The optimization problem based on alternatives aggregation Petri nets as models for industrial discrete event systems

  • Authors:
  • Juan Ignacio Latorre;Emilio Jiménez;Mercedes Pérez

  • Affiliations:
  • Department of Mechanical Engineering, Energetics and Materials, Public University of Navarre, Campus of Tudela, Tudela, Spain;Higher Technical School of Industrial Engineering, Department of Mechanical Engineering, University of La Rioja, Logroño, Spain;Higher Technical School of Industrial Engineering, Department of Mechanical Engineering, University of La Rioja, Logroño, Spain

  • Venue:
  • Simulation
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

The construction, set-up and operation of many systems of interest in sectors such as industry, supply chains and communications are complex processes, which may require significant investment of resources. For this reason, the automation of the decision making for achieving the best design and operation of such systems, which may be regarded as discrete event systems (DESs), constitutes an active research field. In this paper, we present a methodology to cope with this process in an efficient way, optimizing not only the behaviour of the DES but also its structure. This kind of problem is usually associated with the so-called combinatorial explosion, since the number of alternative configurations for the DES might be huge. We present an improved algorithm to transform a set of alternative Petri nets, representing alternative structural configurations, into a more compact model called an alternatives aggregation Petri net. In real decision-making problems, where the different alternative structural configurations may share common subnets, this compact model may allow the development of a much more efficient optimization problem than the classic approach of 'divide and conquer'. The achievement of this objective is performed by developing a single and compact model for all of the alternative structural configurations of the DES and the simulation of the most promising of them. In this paper, the mentioned methodology is introduced and its advantages and drawbacks are described in relation with the classic approach.