基于面向对象贝叶斯网络的船用锅炉炉膛安全性定量分析  

Quantitative Safety Analysis of Hearth of Warship Boiler Based on Bayesian Networks

作　　者：初珠立[1] 杨自春[1] 

机构地区：[1]海军工程大学船舶与动力学院,武汉430033

出　　处：《船海工程》2010年第5期92-95,共4页Ship & Ocean Engineering

基　　金：总装武器装备预研基金(9140A27050106JB11)

摘　　要：在对船用锅炉炉膛进行安全性分析时发现,故障隔离功能丧失在部件之间存在重复故障片段。借鉴面向对象贝叶斯网络建立故障模型,采用基于Bayes的蒙特卡罗仿真得到底事件概率先验信息。与故障树等传统方法比较,该模型直观简洁,有效缩小关键故障因子容量,诊断更加准确。Taking the lost of insulate function as a case of the hearth of ship boiler,the quantitative safety analysis was carried out based on the transition of FTA to Bayesian networks.This technique broke through the short of duality in the FTA theory,with multi-conditions of every node.Meanwhile,the method took the relationships among them in top account,sought for parameters of the nodes before the quantitative calculate.The results showed that this technique is more reasonable and of more engineering practicability.

关 键 词：锅炉炉膛 贝叶斯网络 安全性定量分析 面向对象 

分 类 号：U664.1[交通运输工程—船舶及航道工程] TK229.7[交通运输工程—船舶与海洋工程]