江汉大学学报（自然科学版）2026，Vol.54Issue(3)：15-27,13.DOI:10.16389/j.cnki.cn42-1737/n.2026.03.002
基于Bow-Tie和贝叶斯模型的海底管道动态风险评估
Dynamic Risk Assessment of Subsea Pipelines Based on Bow-Tie and Bayesian Models
摘要
Abstract
The operational environment of subsea pipelines is complex,and the failure paths and modes are diverse.To address this issue,a comprehensive probabilistic risk analysis model integrating multiple failure modes and accident evolution processes is proposed.Initially,a failure-consequence Bow-Tie model for subsea pipelines was established by analyzing the logical relationships among risk factors,failure mechanisms,and safety barriers.Subsequently,a Bayesian network model was constructed according to the mapping rules of the Bow-Tie model.Node states and prior probabilities were determined to analyze the accident evolution paths and identify key influencing events.Finally,the states of key events were updated based on Bayesian inference,thereby enabling the dynamic updating of the probability of subsea pipeline accidents.The results indicate that corrosion-induced perforation is the dominant failure mode during the long-term operation of pipelines.Third-party activities,corrosion protection failure,material defects,and human operational errors are the key events leading to pipeline failure.In addition,inadequate inspection and maintenance strategies,poor cleaning practices,emergency shutdown system failures,and human operational errors are the critical events leading to safety barrier failures.Therefore,enhanced monitoring and maintenance of these critical events are required to improve emergency response capabilities.关键词
海底管道/贝叶斯网络/动态风险分析/多失效模式/贝叶斯推断Key words
subsea pipeline/Bayesian network/dynamic risk analysis/multi-failure modes/Bayesian inference分类
海洋科学引用本文复制引用
李昱捷,邢莉莎,肖昕雨,周晓光,李志..基于Bow-Tie和贝叶斯模型的海底管道动态风险评估[J].江汉大学学报（自然科学版）,2026,54(3):15-27,13.基金项目
国家自然科学基金项目(52109416) （52109416）
湖北省教育厅科学研究计划青年人才项目(Q20244412) （Q20244412）
