Fuzzy reliability using a discrete stress-strength interference model

This paper uses a fuzzy line-segment method to calculate the fuzzy unreliability of a system when only discrete-interval probabilities of stress and strength inside an interference region are available. The discrete-interval probabilities are treated as fuzzy numbers. A stress-strength interference model and extended operations of fuzzy numbers are used to calculate the fuzzy unreliability. Probability density functions are approximated by piecewise fuzzy line-segments that are expressed by linear fuzzy polynomials; the centroid of the membership function for the fuzzy unreliability is treated as the point estimate of unreliability. This method requires less computer time than previous methods in the literature, and is useful since it does not require information on distribution types of stress and strength. Numerical examples demonstrate the method