Performance of Static Slicing Algorithms for Petri Nets

Reggie Davidrajuh and Albana Roci (University of Stavanger, Norway)

This paper focuses on algorithms for static slicing of Petri Nets. This paper presents the implementation of some of these algorithms on the General-purpose Petri Net Simulator (GPenSIM). Also, a new place-invariant based algorithm for static slicing is presented which is more robust to changes in the selection criterion, thus more suitable for model verification stages. However, the experiments presented in the latter part of this paper suggest that all the known static algorithms are impractical for verification of Petri Net models of large real-life systems. Since most of the real-life discrete-event dynamic systems (notably, manufacturing systems) are cyclic (or repetitive), and the Petri Net models of those systems are event graphs. The existing static slicing algorithms are not useful for event graphs as they return the entire Petri Net as the slice. The other type of slicing algorithms - the dynamic algorithms - only works for specific instances of Petri Nets; thus, the applicability of dynamic slicing for model checking is also limited. Therefore, this paper concludes that more work is needed on the static slicing for model verification of large real-life discrete-event systems.

Journal: International Journal of Simulation- Systems, Science and Technology- IJSSST V20

Published: Jan 30, 2019

DOI: 10.5013/IJSSST.a.20.S1.15