基于贝叶斯网络的分层系统可靠性分析

由于分层系统的系统结构复杂,且系统不同层次可靠性信息具有不均衡性的特点,使得分层系统的可靠性建模较为困难.考虑二态情况,通过引入贝叶斯网络描述分层系统元素间的失效关系,并考虑了系统元素间级联失效问题,结合贝叶斯多源信息融合方法,以贝叶斯网络模型为基础整合了系统各层元素的可靠性信息,构成了一种较为通用的二态分层系统可靠性分析方法.算例结果表明:该方法在贝叶斯网络模型计算过程中能够融合系统各层元素可靠性信息并有效处理级联失效问题,提高了系统可靠性后验均值估计的准确性,减小了系统可靠性结论的后验方差. 

Reliability analysis of hierarchical systems based on Bayesian network

Reliability modeling of hierarchical systems is significantly difficult because of the complex system structures and imbalanced reliability information at different system levels. A Bayesian network reliability analysis methodology was proposed to model the reliability of binary-state hierarchical systems considering the combination of Bayesian network and Bayesian information aggregation approach to represent the failure relationship among system elements. An example illustrated that, the proposed methodology can significantly improve the accuracy of system-level reliability modeling and reduce the variances of the posterior failure distribution of system by considering the cascading failure dependency and utilizing all available reliability information throughout the system.