Research On Pipeline Predictive Maintenance Considering Random Correlation Between Degradation And Impact Process

As a common way of energy transmission,natural gas pipeline transportation has developed more mature at this stage and has become an indispensable part of the industrialization process.The natural gas pipeline covers a wide area,but it cannot be repaired or replaced in time due to the limitation of human and material resources.Even corrosion,temperature difference,welding defects and other factors will lead to cracks.External shocks such as pressure and earthquake will lead to further expansion of cracks.Therefore,this paper develops a predictive maintenance strategy driven by reliability-centered data to manage the health prediction of natural gas pipelines to ensure their safe and reliable operation.The main contents of this paper are :Firstly,aiming at the problem that natural gas pipelines are vulnerable to environmental impact and cracking,the random correlation between degradation and impact is considered.The natural degradation process of cracks and the crack propagation caused by external shocks are analyzed respectively.The natural degradation process is modeled as Gamma process,and the external shock process is modeled as homogeneous Poisson process.On this basis,the variable threshold reliability prediction model of pipelines is established.At the same time,the Bayesian parameter update method is used to update the crack propagation rate and variance parameters of the pipeline to analyze the reliability of the natural gas pipeline.;Secondly,considering the influence of random impact on natural gas pipeline cracks,on the basis of pipeline reliability analysis,a data-driven fixed-cycle predictive maintenance strategy is developed,and the corresponding maintenance cost model is established.Under the condition of ensuring its reliable operation,the maintenance strategy threshold is optimized,and the optimal predictive maintenance threshold and detection cycle threshold are obtained.At the same time,considering the influence of imperfect maintenance on pipeline cracks,the uncertainty of imperfect maintenance effect is introduced into predictive maintenance strategy,predictive maintenance strategy under imperfect maintenance is formulated,and a new maintenance cost model is established.The optimal predictive maintenance threshold and imperfect maintenance threshold are obtained by threshold optimization,which can ensure the reliable operation of natural gas pipeline and the lowest maintenance cost rate.Finally,taking the multi-path crack degradation data as an example,the reliability simulation experiments of natural gas pipelines are carried out under different parameters and operating conditions,and the sensitivity analysis of the maintenance threshold and maintenance cost in the developed predictive maintenance strategy is carried out.The maintenance threshold at the lowest cost rate is obtained by experimental comparison,which verifies the validity and accuracy of the model.