Regresar

Control and supervision system development with intelligent agents

Abstract:

In this paper, we present a development approach for control and supervision systems of industrial processes using intelligent agents. The control and supervision systems are defined as MultiAgent Systems (MAS). For supporting these MAS, a functional architecture is presented, which is based on industrial automation platform. Control and supervision tasks are conceived specialized MAS, in order to coordinate, execute and evaluate functions needed for the distributed information processing and decision-making. Control and supervision agents interact with production units in industrial environments, modeled as MAS through logical and functional abstractions of the real processes. In this way, cooperation and negotiation for improving the production performance are achieved through the adequate interaction among the agents community. The agents are built using MASINA (MultiAgent System for INtegrated Automation) methodology and UML.