中国机械工程 ›› 2022, Vol. 33 ›› Issue (19): 2340-2346.DOI: 10.3969/j.issn.1004-132X.2022.19.008

• 智能制造 • 上一篇    下一篇

多维动态贝叶斯网络及其重要度分析方法

陈东宁1,2;胡彦龙1,2;姚成玉3;王宽通1,2;马雷1,2   

  1. 1.燕山大学河北省重型机械流体动力传输与控制重点实验室,秦皇岛,066004
    2.先进锻压成形技术与科学教育部重点实验室(燕山大学),秦皇岛,066004
    3.燕山大学河北省工业计算机控制工程重点实验室,秦皇岛,066004
  • 出版日期:2022-10-10 发布日期:2022-10-20
  • 作者简介:陈东宁,女,1978年生,副教授、博士研究生导师。研究方向为可靠性分析及优化。E-mail:dnchen@ysu.edu.cn。
  • 基金资助:
    国家自然科学基金(51975508);河北省自然科学基金(E2021203061)

Multi-dimensional Dynamic Bayesian Network and Its Importance Measure Analysis Method

CHEN Dongning1,2;HU Yanlong1,2;YAO Chengyu3;WANG Kuantong1,2;MA Lei1,2   

  1. 1.Hebei Provincial Key Laboratory of Heavy Machinery Fluid Power Transmission and Control,
    Yanshan University,Qinhuangdao,Hebei,066004
    2.Key Laboratory of Advanced Forging & Stamping Technology and Science(Yanshan University),
    Ministry of Education of China,Qinhuangdao,Hebei,066004
    3.Key Laboratory of Industrial Computer Control Engineering of Hebei Province,Yanshan University,
    Qinhuangdao,Hebei,066004
  • Online:2022-10-10 Published:2022-10-20

摘要: 贝叶斯网络分析方法是可靠性分析的重要方法,但传统贝叶斯网络分析方法局限于分析单因素影响,当系统可靠性受多因素影响时会产生较大分析偏差。为此,提出多维动态贝叶斯网络分析方法,借助单位阶跃函数与冲激函数进行贝叶斯网络时间连续化构造,建立根节点受多因素影响时系统的失效概率分布函数。在此基础上,对传统重要度分析方法进行多维扩展,提出多维动态贝叶斯网络重要度分析方法。通过对斗轮机张紧机构液压系统进行工程实例分析,并与离散时间贝叶斯网络分析方法分析结果对比,验证了多维动态贝叶斯网络及其重要度分析方法的可行性和优越性,为系统改进与薄弱环节识别提供了更为准确的量化依据。

关键词: 多维动态贝叶斯网络, 重要度, 可靠性分析, 液压系统

Abstract: The Bayesian network analysis method was an important method of reliability analysis, but the traditional Bayesian network analysis method was limited to analyze the influences of single factors, and there was a large analysis deviation when the system reliability was affected by multiple factors. Therefore, a multi-dimensional dynamic Bayesian network analysis method was proposed, which used unit step function and impulse function to construct Bayesian network time continuity, and established the failure probability distribution function when the root node was affected by multiple factors. Then, a multi-dimensional dynamic Bayesian network importance measure analysis method was proposed by expanding the traditional importance measure analysis method. Through the engineering example analyses of the hydraulic system of the bucket wheel machine tensioning mechanisms, and compared with the analysis results of the discrete-time Bayesian network analysis method, the feasibility and superiority of the multi-dimensional dynamic Bayesian network and its importance measure analysis method were verified, which provides a more accurate quantitative basis for system improvement and weak link identification.

Key words: multi-dimensional dynamic Bayesian network, importance measure, reliability analysis, hydraulic system

中图分类号: