Optimization of manufacturing systems modelled by timed Petri nets

Abstract

Integer timed Petri nets (ITPN) are applied in this work to model automated manufacturing systems, where several products are obtained via sequences of operations executed by different machines. In such a model the system state is represented by the joint information consisting of the net marking and of the residual firing times of timed transitions. On this basis, the state equations representing the system dynamics can be written. With reference to this model, it is possible to build the system state diagram, where two main classes of states can be distinguished, named tangible and vanishing, after the terms introduced for GSPN. Then, a specific performance optimization problem is addressed for which the cost to be optimized can only be associated with the permanence of the system in tangible states.