GENERALIZED STOCHASTIC PETRI NETS FOR THE PERFORMANCE EVALUATION OF FMS.

A class of Petri nets in which transitions can fire either immediately or after an exponentially distributed random time since their enabling generalized stochastic Petri nets (GSPN), is used to define performance models for flexible manufacturing systems (FMSs). An example is discussed in which the analysis of the GSPN model yields results that are more reliable than those obtained with a currently used method based on queuing networks. GSPN models are shown that allow accurate analyses of the effects that different scheduling policies and pallet mixes have on system performance.