A Novel Model Repair Approach of Timed Discrete-Event Systems with Anomalies

In this paper, the model repair of timed discrete-event systems where anomalies may occur is considered. The nominal model is assumed to be known, and a set of observed timed sequences is given. The approach works with time Petri net models and is based on the formulation of a mixed-integer linear programming problem. The repaired model is obtained from the nominal one by adding fault transitions as well as by extending the firing interval of transitions. Note to Practitioners - The data collected from the observation of a discrete-event system are usually given in terms of behavioral sequences that may be fixed or may be increased in the course of the system operation (e.g., due to new experiments or simply to the system running). If this occurs, the nominal model of a system may reveal not consistent with these additional observations if they include anomalies. This paper presents an approach to repair the model of the system in order to make it able to also generate the observed timed faulty behavior. The repaired model can be used to perform fault diagnosis.