算例分析验证了模型的可行性,说明了拉格朗日松弛算法的有效性。
[Abstract]:With the increasing trend of market globalization and economic integration, many enterprises have set up branches around the world, engaged in production and sales activities, and promoted the flow of products, raw materials, and technical services among subsidiaries through internal trade. Forming an international production system with multinational corporations as the main body. Therefore, how to carry out reasonable production, transportation and inventory arrangements on a global scale, It is one of the important issues for MNCs to maximize the after-tax profit or minimize the cost of the supply chain. Firstly, the paper expounds the research background and significance of the supply chain production and distribution plan of multinational corporations, summarizes the methods and ideas used in this paper, and extracts the innovation points of this paper. Secondly, it introduces the basic theory of transnational supply chain, including internal trade, transfer price, supply chain planning, key issues and decision models, and briefly reviews and summarizes the existing literature. Then, this paper focuses on the supply chain production and distribution planning under different decision modes. The production distribution model under centralized decision and the production and distribution plan model under decentralized decision mode are established respectively. The two models have similar structures, including a parent company and a multinational company with multiple subsidiaries. A multi-plan period, multi-company, under random production capacity and random demand, Multi-product two-stage supply chain models. However, these complex models are treated differently. The same thing is that the method of dealing with the constraints of stochastic production capacity and stochastic demand with uncertain factors in the model is to bring profit to the model. These constraints are transformed into explicit constraints by using the principle of opportunistic constraint programming. The nonlinear model is transformed into a linear model by variable substitution. The difference is that the decentralized decision model relaxes the coupling constraints between enterprises by Lagrangian relaxation algorithm. The centralized decision problem of the whole supply chain system is decomposed into the independent production and distribution planning decision problem within each enterprise. The parent company acts as the coordination center and iterates through the Lagrangian relaxation algorithm to obtain the optimal solution. At last, the two models give the example analysis. Under the centralized decision, the example analysis verifies the two viewpoints of this paper: first, in order to maximize the global profit under the condition of limited production capacity, Priority is given to markets that bring greater profit returns, leaving other market demands unmet. Second, under certain circumstances, global profits can be maximized at the expense of branch interests; in decentralized decision-making, The feasibility of the model is verified by an example, and the validity of Lagrangian relaxation algorithm is demonstrated.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F276.7;F273

【参考文献】