Applying multi-agent systems to real-time industrial automation

摘要

This work presents a framework which combines distributed real-time objects with adaptive scheduling algorithms that support modes and mode changes for developing multi-agent systems to real-time industrial automation applications. In the proposed model, agents sense environmental changes and request changes in operational modes, which are executed by real-time distributed objects. The paper describes briefly the previous work developed by the authors, on top of which the real-time multi-agents architecture is built: (i) an integrated environment for modeling, simulation, and code generation for distributed real-time objects and (ii) real-time adaptive scheduling algorithms that allows modes changes. The work is at a preliminary stage and the paper basically present the overall strategy to be followed in the ongoing project.