Approach to Integrate Fuzzy Fault Tree with Bayesian Network

Abstract

Fuzzy fault tree (FFT) can offer an efficient method of representing the fault causes and handling fuzzy information in the relationships among events. However, FFT cannot incorporate the evidence into the reasoning as Bayesian Network (BN). To overcome the disadvantage of FFT and BN, an approach of integrating FFT with BN is proposed in this paper. Firstly, the FFT technique of Takagi and Sugeno model that can handle uncertainties in the relationships among different events is introduced. Secondly, the translation rules of converting FFT into BN are presented. The integration algorithm is then demonstrated on an offshore fire case study.