Fault modeling and reliability evaluations using artificial neural networks

In this paper, we present a generalized Markov reliability and fault-tolerant model (including the effects of permanent fault, transient fault and intermittent fault) for reliability evaluations based on the neural network techniques. The desired reliability of the system under design is fed to the neural network and when the neural network converges the parameters of the design are extracted from the weights of the neural network. We also obtained the simulation results which are in agreement with the classical analysis