故障树分析
贝叶斯网络
动态贝叶斯网络
过程(计算)
紧急疏散
风险分析(工程)
计算机科学
应急管理
工程类
运筹学
可靠性工程
人工智能
医学
海洋学
操作系统
地质学
法学
政治学
作者
Huixing Meng,Xuehui An
标识
DOI:10.1016/j.oceaneng.2021.109928
摘要
The risk in emergency operations can amplify accident losses and threaten the safe achievement of emergency response objectives. In this paper, by considering the dynamic characteristics of emergency operations, we propose an integrated model of estimating the probability of emergency failure by integrating fault tree (FT), dynamic Bayesian network (DBN), and fuzzy set theory. In the hybrid model, FT is utilized to identify the risk-influencing factors of emergency operations. DBN is applied to capture the dynamic features in the emergency process. In presence of limited prior knowledge, the fuzzy set theory is employed to determine the prior probabilities of the root nodes. The methodology is utilized to evaluate the risk of oil recovery operations in the deepwater blowout accident. Particularly, we assessed the dynamic risk of lowering, installation and cutting of emergency equipment, as well as the formation of gas hydrate. The risk-influencing factors of emergency operations and their correlations are identified. The influence of the priority order of the process on the emergency operation is expounded. Eventually, a DBN-based emergency operation model for the deepwater blowout is developed. The model captures the spatial variability of parameters and simulates the evolution of emergency operations over time and space. The mutual information is utilized to conduct sensitivity analysis and diagnostic reasoning on the model.
科研通智能强力驱动
Strongly Powered by AbleSci AI