It is clear that a material handling system has a significant influence on the overall performance of a manufacturing system. In order to obtain the best overall system performance, the reliability and availability (R&A) of the material handling system is critical. However, there have been difficulties in evaluating R&A of material handling systems due to the fact that they exhibit stochastic behavior that includes degrading performance over time. Many analytical models have been proposed for the evaluation of R&A of complex degradable material handling systems, however, they are not adequate for a variety of reasons including mathematical complexity, excessive numerical computation, and inability to model transient behavior. Because of the limitations and restrictions of analytical performance models, simulation has been often used as an analysis and evaluation method in the design and operation of large and complex manufacturing systems.; This dissertation presents a simulation-based performance evaluation approach based on the PI method proposed by Usher (2000). The PI method is developed by applying a quadratic loss function, often referred to as “Taguchi loss function”, to characterize the impact of component failures and repair processes on the overall performance of a material handling system. First, the dissertation derives the confidence interval estimates of PI for the three types of performance characteristics, i.e. nominal-is-best, larger-is-better, and smaller-is-better. Next, the effectiveness of PI in two different situations is analyzed, i.e. the performance evaluation of a single system and performance evaluation in comparison of alternative systems. For the performance evaluation of a single system, factor screening experiments, designed to identify the system components whose failure and repair most impact performance, is investigated. For the comparison of alternative systems, a time-based PI approach is proposed. The approaches are illustrated through the analysis of a large and complex distribution center.
展开▼
机译：显然，物料搬运系统对制造系统的整体性能有重大影响。为了获得最佳的整体系统性能，物料搬运系统的可靠性和可用性（R＆A）至关重要。然而，由于材料处理系统表现出包括随时间降低性能的随机行为，因此在评估R＆A方面存在困难。已经提出了许多分析模型来评估复杂的可降解材料处理系统的R＆A，但是由于多种原因，它们不足以解决这些问题，包括数学上的复杂性，过多的数值计算以及无法对瞬态行为进行建模。由于分析性能模型的局限性和局限性，在大型和复杂制造系统的设计和操作中，仿真通常被用作分析和评估方法。本文提出了一种基于模拟的绩效评估方法，该方法基于Usher（2000）提出的 PI 方法。 PI 方法是通过应用二次损失函数（通常称为“田口损失函数”）开发的，以表征部件故障和维修过程对物料搬运系统整体性能的影响。首先，本文针对三种性能特征，即名义上最好，更大更好和较小更好，推导了 PI 的置信区间估计。接下来，分析 PI 在两种不同情况下的有效性，即单个系统的性能评估和替代系统的性能评估。为了评估单个系统的性能，研究了旨在确定故障和维修对性能影响最大的系统组件的因素筛选实验。为了比较替代系统，提出了一种基于时间的 PI 方法。通过对大型复杂配送中心的分析来说明这些方法。
展开▼
