[[alternative]]Design of a Fault Diagnosis System for Diesel Engine Generators

碩士電機工程學系[[abstract]]柴油引擎發電機組系統係由柴油引擎，同步發電機及控制系統設備組成，一旦失效會嚴重影響緊急供電系統的妥善率。因此，如何提升維修人員故障診斷能力，是非常重要的議題。故障診斷是一個複雜的知識能力，維修人員需要擁有一系列故障訊息的經驗累積、整合資訊並配合故障診斷的推理與驗證。故障維修診斷的知識技能與經驗的要求常超過維修人員所能承擔的負荷，因而造成診斷時間的拖延或錯誤。因此，如何縮短故障診斷時間及降低故障診斷疏失為一重要課題。本研究以擷取專家維修診斷知識建立資料庫，並以微控制器掃描比對資料庫找出最有可能的故障。希能提升維修人員之診斷能力，降低診斷時間，減少維修人員診斷疏失，增進維護效率，進而提升柴油引擎發電機組的維護性與妥善率之目標。[[abstract]] The diesel generator set is composed of a diesel engine, synchronous generator and a control system. Once the diesel generator set is out of order, the availability of urgent power supply system will be affected. So it is important to improve the fault diagnosis ability of maintenance workers. Because the fault diagnosis is a complex knowledge, maintenance workers need a series of experience accumulation, and integrate information with inference and verification of fault diagnosis. The demanded knowledge and experience of fault diagnosis often exceed the affordable load taken by maintenance workers, and therefore it results in longer repair time or wrong diagnosis. To decrease the repair time and avoid the wrong diagnosis, this thesis proposed a microprocessor-based system to rapidly compare the possible fault from the expert diagnosis database. This proposed system can improve the diagnosis ability of maintenance workers, decrease the repair time, avoid the wrong diagnosis, promote the maintenance efficiency, and enhance the availability of diesel generator sets