摘要
针对多态元件组成的二态系统可靠性定量分析的难点,提出了一种基于贝叶斯网络的分析方法.该方法首先将多态故障树转化成贝叶斯网络,然后利用该网络计算各事件发生概率和元件重要度.同时利用该贝叶斯网络还可以得到其他一些有意义的结果,可用于系统故障诊断和实时监控.通过一个供电系统实例说明了该方法的有效性.
According to the difficulty of quantitative reliability analysis for binary-state system with multi-state components, a new method based on Bayesian networks is proposed. Multi-state fault tree was converted into Bayesian networks firstly, then the probability of each event and the importance of each component were calculated. And additional information used for failure diagnosis and realtime monitoring was obtained at the same time. The correctness of this method was validated through an example of power supply system.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2009年第6期232-235,共4页
Journal of Harbin Institute of Technology
基金
国家杰出青年科学基金资助项目(70825006)
关键词
贝叶斯网络
多态系统
重要度
故障诊断
供电系统
Bayesian networks
multi-state system
importance measures
failure diagnosis
power supply system