Fault Tree and Fuzzy D-S Evidential Reasoning combined approach: An application in railway dangerous goods transportation system accident analysis

Abstract In order to manage the railway dangerous goods transportation system (RDNGTS) successfully, an explicit and effective previous accident analysis and accident control approach is essential and necessary. In this study a Fault Tree and Fuzzy D-S Evidential Reasoning combined approach is proposed to analyze the RDNGTS accident, which can solve the uncertainty modeling and information fusion problems existing in RDNGTS accident analysis. The approach has six steps: (i) identify causes of accident and calculate their weights based on Fault Tree, (ii) establish the Fuzzy Belief Structure model of causes of accident, (iii) handle initial qualitative and quantitative data, (iv) fuse the pre-processing data based on Fuzzy D-S Evidential Reasoning algorithm, (v) allocate Confidence Level of fuzzy intersection and, (vi) rank the final Fuzzy Belief Structure of each component based on trapezoidal fuzzy numbers and triangular fuzzy numbers. A historical lithium battery railway transportation accident happened in 2016 in China is applied as the background to examine the approach mentioned in this paper. The results show that professional skills and attitudes of transportation staffs are the weakest component in this lithium battery railway transportation accident. Managers of RDNGTS in China should pay more attentions to the professional skills and attitudes of transportation staffs. Some measures such as improve the awareness of safety and protection, train and examine the professional skills of transportation staffs, may be helpful in curbing the negligent working attitude of transportation staffs. Furthermore, the results also show that D-S Evidential Reasoning could provide a unified modeling framework for uncertain, incomplete, inaccurate and even ignorant information. It could solve the limitations in probability reasoning processes effectively.