按单装配供应链的协同库存管理研究

发布时间：2018-04-10 05:20

本文选题：并行遗传算法　切入点：供应链　出处：《兰州交通大学》2014年硕士论文

【摘要】：近几十年来,随着科学技术的飞速发展和世界经济一体化进程的加速，企业面临的市场环境发生了巨大变化。消费者的购买欲望和购买力在逐年增加,消费者成为市场运作的关键，复杂多变的产品需求和不可预测的市场动态对企业的运营方式造成了巨大冲击。在这样的背景下,按单装配这种新的供应链管理方式被提出,并逐渐受到企业管理者们的重视。同时，由于大规模机器化生产的普及，产品更新换代的时间大大的减少，急剧缩短的产品生命周期使库存成为了供应链管理的瓶颈，制约着供应链的整体性能。以制造企业为核心的传统供应链管理已经无法满足客户在产品质量和响应时间上的需求，为企业提供的利润越来越小。因此，协同管理成为供应链赢利的突破口，通过对节点企业进行信息共享、统一规划和协同管理，使得供应链的总成本得到降低。所以,探讨按单装配供应链的协同库存管理问题具有重要意义。本文首先介绍了国内外有关按单装配供应链和协同库存管理研究的现状。其次，研究了供应链相关理论、库存管理的相关知识、按单装配和协同库存等相关概念，分析了供应链中现存的问题。然后，将经济学原理和供应链管理相结合，提出按单装配供应链协同库存管理这一策略，，为下文模型的建立奠定了理论基础。本文通过对服从泊松分布的产品随机需求及服从泊松分布的供应商和制造商随机订货提前期进行分析，建立了以求解最少成本为目标的多产品、多需求、允许缺货的两级按单装配供应链的协同库存模型。并引入混合并行遗传算法，以改进的步长加速法作为交叉算子，十进制到二进制的转换的截取作为变异算子，增加迁移算子，利用最大适应度的迭代次数作为终止规则，利用多进程并行的方式求成本最优时的再订货点的库存量和固定订货量。混合并行遗传算法缩短了求解模型所用的时间并增强了最终解的准确性。通过一定规模的测试数据对构建的模型进行仿真模拟及结果分析。通过仿真实验数据验证按单装配供应链的协同库存模型的收敛性和稳定性，并证明在不同的环境中该模型对库存成本的降低都起着作用。最后，总结了本文的研究成果和结论，着重分析了模型建立和算法设计的不足，对今后按单装配供应链协同库存管理研究的发展从更具普遍性的角度提出了一些建议和思考。
[Abstract]:In recent decades, with the rapid development of science and technology and the acceleration of the process of world economic integration, the market environment faced by enterprises has undergone tremendous changes.The purchasing desire and purchasing power of consumers are increasing year by year. Consumers become the key of market operation. The complex and changeable product demand and unpredictable market dynamics have a great impact on the operation mode of enterprises.Under this background, the new supply chain management mode of single assembly has been put forward, and gradually paid attention to by enterprise managers.At the same time, because of the popularization of large-scale machine production, the time of product renewal is greatly reduced, and the sharply shortened product life cycle makes inventory become the bottleneck of supply chain management, which restricts the whole performance of supply chain.The traditional supply chain management with manufacturing enterprises as the core can not meet the needs of customers in product quality and response time, and the profit for enterprises is becoming smaller and smaller.Therefore, collaborative management has become a breakthrough in the profit of supply chain. By sharing information, unified planning and collaborative management, the total cost of supply chain can be reduced.Therefore, it is of great significance to discuss the collaborative inventory management of single assembly supply chain.This paper first introduces the research status of single assembly supply chain and collaborative inventory management at home and abroad.Secondly, the related theories of supply chain, the knowledge of inventory management, the related concepts of single assembly and collaborative inventory are studied, and the existing problems in supply chain are analyzed.Then, by combining economic principles with supply chain management, the strategy of collaborative inventory management in single-assembly supply chain is put forward, which lays a theoretical foundation for the establishment of the following model.Based on the analysis of the random demand of the products distributed by Poisson and the random ordering lead time of the suppliers and manufacturers of the service from the Poisson distribution, a multi-product and multi-demand system with the aim of solving the minimum cost is established in this paper.Collaborative inventory model of two-level single-level assembly supply chain that allows shortage.A hybrid parallel genetic algorithm is introduced, in which the improved step size acceleration method is used as the crossover operator, the decimal to binary transformation is used as the mutation operator, the migration operator is increased, and the maximum fitness iteration is used as the termination rule.The inventory and fixed order quantity of reorder point are obtained by multi-process parallel method when the cost is optimal.Hybrid parallel genetic algorithm shortens the time to solve the model and enhances the accuracy of the final solution.Through a certain scale of test data, the model is simulated and analyzed.The convergence and stability of the collaborative inventory model of single-assembly supply chain are verified by simulation data. It is proved that the model plays an important role in reducing inventory cost in different environments.Finally, the research results and conclusions of this paper are summarized, and the shortcomings of model building and algorithm design are analyzed, and some suggestions and thoughts on the future development of collaborative inventory management in single-assembly supply chain are put forward from a more universal point of view.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F253;TP18

【参考文献】