| Summary: | 碩士 === 國立交通大學 === 工業工程與管理系 === 89 === Since the competition in wafer fab is keener and keener, operational staff needs a quick response tool for supporting the make decision in production planning and scheduling. The thesis thus adopts eM-Plant software to construct a production scheduling system and simulation platform for the wafer fab.
To construct the system, the object-oriented analysis and design as well as the structured analysis and design are inclusive in the system development process. UML, which is the tool used in object-oriented analysis and design, can benefit object-oriented simulation. Additionally, the structured analysis and design can transmit the methodology and algorithms into model easily and correctly.
The system contains six modules, which are production planning and scheduling module, shop floor control module, virtual fabrication module, demo module, database module and auto-creation module. In the production planning and scheduling module, production performance indices such as cycle time estimation, utilization rate and WIP level can be quickly calculated. Release sequence and due date setting are also generated. As for shop floor control module, wafer release control such as CONWIP and TB under load regulating considerations (LTRB) are built. Besides, the virtual fabrication module provides user to set specialized simulation environments, and demo module can present the production performance indices and simulation results in static and movement chart. In addition, input/output and temporary storage data are saved in database module. The auto-create module is finally used to duplicate a class as an object so as to integrate all modules to be a system.
The experimental studies show that production performance indices such as product and layer cycle time for each priority class of orders, WIP level, and utilization of each workstation can be quickly generated by the designed system. Besides, operational staff can easily get the results from different control methods without spending a lot of time to construct simulation models. Thus, the designed system could be treated as a production support tool for production planning and making decisions.
|
|---|
