Research On The Reliability Evaluation Methodology Of Subsea Bop System Based On Bayesian Networks

The mining of marine oil and gas resources is a project related to the lifeblood of national economic,military,political and other aspects.Subsea blowout preventer(BOP)is one of the high-end equipment,and also the the most important security barrier of the entire drilling process.To avoid the occurrence of the accident requires reliable design,reliable manufacturing,reliable use and maintenance and other aspects of common protection.In this paper,Bayesian network is used as a basic tool to systematically study its reliability assessment methods to improve the reliability of modeling simplicity,accuracy and reasoning the rapid,and in that way lay the foundation of the water test and practical engineering applications of domestic subsea BOP.A method of evaluating the reliability of electronic systems considering intermittent faults is proposed,and the three-state Markov model is embedded in the transition condition probability table of the dynamic Bayesian network.The reliability evaluation model of the three-mode redundant and dual-mode redundant two electronic control systems is established by using this method.The reliability and availability of the two electronic control systems are counted and the importance of each component in the two systems is calculated by applying the concept of mutual information order.The effects of intermittent fault parameters,including the intermittent fault coefficient,the permanent fault coefficient,and the intermittent maintenance factor on the reliability and availability of the two systems are investigated.A real-time reliability evaluation method based on Bayesian network is proposed.Through the intermediate node,the static Bayesian network is input to the dynamic Bayesian network reliability evaluation model by inputting in the fault diagnosis result of the fault symptom.By using this method,the real-time reliability model of the hydraulic control system of the sluice bunker is established,and the mutual information of the model is calculated to determine the order of importance of the components.Values of real-time reliability and changes of the four different cases are calculated thus providing the recommendations for the design and maintenance to improve the reliability of the system.A dynamic object-oriented Bayesian network is proposed to evaluate the reliability of complex systems.The same or similar fragments and the whole time slice are defined as class for encapsulation modeling.The model of two kinds of deep water BOPs is established by using this method,and its reliability and usability are calculated.A set of the experimental prototype of deep water BOP system and its real-time reliability evaluation system are developed.On the basis of the original two sets of gate BOP mechanical system,a ring BOP,two-mode redundant hydraulic control system and NI acquisition card electronic control system based on the relay are developed.The real-time reliability evaluation system based on LabVIEW and MATLAB is developed to verify the real-time reliability evaluation theory.