A fault diagnosis method for marine hydraulic steering gear based on Bayesian networks

Marine steering gear is the mechanical equipment of ship course-keeping and turning. However, the research level of fault theory and fault diagnosis of marine steering gear is underdeveloped, which leads to the poor reliability of ship steering system. Therefore, a fault tree model is established with the historical data and knowledge and experience of steering gear fault to solve the problems, which is transformed into a Bayesian network model to realize the forward fault probability inference and fault cause tracing by using Bayesian network to realize the fault diagnosis and location of ship hydraulic steering gear. Subsequently, forward fault probability inference and fault causes tracking are realized by using Bayesian networks in two cases respectively. On the one hand, this method can assist the staff to judge the possibility of system failure according to the status of parts and components to make maintenance plan. On the other hand, this method can locate the fault of the component when a fault occurs in the system and arrange the inspection in turn through the probability distribution to improve the efficiency of fault diagnosis.