FAULT DIAGNOSIS OF MARINE DIESEL ENGINE BASED ON MIXED SIMILARITY ALGORITHM

Abstract

In view of the problems of inlet and exhaust faults and clogging of the marine diesel engine, the appropriate thermal parameters are selected as the basis for fault diagnosis and positioning. In this paper, the improved Pearson correlation coefficient and grey relational diagnosis analysis are combined, and a hybrid similarity collaborative filtering algorithm is proposed. At the same time, the simulation model of the diesel engine is built by AVL-BOOST software, and the fault samples are simulated. The mixed similarity collaborative filtering algorithm is used to calculate the correlation degree of the fault data, and the final diagnosis result is given accordingly. The results show that the hybrid similarity diagnosis algorithm has excellent diagnosis speed and accuracy, which can ensure the fault diagnosis and location of diesel engine is more accurate and reliable.