Fault detection in acyclic Petri net models

This paper proposes an approach for fault detection in a system that is modeled as an acyclic Petri net. Places in the net can be either observable or unobservable, and faulty behavior in the system are modeled as faulty transitions in the net. Our goal is to determine whether faulty transitions occur or not based on the observation of system state (marking) changes. We develop an algorithm that is able to determine the possible occurrences of faulty transitions in the net. An example is also provided to illustrate our approach.