Modelling of failure prediction Bayesian network based on fault tree analysis

Based on the useful failure relationships embedded in the fault tree analysis (FTA), the paper proposes a failure prediction Bayesian Network (FPBN) modelling method to establish the practical model and facilitate its applications. In the modelling method, the transformation of events and logic relationships in FTA to variables, edges and conditional probability distributions (CPD) in FPBN are illustrated. Moreover, considering the developments of advanced detection devices, the detection information is also introduced in the FPBN model. At last, the case study of a tank clutch pedal (TCP) is carried out by transforming its FTA to a corresponding FPBN. The results show that, the proposed method can perform effectively in modelling FPBN based on FTA.