FAULT DETECTION AND IDENTIFICATION OF MARINE DIESEL ENGINE USING EXPERT KNOWLEDGE AND FUZZY LOGIC APROACH

Marine propulsion engine is the most important system onboard ship for ship’s successful mission. Therefore, it’s very useful to estimate its future state and availability. In this paper, one effective method for marine diesel propulsion engine faults detection and identification based on fuzzy logic, expert knowledge and experience in real conditions at sea will be presented. The faults diagnosis method is based on real-time diagnostic signals i.e. symptoms and events and their relation to faults in an extended manner with tracking process relevant variables during normal as well as faulty operation and times of symptoms and events occurrence. Engine expert knowledge and especially long time experience will be used as key decision method in fault detection and identification process. The effectiveness and power of this method has been evaluated simulating faults of engine bearings. The simulation has been done using real ship's propulsion plant simulator "Kongsberg, Full mission engine room simulator". The simulation results has been compared with expert decision in faults diagnosis. This paper represents a part of research results conducted within scientific project “New technologies in diagnosis and supported by Ministry of Science, Education and sports of the Republic of Croatia 1 .