Fault diagnosis in discrete-event systems: incomplete models and learning

Most state-based approaches to fault diagnosis of discrete-event systems require a complete and accurate model of the system to be diagnosed. In this paper, we address the problem of diagnosing faults given an incomplete model of the system. We introduce the learning diagnoser, which estimates the fault condition of the system and attempts to learn the missing information in the model using discrepancies between the actual and expected output of the system. We view the process of generating and evaluating hypotheses about the state of the system as an instance of the set covering problem, which we formalize by using parsimonious covering theory. We also explain through an example the steps in the construction of the learning diagnoser.