基于证据理论和贝叶斯网络的列控车载子系统可靠性分析

Reliability analysis of on-board subsystem of train control system based on evidence theory and Bayesian network method

针对在列控车载子系统可靠性分析中存在不确定信息、共因失效、恢复机制和降级运行的问题,采用融合证据理论的贝叶斯网络模型对其进行可靠性分析。在分析CTCS-3级列控车载子系统结构的基础上构建故障树,并将故障树转化为贝叶斯网络。利用贝叶斯网络的正向推理能力,综合考虑共因失效和恢复机制因素,结合信任测度和似然测度得到列控车载子系统的可用度并讨论降级运行对系统可用度的影响。利用贝叶斯网络的反向推理能力,得到车载子系统薄弱环节。通过求解重要度,得到存在认知不确定性的基本单元对车载子系统可靠性的影响程度。研究结果表明:该方法增强了贝叶斯网络处理不确定信息的能力,运用该方法得到的评价结果比完全信息条件下得到的更符合实际情况。

To deal with uncertain information,common cause failure,recovery mechanism and degraded operation in the reliability analysis of on-board subsystem of train control system,the Bayesian network model integrating evidence theory was used to analyze its reliability.Based on the analysis of the structure of on-board subsystem of CTCS-3 train control system,the fault tree was constructed and transformed into Bayesian network.By using the forward inference ability of Bayesian network,considering the factors of common cause failure and recovery mechanism,the availability of on-board subsystem was obtained by combining the belief measure and plausibility measure.And the effect of degraded operation on system availability was discussed.By using the backward inference ability of Bayesian network,the weaknesses of on-board subsystem were recognized.By solving the magnitude of importance,the effects of basic units with epistemic uncertainties on the system are obtained.The results show that this method enhances the ability of Bayesian network to deal with uncertain information,and the evaluation results obtained by this method are more accordance with the actual conditions than those obtained under the condition of complete information.