Fuzzy Reliability Estimation Using Chi-Squared Distribution

In traditional reliability analysis, the failure rate probabilities of components of a system are considered as exact values. But in real world application there exist uncertainty in failure data obtained from historical data or personal judgment. In this paper the fuzzy reliability of a repairable system comprising of four independent and identical modules is estimated, where the failure rate for each module is assumed to follow an exponential distribution. To handle the uncertainty in calculation of failure rates, these parameters are estimated through a fuzzy triangular number. These fuzzy numbers can be calculated by using point estimation and %(1-β) confidence interval of failure-rate parameters. A numerical example is given to illustrate the procedure and validate the result.