A platform independent communication support for distributed controller systems modelled by Petri nets

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper addresses the distributed execution of Input-Output Place Transition Petri nets (IOPT-nets) models through a network of distributed controllers by providing platform independent modules for communication. Each controller is associated with an IOPT-net sub-model and a set of static communication modules characterized dynamically by the communication meta-model. The sub-models use communication channels to exchange events allowing the global evolution of the IOPT-net model. Whenever controllers are interconnected through some kind of network, communication channels are implemented by platform specific modules following a layered architecture. Communication interfaces, buffers size, routing, etc. are parameterized through the communication meta-model at the application description level, providing the necessary specifications for code to be generated dynamically and therefore support the distributed controller's communication. An application example is introduced, composed by a network of nine distributed controllers interconnected following different topologies and using different communication interfaces. Arduino boards are used together with Ethernet shields and serial lines as implementation platforms for proof-of-concept purposes.
Original languageEnglish
Title of host publicationIEEE International Conference on Industrial Informatics INDIN
Pages88-93
DOIs
Publication statusPublished - 2014
Event12th IEEE International Conference on Industrial Informatics (INDIN) -
Duration: 1 Jan 2014 → …

Conference

Conference12th IEEE International Conference on Industrial Informatics (INDIN)
Period1/01/14 → …

Keywords

  • Arduino
  • IOPT
  • Model-based development
  • Petri nets
  • Platform Independent

Fingerprint Dive into the research topics of 'A platform independent communication support for distributed controller systems modelled by Petri nets'. Together they form a unique fingerprint.

Cite this