Abstract:

Based on a MonteCarlo (MC) estimation algorithm of multi state network reliability, the author considered an M-C estimation algorithm based on fusion of hierarchical sampling and dynamic sampling. The hierarchical sampling was realized by setting the probability threshold α to change hierarchical principle based on the hierarchical sampling method of state tree search. Using dynamic sampling, the capacity value of each side of the network was dynamically generated when the network was invalid, so that the invalid network state could be generated without the sampling of all edges, and the simulation time was shortened. Simulation results show that the dynamic sampling can shorten the simulation time, but the advantage will gradually disappear with the increase of the network reliability, and it is more suitable for the multi state network with low reliability.

Key words: network reliability, multi state network, Monte-Carlo estimation