Fuzzy Fault Tree Based Fault Detection

Abstract

In this article, for the Linux operating system environment, with the characteristics of ambiguity and uncertainty for the occurrence probability of system failures, fuzzy theory is introduced into the fault tree analysis. The occurrence probability of basic events in the conventional fault tree is made fuzzy by introducing the concept of fuzzy sets. Using the upstream method for solving the minimum cut sets and transferring the different fuzzy numbers into triangular fuzzy numbers, the method is validated with the CPU error detection. This way provides a theoretical basis and implementation for system reliability evaluation, fault diagnosis and maintenance decisions.