Physical diagnostic for profibus DP networks based on Artificial Neural Network

PROFIBUS (Process Field Bus) DP is the most used field communication bus in the worldwide industry. The increasing use of this fieldbus network protocol in industrial plants has made quick diagnostics for fails extremely necessary and important in order to minimize halt time during installation and the consequent financial losses in the production process. This work describe a tool based on Artificial Neural Networks (ANN), which will be used to promptly diagnose the physical layer of a PROFIBUS network in case of a failure, providing a criterial performance analysis through the basic concepts presented. In this context, this paper will briefly describe the physical layer of the mentioned protocol and the related problems, as well as the results from the performed ANN training.