Maximal Good Step Graph Methods for Reducing the Generation of the State Space - Archive ouverte HAL Access content directly
Journal Articles IEEE Access Year : 2019

Maximal Good Step Graph Methods for Reducing the Generation of the State Space

Hao Dou
  • Function : Author
Kamel Barkaoui

Abstract

This paper proposes an effective method based on the two main partial order techniques which are persistent sets and covering step graph techniques, to deal with the state explosion problem. First, we introduce a new definition of sound steps, the firing of which enables to extremely reduce the state space. Then, we propose a weaker sufficient condition about how to find the set of sound steps at each current marking. Next, we illustrate the relation between maximal sound steps and persistent sets, and propose a concept of good steps. Based on the maximal sound steps and good steps, a construction algorithm for generating a maximal good step graph (MGSG) of a Petri net (PN) is established. This algorithm first computes the maximal good step at each marking if there exists one, otherwise maximal sound steps are fired at the marking. Furthermore, we have proven that an MGSG can effectively preserve deadlocks of a Petri net. Finally, the change performance evaluation is made to demonstrate the superiority of our proposed method, compared with other related partial order techniques. INDEX TERMS Petri nets, state explosion problem, covering step graph methods, persistent sets.
Fichier principal
Vignette du fichier
08880692.pdf (8.75 Mo) Télécharger le fichier
Origin : Publisher files allowed on an open archive
Loading...

Dates and versions

hal-02476585 , version 1 (06-04-2020)

Licence

Attribution

Identifiers

Cite

Hao Dou, Kamel Barkaoui, Hanifa Boucheneb, Xiaoning Jiang, Shouguang Wang. Maximal Good Step Graph Methods for Reducing the Generation of the State Space. IEEE Access, 2019, 7, pp.155805-155817. ⟨10.1109/ACCESS.2019.2948986⟩. ⟨hal-02476585⟩
52 View
11 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More