Reliability Analysis Of Compressor Unit Based On Dynamic Bayesian Network

With the development of the economy,the demand for natural gas is increasing day by day.The compressor unit is an important part of the natural gas industry.It represents the high-tech level of major national equipment and is the key to promoting the transformation and upgrading of the national equipment manufacturing industry.The compressor unit is regarded as the "heart" of the natural gas pipeline transportation process and plays a key role in the reliability of natural gas energy transportation.Therefore,it is necessary to conduct reliability research on natural gas compressor units to ensure the safe and efficient operation of natural gas compressor units.In this thesis,the reliability analysis of the compressor unit is carried out.The main research contents are as follows:(1)Carry out failure mode and effect analysis for compressor units.First of all,according to the structure and function of the compressor unit,the failure mode and impact analysis is carried out on the maintainable parts of the 12 main engines and auxiliary equipment of the compressor unit,and the risk priority number method in the traditional failure mode and effect analysis is used for the detection degree.In order to solve the problem of unclear classification of grade,occurrence frequency grade,and severity grade,a risk priority number(RPN)quantification method for compressor units driven by actual data is proposed,and on this basis,the RPN interval division rule is given,get the high,medium and low risk failure modes of the system and put forward targeted improvement measures.(2)Carry out dynamic fault tree analysis on the lubricating oil subsystem in the compressor unit system.According to the results of the failure mode and effect analysis,it is concluded that the failure frequency of the lubricating oil system is high,the number of medium-risk failure modes is large,and the failure of the lubricating oil system has the characteristics of dynamic and sequential characteristics.The dynamic fault tree modeling of the lubricating oil system of the compressor unit is carried out,and the Markov model is used to solve the established dynamic fault tree,and the reliability curve of the lubricating oil system is obtained,and the reliability of the lubricating oil system at different times is calculated.(3)Aiming at the problem that the Markov model has a state space explosion when the system complexity is high,the dynamic fault tree is mapped to the Bayesian network,and the reliability analysis method of the compressor unit lubricating oil system based on the discrete-time Bayesian network is studied.Under the same task time,the system failure probability based on the Markov method is compared with the results obtained by the discrete-time Bayesian network to verify the correctness of the researched method.(4)On the basis of the reliability analysis of the lubricating oil system based on the discrete time Bayesian network,the reliability analysis of the overall compressor unit is carried out to solve the failure probability of the compressor unit system in different time periods and the failure within the task time Probability,introduce the importance of static Bayesian network into the time dimension,study the root node probability importance and key importance algorithm of compressor unit system based on discrete time Bayesian network,and propose an important method based on compressor unit maintenance strategy Degree,through the root node importance analysis to get the weak link of the compressor unit system.