基于模糊故障树和贝叶斯网络的FPSO碰撞风险分析方法研究  

Collision risk analysis method of deepwater general FPSO based on fuzzy fault tree and Bayesian network

作　　者：刘昆[1] 纵帅 王加夏 卜淑霞[2] 王秀飞 LIU Kun;ZONG Shuai;WANG Jia-xia;BU Shu-xia;WANG Xiu-fei(School of Naval Architecture and Ocean Engineering,Jiangsu University of Science and Technology,Zhenjiang 212003,China;China Ship Scientific Research Center,Wuxi 214082,China)

机构地区：[1]江苏科技大学船舶与海洋工程学院,江苏镇江212003 [2]中国船舶科学研究中心,江苏无锡214082

出　　处：《船舶力学》2022年第4期574-583,共10页Journal of Ship Mechanics

基　　金：国家自然科学基金项目(51779110;51809122;52171311;52101359);江苏省自然科学基金项目(BK20191461;20190966);工信部高技术船舶科研项目(深水通用型FPSO工程开发);江苏省研究生创新计划项目(SJCX21_1778);绿色智能内河船舶创新专项。

摘　　要：穿梭油轮与FPSO碰撞事故是FPSO油气外输过程中常见的灾害事故之一。为将碰撞事故的可能性降至最低,并为风险管理提供决策依据,本文将模糊故障树与贝叶斯网络(BN)结合,提出了一种基于模糊故障树与贝叶斯网络的FPSO油气外输系統碰撞风险评估方法。依据模糊故障树与贝叶斯的映射算法,构建BN模型及确定条件概率表(CPT);运用BN网络的推理算法计算事故发生的概率和节点的后验概率进行故障诊断,并且可以依据节点的重要度对结果进行故障的分析排查。最后,通过FPSO外输油气系统工程验证算法的可行性及有效性。The collision between a shuttle tanker and an FPSO is one of the common disasters in the process of FPSO oil and gas export.In order to minimize the possibility of collision accidents and provide decision-making basis for risk management,a risk assessment method of FPSO oil and gas export system collision based on fuzzy fault tree and Bayesian network(BN)was proposed by combining fuzzy fault tree and BN.Based on the mapping algorithm of fuzzy fault tree and Bayes,BN model and conditional probability table(CPT)were constructed.The BN inference algorithm was used to calculate the probability of the occurrence of accidents and the posterior probability of nodes for fault diagnosis,and the results were analyzed according to the importance of nodes to troubleshoot faults.Finally,the feasibility and effectiveness of the algorithm were verified by FPSO oil and gas transport system engineering.

关 键 词：FPSO外输系统 碰撞风险评估 模糊故障树 贝叶斯网络 重要度 

分 类 号：U661.4[交通运输工程—船舶及航道工程]