LAUSR

In view of the incomplete knowledge and lack of data, the system has epistemic uncertainties. This paper uses Bayesian network fusion evidence theory to analyze the reliability of high-speed railway ATP on-board equipment. According to the requirements of safety-critical system reliability analysis, this paper starts with the analysis of system structure and unit module status, comprehensively considering the influence of uncertain information, common cause failure, recovery mechanism, and degraded operation on system reliability. In this paper, with the advantage of the Bayesian network in the description of events in multiple states, evidence theory is used to reason about the system under incomplete information conditions, obtain the availability interval of on-board subsystems and discuss the impact of the degraded operation on system availability. The α factor model is used to analyze common cause failures, and then the Bayesian network modeling of common cause failures is realized by adding common cause failure nodes. The results show that the method enhances the Bayesian network’s ability to process uncertain information, and the common cause failure data of the train control on-board subsystem is continuously accumulated. The factor model can be used to obtain a more practical common cause failure rate.