| Summary: | 博士 === 中原大學 === 工業與系統工程研究所 === 99 === In this research, we consider the imperfect production processes of the Economic Lot Scheduling Problem (ELSP) which deals with multi-product produced on a single production facility. The basic assumption of the classical ELSP model is that the production process is perfect. In fact, this assumption may not be valid for most of the manufacturing environments. Owing to aging, many production processes may deteriorate to produce defective products. Due to imperfect production processes, decisions regarding whether and when to implement a screening process are typical of every inventory problem. Besides, rework of defective products, which may eliminate waste, reduce disposal costs and comply with environmental legislation, is also an important topic.
This research is concerned with the joint effect of scrap, imperfect repair, and shortage on the ELSP involving three types of screening processes with or without rework considerations. In general, a production facility is assumed to be in control at the start of a production run. Owing to aging, it may deteriorate and shift at a random time to an out-of-control state, and consequently produces a fixed fraction of nonconforming items. There are two situations to handle the defects in this research: (1) these defective items cannot be repaired or reworked, and thus must be scrapped with an additional cost; (2) some proportion of defective items cannot be reworked and must be scrapped. The rest of these defective items can be reworked off line. However, since the repair process is imperfect, scraps are produced during the manufacturing and/or rework processes. In the first situation, three models are presented to consider whether and when to implement a screening process, which are: (1a) no screening process, (1b) after-production screening, and (1c) in-production screening. On the other hand, there are several models are proposed in the second situation with various screening options that are (2a) under no screening process with repair before/after depletion, (2b) after-production screening with repair right after screening, and (2c) in-production screening with repair right after production.
In this research, a Common Cycle (CC) approach is applied to solve ELSP under imperfect production process. The objective is to determine an optimal common production cycle time that minimizes the expected total cost per unit of time. A solution procedure is developed to find near-optimal solutions for the models. By approximating exponential functions included in the objective functions, certain closed-form solutions can be developed. A genetic algorithm heuristic is implemented to verify the effectiveness of the solution procedure we proposed in this dissertation. Numerical examples are presented to illustrate the derivation of the optimal common production cycle time. These results show apparently that the less reliable a production process is, the shorter a production common cycle time should be. Without rework considerations, strategy without screening is properly adopted when the process is reliable. As the process becomes less reliable, after-production screening process is implemented. On the other hand, with rework considerations, strategy without screening is properly adopted.
|
|---|
