綜合切削中心機維修診斷專家系統設計

Abstract

本論文提出從客戶的維修資料、 運用 IDEF0 的技術、和關聯性矩陣測試 的系統化方法來設計綜合切削中心機的維修診斷專家系統的知識庫。從客 戶的維修資料開始，構建出包括目錄、故障現象、故障的零組件、測試和 修理的階層知識庫。綜合切削中心機零組件的目錄和故障現象之間的關係 是透過 IDEF0 的技術來建立。在知識庫和客戶的維修資料之間的驗證是 透過關聯性矩陣來加以測試，測試無誤之後，再經過使用者的測試來修正 知識庫的正確性。 此外，運用 InSupport 所建立的綜合切削中心機維修 診斷專家系統的原型所需的資料是以友嘉實業公司所生產的 FV-800A 機 器為例子，作為本論文的實例運用。 A systematic methods based on customer maintenance data, IDEF0 skills, and adjacency matrix test is proposed to design the knowledge base of diagnostic expert system for CNC machining center. From the customer maintenance data, a hierarchical knowledge base consisting of category, symptom, failure, test and repair is constructed, The relationship between component category of CNC machine center and symptom is specified using IDEF0 skills. Then the consistency between knowledge base and customer maintenance data is verified through adjacency matrix test. After this, a second test for the knowledge base is through many feedbacks from users. Also a prototype diagnostic expert system based on InSuppoert for a CNC machining center FV-800A, Fair Friend Ent. Group, has been implemented for demonstration purpose.