Petri nets supervisory control for deadlock prevention and avoidance

This paper deals with supervisory control for deadlock prevention and deadlock avoidance in processes modelled by Petri nets (PN´s). In this paper the problem of deadlocks prevention is solved for a particular class of controlled PN´s, the regular Petri nets, i.e. the PN´s which generate a regular language. For this class of PN´s, the problem can be faced in the Ramadge-Wonham framework by the construction of an automaton that is linguistically equivalent to the given PN and transferring the problem of deadlocks search in PN to a path problem on such automaton. Finally the problem of deadlock avoidance is faced in a more general framework making reference to a PN which is not necessarily regular