An optimization approach for identification of Petri Nets

The paper addresses the identification problem of discrete event systems by determining the structure and the initial marking of a Petri Net (PN) modeling the system. More precisely, we assume that the numbers of places and of transitions are given and that a finite sequence of transitions and the corresponding markings are completely or partially known. Hence, the conditions to univocally identify a pure PN are established. On the other hand, if the singleness of the identification problem solution can not be guaranteed, we introduce an approach based on the solution of an integer linear programming problem. The linear constraint definition utilizes the knowledge of the observed firing sequence and the properties imposed on the PN.