The Research Of Long-distance Gas Pipeline Pre-repair Strategies Based On RCM Theory

China’s long distance natural gas pipeline has been increasingly improved, at thesame time as the early laying all aspects of technology limited. Now most of thesepipeline has worked for20-30years, and their aging and various hidden troubles havebeen gradually emerging. Aiming at this situation, the papering uses reliability centeredmaintenance (RCM) strategy as the method of repair the long distance natural gaspipeline to ensure the safe operation of the pipeline.This paper studies mainly:(1)For long distance natural gas pipeline’s characteristics of inflammable,explosive, long distance buried gas transmission, the pipeline’s trouble factors areanalyzed, and the long gas pipeline trouble trees are established. Using the calculationmethod of triangular fuzzy number, the main reasons of the pipeline troubles can bearrived. The main cause is the pipeline corrosion perforation, the third party damageand the design flaw. According this result, the paper analyzes the various factors’trouble reason and effect.(2)On the basis of factor analysis of pipeline failure, we choice the method ofreliability centered maintenance. Using the RCM logic decision tree, we can analysisthe pipe hidden-failure consequences, the security consequences, task consequencesand economic consequences these four aspects, then achieve the main parameters ofthe preventive maintenance, and the predictive maintenance. Next we establish themodel that reliability centered in the long distance natural gas pipeline preventiverepair. (3)When establishing the long-distance natural gas pre-maintenance model, we getthe model for predicting maintenance of pipeline forecast maintenance cost model andthe available model on the basis of established model assumptions, parameters andpipeline failure probability. At the same, we also get the preventive maintenanceavailability model on the basis of established the model assumptions and parameters.(4)In order to achieve more accurate repair information, before using theapplication of repair strategies, we first use the neural network method. Through theMATLAB training comparison, we can choose the optimal BP neural network model.Prediction the results of the acquired data time series obtained, we can get the pipelinedevelopment trend data. Combined with the already available data, and applied to themodel that we have built. Through comprehensive analysis and calculation, we can getthe reliable repair strategy.Through the research analysis, the long distance natural gas pipeline repair modelstrategy can be effectively applied to the daily management of the pipeline. In theactual work, we can more perfect the shortcomings of the method. Then the maturepremise can be extended to other areas of the pipeline repair, and the theory methodwill be used for more extensive field to improve the reference.