PARAMETERS ESTIMATION FOR TIMED AND CONTINUOUS PETRI NETS: APPLICATION TO THE IDENTIFICATION AND MONITORING OF HYBRID SYSTEMS
Dimitri Lefebvre, Philippe Thomas · Cybernetics & Systems · 2005
Petri net (PN) models are useful tools for the modeling of discrete event systems and hybrid systems. This paper is about the parameters estimation for two classes of timed PNs: T-timed PNs and continuous PNs with variable speeds. The parameters to be estimated are either incidence or temporal parameters. Data- and model-based methods are proposed. In each case, on-line and off-line implementations are detailed. As a consequence, identification and monitoring applications for hybrid systems are described.