基于动态贝叶斯网络的甲醇船对船加注风险评估
Methanol ship-to-ship bunkering risk assessment based on dynamic Bayesian network
张元元 1张春昌 2孙虎 1李根 1李世博1
作者信息
- 1. 上海海事大学商船学院,上海 201306
- 2. 上海海事大学商船学院,上海 201306;船舶与海洋工程特种装备和动力系统国家工程研究中心,上海 201306
- 折叠
摘要
为降低甲醇船对船加注泄漏的风险,构建动态贝叶斯网络模型对甲醇船对船加注泄漏风险进行评估,同时结合蝴蝶结模型、模糊集理论和遗漏概率模型对这类事故的致因和后果进行分析.结果表明,甲醇船对船加注泄漏事故的发生概率为0.040 10,导致这类事故发生的主要因素为软管及法兰损坏.事故发生后发生闪火/池火的概率为0.003 57,发生喷射火的概率为0.004 01,发生蒸气云爆炸的概率为3.61 ×10-5.
Abstract
In order to reduce the risk of methanol ship-to-ship bunkering leakage,a dynamic Bayesian network model is constructed to evaluate the risk of methanol ship-to-ship bunkering leakage,and the causes and consequences of the accidents are analyzed by combining the bow tie model,the fuzzy set theory and the omission probability model.The results show that,the occurrence probability of methanol ship-to-ship bunkering leakage accidents is 0.040 10,and the main factors leading to the accidents are hose damage and flange damage.The occurrence probability of flash/pool fire after the accidents is 0.003 57,the occurrence probability of jet fire is 0.004 01,and the occurrence probability of vapor cloud explosion is 3.61 × 10-5.
关键词
船对船加注/动态贝叶斯网络/蝴蝶结模型/模糊集理论Key words
ship-to-ship bunkering/dynamic Bayesian network/bow-tie model/fuzzy set theory引用本文复制引用
出版年
2024
NETL
NSTL
万方数据