ABSTRACT

Discrete Event Simulation (DES) is the main tool used in analyzing supply chains (SCs). In many simulation studies the effect of the modeling approach on SC performance is ignored. Many researches have overlooked the fact that in DES where events are the drivers of variable changes; the modeler has the option to change these variables pre or post delay event occurrence. In SCs this choice is critical and has a significant effect on the results obtained. Another modeling approach is System Dynamics (SD) which overcomes the difficulties of DES because variables do not change in discrete jumps but rather smoothly overtime. Both DES and SD approaches are combined in one model and are compared to two types of DES modeling approaches; Pre-delay and Post-delay DES. A case study is implemented in a local cement SC. Inventory levels (IL), holding costs (HC) and ordering costs (OC) for the Combined DES/SD model are shown to be between those of the other two models. Post-delay DES is shown to overestimate IL and HC but underestimate OC; while, Pre- delay DES is shown to underestimate IL and HC while overestimating OC.

摘要

离散事件模拟 (DES) 是分析供应链 (SC) 的主要工具。在许多仿真研究中，建模方法对 SC 性能的影响被忽略了。许多研究忽略了一个事实，即在 DES 中，事件是变量变化的驱动因素；建模者可以选择在延迟事件发生之前或之后更改这些变量。在 SCs 中，这一选择至关重要，并且对获得的结果有显着影响。另一种建模方法是系统动力学 (SD)，它克服了 DES 的困难，因为变量不会在离散的跳跃中变化，而是随着时间的推移而变化。DES 和 SD 方法都结合在一个模型中，并与两种类型的 DES 建模方法进行比较；预延迟和后延迟 DES。在当地的水泥 SC 实施案例研究。库存水平 (IL), 组合 DES/SD 模型的持有成本 (HC) 和订购成本 (OC) 显示在其他两个模型之间。延迟后的 DES 高估了 IL 和 HC，但低估了 OC；而预延迟 DES 被证明低估了 IL 和 HC 而高估了 OC。