Estimating Reliability of Degradable Computing System using Fuzzy Logic

Abstract

Conventional reliability theory is based mainly on probist reliability, which uses a binary state assumption and classical probabilistic distributions, which is often inappropriate for handling the uncertainty and imprecision of reasoning processes in real world applications due to lack of sufficient probabilistic information. To overcome this situation, Fuzzy rule based methods have been considered useful in the evaluation of reliability of the system as these methods allow the modeling and computation of the heuristic knowledge and linguistic information. In this paper, a four unit gracefully degradable system is developed and analyzed with respect to the effects of repair and coverage factors on its reliability. On the other hand Markov Model has proven its applicability in handling degradation, multiple-failures, imperfect fault coverage and other sequence dependent events, the reliability of the aforesaid system has been computed by incorporating the features of both Markov and Fuzzy Linguistic Modeling approaches.