Implementation of Intelligent System for Interpretation, Classification and Analysis of Abnormal Situations in Industrial Automation Systems at Petrobras
Authors
Kaku Saito
PETROBRAS, Av. Henrique Valadares, Rio de Janeiro
Marcos M. Calôr Filho
PETROBRAS, Av. Henrique Valadares, Rio de Janeiro
Bruno Máximo Menezes
PETROBRAS, Av. Henrique Valadares, Rio de Janeiro
Adler Fonseca de Castro
PETROBRAS, Av. Henrique Valadares, Rio de Janeiro
Eliane Valvano C.S. Mello
PETROBRAS, Refinaria Duque de Caxias – REDUC, Rio de Janeiro
Danilo Curvelo de Souza
Universidade Federal do Rio Grande do Norte, Rio Grande do Norte
Gustavo Leitão
Universidade Federal do Rio Grande do Norte, Rio Grande do Norte
Diego Cabral Silva
Universidade Federal do Rio Grande do Norte, Rio Grande do Norte
Luiz Affonso H. Guedes de Oliveira
Universidade Federal do Rio Grande do Norte, Rio Grande do Norte
This work presents the development and implementation of a system for detecting, interpreting and classifying failures of industrial automation system infrastructures. The work was implemented at Petrobras' Duque de Caxias Refinery, currently covering about 70% of the refinery's infrastructure. The methodology adopted makes it possible to extend the results to other industrial automation equipment, enabling the unification of data in order to compose a large environment of asset reliability. This work addresses the project methodology, the technologies adopted, the solution architecture and the results achieved.
