Studies on applications of Markov chain Monte Carlo in large-scale system reliability evaluation

A new Monte Carlo simulation method for large-scale system reliability evaluation is presented in this paper, which is Markov Chain Monte Carlo (MCMC). In this method, the Markov chain is generated by Gibbs sampler. The Gibbs sampler utilizes a set of full conditional distributions associated with the target distribution of interest in order to define a Markov chain with an invariant distribution equal to the target distribution. The system states are sampled one by one from this Markov chain. Comparing with the classical Monte Carlo simulation method, the relativities between these states are considered. So, MCMC is a kind of dynamic Monte Carlo simulation method and can reflect the inherence of the system states. The results of the IEEE RTS 24-bus test system show that the proposed method is efficient in large-scale system evaluation. The proposed model improves the convergence, stability and computation speed of the reliability evaluation dramatically.