선박엔진 통합시스템의 실시간 고장진단 기법에 관한 연구

A Study on the Real-time Fault Diagnosis of Integrated Ship Engine Systems

This paper presents the diagnostic method of an integration system where both diagnosis system and monitoring system are combined. Knowledge base for ship engine fault diagnosis is implemented using hybrid knowledge representation technique by EE(Element Expert) shell, and using a certainty-factor method to handle uncertainties of evidences and rules. Whenever alarm occurs, the monitoring system sends alarm data to the diagnosis system. At the same time, the diagnosis system converts alarm data type to diagnosis data type, verifies alarm data, and calculates data's trend. To map alarm data to evidences of knowledge base, the diagnosis system uses two databases. The real-time diagnosis results are provided by a reasoning mechanism implemented using visual C++ for driving OLE automation server of EE shell.