Detecção, diagnóstico e correção de falha em sensores industriais foundation fieldbus utilizando agentes inteligentes

Abstract

This work performs the development of a system of detection and diagnosis of faults in field sensors with dynamically signal correction, which this system is based on the concept of intelligent agents, using architecture of multagents. The system operates in environment industrial network Foundation Fieldbus. To perform the detection and diagnosis of failures are used the algorithms of neural networks and fuzzy logic (acting like a fuzzy classifier), which are embedded in industrial network Foundation Fieldbus. Furthermore, in order to work with the system of detection, diagnosis and correction of failures was developed a software to perform the monitoring and supervision of the system, which was instantiated an agent communicating with the industrial network. The other agents are instantiated on the functional blocks of industrial network.