A Novel Probability Evaluation Method Based on Fuzzy Fault Dependent Matrix

Abstract

Safety-critical system is a system in which any design error or failure has the potential to lead to high economic costs or loss of life, and the reliable operation of safety-critical system requires high safety of critical data. Therefore, it is necessary to use fault tree analysis or other traditional methods to analyze the availability and integrity of critical data during the design process. However, traditional methods cannot handle the uncertain fault causal relationships of these safety-critical systems, and availability and integrity cannot be evaluated. Here we propose a fuzzy fault dependent matrix to evaluate the probability of failure condition related to uncertain fault causal relationships based on the thinking of statistics. The qualitative and quantitative analysis methods from the perspective of probability evaluation are given. We take a fault-tolerant display unit an example to illustrate the use of fuzzy fault dependent matrix. A Monte-Carlo simulation is carried out to verify the evaluated results, which proved the correctness of the proposed method.