Fault Diagnosis of Diesel Engine Lubrication System Based on Bayesian Network

Abstract

The existing fault diagnosis techniques are mostly manual inspections. In recent years, support vector machines have been used for fault diagnosis. However, this method is a classification method, which can only judge whether the fault is faulty and cannot diagnose the specific fault cause. In order to achieve accurate positioning of diesel engine lubrication system fault diagnosis, avoid more serious faults of diesel engine. Proposes a fault diagnosis method for Bayesian networks. Taking a diesel engine as the research object, the fault analysis of the lubrication system is carried out to establish a fault tree, and then the Bayesian network diagnosis model of the lubrication system is established by transforming the fault tree into the Bayesian network. Finally, the model is verified by an example. The actual case verification shows that the Bayesian network can be used for fault diagnosis of the oil system, which can realize the fast and accurate positioning of the fault cause. This method can be used for fault diagnosis of marine diesel engines.