Life cycle reliability analysis using imprecise failure data

Abstract

Imprecise life data of machine element represent precise values along with it’s variation proportional to its degree of impreciseness. In this paper, it is attempted to develop a fuzzy Weibull model, where model parameters are in the form of fuzzy variable, with an objective to perform reliability analysis of life cycle data with simultaneous estimation of epistemic uncertainty and deterministic result. Imprecise life data of machine component is modeled as a fuzzy set with triangular membership function. Then entire range of the fuzzy variable is decomposed into a set of $$\alpha$$ -cut intervals corresponding to a various degree of impreciseness. Using these intervals, Weibull model parameters are estimated through the interval least square fitting method. The estimated parameters are in interval form. Integration of these estimated intervals of the model parameters construct a fuzzy variable. Detailed methodology including all processes involved is demonstrated for ball bearing life cycle data.