Abstract: In order to obtain the dynamic change law of reliability of redundant repairable system under the influence of common cause failure, a new algorithm combining improved GO method with dynamic bayesian network was proposed. Firstly, the GO diagram model was established according to the schematic diagram and flow chart of the system. Then, the GO model was converted into a DBN by using the conversion mapping rules of each operator and signal flow to the DBN. Finally, the model was solved by using DBN mature software and algorithms. After verification and comparison, it can be seen that this method is not only able to obtain the dynamic curve of system reliability and availability changing with time, but also to treat the influence of maintenance factors and common cause failures on the analysis results. In additional, that this method with the powerful ability of reverse reasoning, is able to obtain the weak links quickly and provide reference the fault diagnosis for the system

Key words: GO method, Dynamic bayesian network, Redundant systems, Reliability analysis

摘要： 为获得冗余可修系统在共因失效影响下的可靠性动态变化规律，提出了一种将改进GO法与动态贝叶斯网络相结合的新算法。首先，根据系统的原理图与流程图建立GO图模型；然后，基于信号流与各操作符向动态贝叶斯网络的映射规则，将GO图模型转化为动态贝叶斯网络；最后，凭借动态贝叶斯网络成熟的软件与算法对模型进行求解。通过对算例的验证和比较可以得出，该方法不但能够得到冗余系统的可靠度与可用度随时间变化的动态曲线，还能处理维修因素与共因失效对分析结果的影响。此外，新算法具有强大的反向推理能力，可迅速获得系统的薄弱环节，为实现系统的故障诊断提供理论支持

关键词: GO法, 动态贝叶斯网, 冗余系统, 可靠性分析