Random Fuzzy Variable Modeling on Repairable System

Repairable system analysis is in nature an evaluation of repair effects. Recent tendency in reliability engineering literature was imposing repair regimes and estimate system repair effects or linking repair to certain covariate to extract repair impacts. Hinted by engineering tune up exercises, we propose a repair model in terms of random variable distribution with a fuzzy parameter because fuzziness reflects the evolution of system dynamic rule changes according to its design specifications. In this paper, we develop an average chance distribution for random fuzzy lifetimes based on the foundational work of self-dual fuzzy credibility measure theory proposed by Liu (2004) and the traditional probability measure theory. We further propose a maximum average chance principle for data-assimilated parameter estimation, which will lead to two empirical distributions – an average chance empirical distribution and an empirical probability distribution with the expected fuzzy parameter as the point estimate for its parameter. The differences between the two filtered lifetimes will facilitate the repair effects. © 2007