Study On Risk Assessment Of Urban Buried Gas Pipeline Based On Bayesian Network

City Gas Pipeline is the city’s major public infrastructure, it is one of the city’s lifeline, its substance has delivered flammable, explosive, poisonous specific hazards, so that it’s safety should not be ignored. Effective urban buried gas pipeline risk assessment to reduce and prevent the occurrence of city buried gas pipeline spill could protect the city operation safely.Paper mainly is based on Bayesian network model of city buried gas pipeline leak failure probability evaluation. In the domestic and international gas pipeline accident statistics analysis and city buried gas pipeline leakage fault tree analysis, based on the summary of the main factors of city buried gas pipeline spill and leak in the pipeline failure possibility Bayesian network for the evaluation of the target model, design and construction factors, corrosion factors, third-party damage, natural disasters into four main aspects of the model, specific to the three parent; use noisy-or gate node to determine the conditional probability of Bayesian network structure powerful processing uncertainty ability to reasonably comprehensive expert disagreement to update the node marginal probabilities. Finally, the possibility of leakage of the buried gas pipeline in Cao Huilu Ch1 segment of Beijing city is calculated by using the Bayesian network model.The key to evaluate the possibility of the leakage failure of urban buried gas pipeline by Bayesian network is the construction of Bayesian network model and the effective aggregation of expert opinion.Paper based domestic and international city gas pipeline leakage accident statistical analysis and fault tree analysis, summarized the main factors for third-party damage, aged improper corrosion, equipment and operation of the pipeline itself is defective, land subsidence, natural factors, human sabotage, pipeline each other Effect of destruction, were constructed according to the structure of the Bayesian network model; use powerful processing noisy-or gate and Bayesian network structure with the uncertainty of the ability to determine the conditional probability of nodes and node marginal probability, complete model construction; With the model of Beijing’s Cao Hui road Ch1 period of gas transmission pipeline section corrosion factors under the influence of leakage failure possibility to carry on the quantitative evaluation, the evaluation results showed that the section of buried gas pipeline leak failure likelihood probability value of 0.735, referring to Chapter 7 failure the possibility of grading table, the corresponding pipeline leakage risk rating for IV level, the evaluation results indicate that the use of Bayesian network model of city buried gas pipeline leak failure may line safety assessment, it is possible for the gas pipeline system of city’s area security management provide an effective basis.