A Bayesian Belief Network Based Method for Performance Evaluation and Troubleshooting of Multistate Systems

This article proposes a new method for evaluating the performances of a class of industrial multistate systems, i.e., production units with variable instantaneous output. The method is based on the construction of a Bayesian Belief Network. The main advantage of the approach is that the model can be used for troubleshooting, as well as for evaluation. The resolution of a test case illustrates the use of the method.