An original Petri net state estimation by a reduced Luenberger observer

The industrial processes considered can be represented by a generalised Petri net model. An approach for Petri net (PN) state estimation and verification is proposed and experimental results are given. This methodology is based on a reference model, for partially known discrete event systems modeled by generalised Petri nets. With the help of the theory of generalised state-space systems, an extended reduced Luenberger observer is proposed to reconstruct the marking and the firing count, and gives an alternate way for PN state computation. The necessary and sufficient existence conditions of this estimator are given and proven. This diagnosis methodology is implemented on a basic example, and simulation results show the efficiency of this approach