| Summary: | 碩士 === 國立成功大學 === 工學院工程管理專班 === 92 === Technological advancement and highly intensive capital investment within the semi-conductor industries in the past decade have meant that productivity, yield, process safety, and diagnosis of equipment failures in the process are very important. Among these factors, the fault diagnosis of equipment is especially critical because timely correction of errors will lead to higher productivity, yield and safety. Previously, maintenance engineers judged from alarm status of the control monitoring systems, and their own knowledge, experience to identify the causes, then suggest the preventative strategies and actions. Nowadays, more intensive capital investment, higher risk, and increasingly more complex equipment have raised the industry standard to demand an even faster response in the failure diagnosis. This paper, therefore, attempts to construct a prototype of alarm expert system based on FMCS (Facility Management Control System), while using AAS (Air Abatement System) as an example. This expert system operates according to the inference based on a knowledge base consisting of senior engineers’ operating experiences and relevant information. Once the irregularity happens, FMCS will issue a timely warning and show the corrective procedures by the graphic displays. This system minimizes the judgment errors, manpower and material losses, while enabling factory engineers to diagnose errors quickly and reliably. As a result, with downtime minimized, yield can be enhanced and risk reduced.
|
|---|
