System reliabibility measures using fuzzy sets

This paper proposes a new method for system reliability evaluation by redefining the traditional crisp success event as a fuzzy success event. We use the substitute characteristic to define the membership function for the fuzzy success/failure event. Through system function analysis and reliability block diagrams, the system performance variable is expressed as a function of the components performance variables. The membership function for the system fuzzy success event is further evaluated as a function of the system performance variable and utilized for system reliability performance evaluation. A generalized extension principle is proposed to investigate the dynamic behavior for time to fuzzy failure (TTFF) under given stochastic performance degradation processes. Numerical examples illustrate how the proposed method can be applied in system reliability performance evaluation as well as the advantages over traditional system reliability measures.