A numerical simulation approach for reliability analysis of fault-tolerant repairable system

This paper proposes a numerical simulation approach for reliability analysis of fault-tolerant system with repairable components. In the traditional method for the reliability analysis of fault-tolerant system, the system structure is described by means of binary decision diagram (BDD) and Markov process, and then the reliability indexes are calculated. However, as the size of system augments, the size of state space will increase exponentially. In addition, Markov approach requires that the failure and repair time of the components obey exponential distribution. In this paper, by combining dynamic fault tree (DFT) and numerical simulation based on the minimal sequence cut sets (MSCS), we propose a new method to evaluate reliability of fault-tolerant system with repairable components. The approach presented does not depend on Markov model, so that it can effectively solve the problem of the state-space combination explosion. Moreover, our method does not require the system to have Markov property, and it is suitable for systems whose failure and repair time obey arbitrary distributions. Therefore, our method is more flexible than traditional method. At last, one example is given to verify the method.