Stochastic simulation and optimization in supply chain management
Stochastic simulation and optimization in supply chain management
复制标题
供应链管理中的随机模拟与优化
DOI：
10.1177/0037549718772527
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
Gang Chen
中科院分区：
文献类型：
--
作者：
Baozhen Yao;Gang Chen
供应链管理（SCM）是指从采购开始到最终客户的过程，以最小的成本优化SCM。它被视为一种商业哲学，将伙伴关系的概念扩展到多公司努力管理从供应商到最终客户的货物总流量。它包括原材料、在制品库存和产成品从原产地到消费地的移动和储存。供应链管理面临着多种不确定性，包括需求信息偏差逐渐夸大所导致的供应商库存不确定性和物流供应中时间延迟累积效应所导致的交货期不确定性。然而，不确定性和复杂性不断上升，需要更复杂的模型和更先进的新方法来解决这些问题。这推动了基于随机模拟和优化的计算密集型SCM方法的发展。随机模拟和优化可以处理具有非线性和随机元素的复杂模型，以获得准确和有洞察力的结果。它为SCM的优化、随机优化算法和有效的仿真技术提供了有前途的机会。广泛的主题进行了讨论，特别是在以下领域：优化单和多级库存/能力管理，包括应用随机规划方法的库存管理，物流配送管理，并与风险考虑客户需求;和可持续供应链，包括回收和逆向物流，物流、交通和运输中的随机方法和算法，供应链管理中的时间相关路由问题，采购管理，分析性能模型，仿真和信息模型，贝叶斯全局优化，离散事件仿真，以及供应链管理中的蒙特卡罗方法。在题为“多级供应链库存转运模型在不同层次上的仿真”的论文中，Yan和Liu使用系统动力学方法建立了库存转运模型。然后通过仿真对不同层次的单链、双链、三链和四链库存转运模型进行了比较分析。模型分析的结果表明，平均库存水平的变化很小，并随着库存转运模型中链的数量从一个增加到四个而下降。库存转运系统的平均客户需求满意度随着转运系统增加链条数量的增加而持续上升，平均客户满意度的增长幅度随着转运成本的增加而持续下降。研究供应链中的复杂库存转运问题，可以在产业集群中有所作为，实现双赢和规模经济。在题为“报废汽车零部件区域配送中心选址的优化”的论文中，Sun等人通过双层规划模型获得了配送中心选址成本和运输成本之间的权衡。在上层，确定配送中心的选址。下层代表了配送中心候选点在任意位置模式下的配送路径。采用混合复杂进化算法（SCE-UA）对配送中心选址的上下层模型进行求解。以辽宁省为例进行了数值试验，验证了模型的有效性.结果表明，该模型能够有效地确定报废汽车的配送中心，提高资源利用率，降低报废汽车逆向物流网络的成本。在题为“基于叶片光合作用运输的启发式算法”的论文中，Yu等人提出了一种基于运输网络设计的非确定性多项式时间的运输网络设计的双层规划模型。上层追求用户总出行时间最小，
Supply chain management (SCM) refers to the processes from the beginning of procurement to the end customer, optimizing SCM with minimal cost. It is viewed as a business philosophy, which extends the concept of partnerships into a multi-firm effort to manage the total flow of goods from the supplier to the ultimate customer. It includes the movement and storage of raw materials, work-in-process inventory, and finished goods from point of origin to point of consumption. SCM is facing a variety of uncertainties, covering supplier inventory uncertainty caused by gradually exaggerated demand information deviation and delivery date uncertainty caused by the cumulative effect of time delay in logistics supply. However, the uncertainties and complexities are constantly rising and more complex models and more sophisticated novel methodologies are required to tackle them. This has driven the development of computation-intensive SCM methods based on stochastic simulation and optimization. The stochastic simulation and optimization can handle complex models with nonlinear and stochastic elements to get accurate and insightful results. It offers promising opportunities for the optimization of SCM, stochastic optimization algorithms, and efficient simulation techniques. A broad range of topics are discussed, especially in the following areas: optimization of single and multi-echelon inventory/capacity management, including application of the stochastic programming approach to inventory management, logistics distribution management, and customer demand with risk considerations; and sustainable supply chain, including recycling and reverse logistics, the stochastic method and algorithm in logistics, traffic and transportation, the timedependent routing problem in SCM, procurement management, the analytical performance model, the simulation and information model, Bayesian global optimization, discrete-event simulation, and Monte Carlo methods in SCM. In the paper entitled ‘‘Simulation of multi-echelon supply chain inventory transshipment models at different levels,’’ Yan and Liu established an inventory transshipment model using the system dynamics method. Then the authors analyzed the single-, double-, three-, and fourchain inventory transshipment models at different levels comparatively through simulation. The results of the model analysis indicated that the average stock level was minimally altered and declined as the number of chains in the inventory transshipment model increased from one to four. The average customer requirement satisfaction rate of the inventory transshipment system consistently increased as the number of chains added to the transshipment system increased, and the growth range of the average customer satisfaction rate continuously decreased with the increasing transshipment costs. The study of complex inventory transshipment of the supply chain could make a difference in the industrial cluster to achieve win–win and economies of scale. In the paper entitled ‘‘Optimization of a regional distribution center location for parts of end-of-life vehicles,’’ Sun et al. obtained a trade-off between the location cost of the distribution center and the transportation cost by a bilevel programming model. In the upper level, the sites of distribution centers were determined. The lower level represented the delivery routes under any location pattern of candidate sites of distribution centers. The shuffled complex evolution algorithm (SCE-UA) was applied for the upper and lower models of the distribution center location. A numerical test based on Liaoning Province in China was employed to validate the proposed model. Compared with the current scheme, the results demonstrated that the model can be an effective method to determinate a distribution center for end-of-life vehicles (ELVs), which could help improve the utilization of resources and lower the cost of the reverse logistics network of ELVs. In the paper entitled ‘‘A heuristic algorithm based on leaf photosynthate transport,’’ Yu et al. presented a bilevel programming model for transportation network design based on the non-deterministic polynomial-time of the transportation network design. The upper level layer pursued the minimum total travel time of users and the