近年來隨著機械製造技術的進步，因此機械功能性也更顯複雜，對於操作人員而言，擁有高效率的監控與維護的平台才能即時處理機械的突發狀況。以往機械單元若發生故障時，往往必須花費相當大的成本在故障排除的工作上，若能建構一即時診斷環境，提供使用者隨時掌握生產的各項重要資訊，並於異警發生時提供適當的輔助機制，達成網絡化診斷維修之目的，此一課題值得深入探討與研究。 本文提出以網際網路技術（Internet）對於射出成型機進行遠距維護，並利用模糊故障診斷來對設備作故障分析。遠距維護系統是依據案例式推理為基礎，分析射出成型機之異常狀況並建立成案例庫，運用過去案例的經驗，來解決新的問題。 故障模糊診斷是針對案例無法解決故障，則對設備進行診斷分析，找出最有可能發生故障的設備組件，因此利用到故障樹分析法、模糊理論與貝氏定理來建構。 The manufacturing technology of machines is mature gradually, and to promote the machine functions is becoming more complex gradually in the same time. For the manufacturer, to possess a high-effect remote & maintenance platform, and then to handle machine alarm states in real time are very important. Formerly if the machine unit gave the alarms, manufacturers had to spend a lot of time to eliminate where the machine is breakdown. If it enables to build a real-time diagnosis environment, users might be able to obtain the information of manufacturing in real time. This function can offer users the problem solutions of manufacturing and breakdown. The topic merits discussion and study. This paper presented a web-based remote diagnosis and maintenance solution for the injection molding machine, and analyzed the breakdown by fuzzy fault diagnosis. The Remote Diagnosis and Maintenance System is based on Case-based Reasoning, and diagnosed the unusual condition of the infection molding machine, and then built the database. The system is used the past case experience to solve the new problem. The fuzzy fault diagnosis is aimed at the case-based reasoning to be unable to solve the breakdown. Diagnosing the equipment, and discovered the most possibility to case breakdown of the component of the machine. Therefore the diagnosis structure applies the Fault-Tree Analysis, Fuzzy Theorem, and the Bayes Theorem.
