Research on possibility of leakage of overhead gas pipeline in plant area based on FUZZY-BN-FTA model
In order to study the possibility of overhead gas pipeline leakage in the plant area,an ensemble model combining the fuzzy mathematical method,Bayesian network model and fault tree model was proposed.A fuzzy Bayesian network model of overhead gas pipeline leakage accident was established by GeNIe software.The Leaky Noisy-or Gate extension model was in-troduced to modify the conditional probability table of nodes in fuzzy Bayesian network,and the probability value of overhead gas pipeline leakage was predicted by combining the prior probability of basic events.The importance of three basic events was introduced to determine the importance ranking of basic events,and the sensitivity analysis was carried out to determine the main influencing factors and control measures.The results show that the main factors leading to the leakage of overhead gas pipelines are the vehicle impact in plant area,pipeline fatigue loss,poor pipe quality,unreasonable installation design,and welding defects.The research results can provide a reference for the safety management personnel of industrial enterprises to formulate targeted control measures and reduce the accident occurrence.
overhead gas pipeline leakageprobabilities analysisBayesian networktrapezoidal fuzzy numberLeaky Noisy-or Gate extension model