Application of Artificial Intelligence Methods for the Diagnosis of Marine Diesel Engines

The paper presents a diagnostic system for marine diesel engine based on an expert system model. The research relevant to knowledge acquisition for this system was done, knowledge base was built and general structures of the expert system was proposed. Basic sources of knowledge which can be used for construction of knowledge base are also identified. The basic knowledge related to the diesel diagnostic was undertaken from experts and diagnostic database. The paper questionnaire was used to the knowledge acquisition from experts. The basic knowledge related to the marine diesel exploitation was undertaken. The rule induction algorithms was used to knowledge acquisition from database. During the experiment efficiency of LEM induction algorithms was compared to new MODLEM and EXPLORE algorithms. Training and test data were acquired from experiment on marine engine Sulzer 3AL 25/30.