A Multiagent Architecture Based in aFoundation Fieldbus Network Function Blocks

The industrial automation is directly related to the technological development of information. Better hardware solutions, as well as improvements in software development methodologies have made possible the rapid development of the productive process control. In this Chapter, it is proposed an architecture that permits to join two technologies in the same hardware (Industrial Network) and software context (Multiagent Systems – MAS). We show a multiagent architecture which uses an algorithm-based Artificial Neural Network (ANN) to learn about fault problem patterns, detect faults, and adapt algorithms that can be used in these fault situations. We also present a dynamic Function Block (FB) parameter exchange strategy which allows agent allocation in fieldbus. This proposed architecture reduces the supervisor intervention to select and implement an appropriate structure of function block algorithms. Furthermore, these algorithms, when implemented into device function blocks, provide a solution at fieldbus level, reducing data traffic between gateway and device, and speeding up the process of dealing with the problem. We also present some examples for our approach. The first one introduces FBSIMU which simulates Foundation Fieldbus function blocks architecture. This software has a controlled process and allocates the MAS to detect and correct faults. The second example shows a multiagent architecture that implements the neural network change in a laboratory test process which imitates fault scenarios. 18