A NEW FAULT TREE ANALYSIS METHOD: FUZZY DYNAMIC FAULT TREE ANALYSIS
Yan-Feng Li,Hong-Zhong Huang,Yu Liu,Ningcong Xiao,Haiqing Li
2012-01-01
Eksploatacja i Niezawodnosc - Maintenance and Reliability
Abstract:Fault tree analysis (FTA) is a widely used reliability assessment tool for large and complex engineering systems. The conventional fault tree analysis method, which contains AND, OR, and Voting gates, etc., can efficiently build an analytical model to represent combinations of component failures that cause the failure of a system. However, due to its limited modeling capability, we may confront difficulties when modeling dynamic systems which involve complicated dynamic characteristics such as sequence dependency and functional dependency. Markov-based dynamic fault tree analysis (DFTA) extends the static FTA by introducing additional gates to model such complicated interactions among events. In many circumstances, it is quite difficult to obtain an accurate system reliability estimate due to limited data. To overcome this issue, a fuzzy dynamic fault tree model is put forth to assess system reliability. To obtain the membership function of the fuzzy probability for the top event of the studied fault trees, the extension principle is employed to calculate the associated membership function via a pair of parametric programming problems. Finally, a case study is presented to demonstrate the application of the proposed approach for the hydraulic system of a CNC machining centre.