现代防御技术 ›› 2021, Vol. 49 ›› Issue (6): 90-97.DOI: 10.3969/j.issn.1009-086x.2021.06.014

• 综合保障性技术 • 上一篇    下一篇

基于区间贝叶斯网络的复杂系统可靠性分析

熊娜1, 梅天宇2, 李飞1   

  1. 1.南昌理工学院,江西 南昌 330044;
    2.湖北宜昌长江三峡游轮中心开发有限公司,湖北 宜昌 443000
  • 收稿日期:2021-04-15 修回日期:2021-07-07 发布日期:2022-01-12
  • 作者简介:熊娜(1985-),女,江西樟树人。副教授,硕士,主要从事工程管理研究。通信地址:330044 江西省南昌市青山湖英雄大道901号南昌理工学院 E-mail:3240845791@qq.com
  • 基金资助:
    军队预研基金资助项目

Reliability Analysis of Complex System Based on Interval Bayesian Network

XIONG Na1, MEI Tian-yu2, LI Fei1   

  1. 1. Nanchang Institute of Technology,Jiangxi Nanchang 330044,China;
    2. Hubei Yichang Yangtze River Three Gorges Cruise Ship Center Development Co.,Ltd.,Hubei Yichang 443000,China
  • Received:2021-04-15 Revised:2021-07-07 Published:2022-01-12

摘要: 鉴于传统故障树模型难以表达事件多态性、模糊性的情况,提出了一种基于区间贝叶斯网络的复杂系统可靠性分析方法。针对故障树模型中事件具有多状态的问题,提出了贝叶斯网络正向推理与反向推理方法,确定了底事件/顶事件处于不同状态下,顶事件/底事件处于不同状态的概率指标;针对底事件故障失效率难以确定的问题,引入三角模糊子集和区间三角模糊子集确定了不同输入条件下复杂系统的可靠性指标。可靠性分析验证了建模的准确性和可行性,在工程实践中具有较大的实用价值,可推广应用于小样本失效数据复杂系统的可靠性分析与故障诊断中。

关键词: 故障树, 多状态, 贝叶斯网络, 复杂系统, 三角模糊子集, 可靠性分析

Abstract: In view of the fact that it is difficult for the traditional fault tree model to express the multi-state and fuzziness of events,a reliability analysis method of complex system based on interval Bayesian network is proposed.Aiming at the problem of multi-state events in the fault tree model,the forward reasoning and reverse reasoning methods of Bayesian network are proposed to determine the probability indexes of the top event/bottom events in different states when the states of the bottom events/top event are determined.Aiming at the problem that it is difficult to determine the failure rates of bottom events,triangular fuzzy subset and interval triangular fuzzy subset are introduced to determine the reliability value of complex system under different input conditions.The accuracy and feasibility of the model is verified by the reliability analysis results,which has great practical value in engineering practice,and can be applied to the reliability analysis and fault diagnosis of complex system with small sample failure data.

Key words: fault tree, multi state, Bayesian network, complex system, triangular fuzzy subset, reliability analysis

中图分类号: