Design of an Improved Think Globally Act Locally Approach for the Computation of Petri Nets Based Liveness Enforcing Supervisors of FMSs

An improved think-globally-act-locally (ITGAL) method is proposed in this paper for the computation of a liveness enforcing/deadlock prevention supervisor containing of a set control places (CPs) for a Petri net (PN) model of a flexible manufacturing system (FMS) suffering from deadlocks. The proposed method is especially suitable for generalized PN classes containing weighted arcs such as S4R and S4PR. It leads to optimal or near-optimal liveness-enforcing supervisors without solving intractable integer linear programming problems. By using a recently proposed optimality test for CPs, the proposed ITGAL method provides improved behavioral permissiveness and/or reduced structural complexity of the CPs. The applicability of the proposed method is shown by means of a number of typical FMS examples.