Modeling of discrete event systems: A holistic and incremental approach using Petri nets

  • Authors:

  • Carmen-Veronica Bobeanu;Eugene J. H. Kerckhoffs;Hendrik Van Landeghem

  • Affiliations:

  • Ghent University, Technologiepark, Ghent, Belgium;Delft University of Technology, Mekelweg, Delft, The Netherlands;Ghent University, Technologiepark, Ghent, Belgium

  • Venue:
  • ACM Transactions on Modeling and Computer Simulation (TOMACS)
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this article, the authors provide an alternative view on Petri nets modeling of discrete event systems. The proposed modeling procedure follows the Systems Specification guidelines underlying the well-known DEVS modeling formalism. The authors' endeavour is towards perfecting the design of reusable Petri nets-based models by searching for a good primitive for a modular model construction and the introduction of coupling templates as standardised means to couple building block components. Assuming that the real-world system to be modeled has been analyzed in depth beforehand through a suitable system analysis method (which itself is beyond the scope of the article), we present a systematic step-by-step approach to construct a model in the Petri nets domain together with its experimental frame. The construction adheres to well-defined rules, which enable computer-based model construction. The input for this systematic bottom-up construction of Petri nets-models is information (about, e.g., primitive system components, entity flows, routing constructs) gathered from the top-down system analysis. In the article, attention is also paid to the algebraic backgrounds underlying the proposed model construction. These provide the basis for formal correctness proofs, mapping Petri nets onto DEVS-models, and complexity reduction of the found Petri nets-models. By offering to the model builder the possibility to handle multiple abstraction levels and by addressing important issues related to the interfacing question of coupled models and model components described in Petri nets and DEVS formalism, the authors' work addresses two of the main research directions of Computer Automated Multi-Paradigm Modeling ([Mosterman and Vangheluwe 2002]): model abstraction and multiformalism modeling. The article concludes with an illustrative application example.