Hybrid System based on Fuzzy Inference and Colored Petri Nets to Identify Electrical Fault Events in Real Time

The modeling and validation of a hybrid system of artificial intelligence conformed by a fuzzy inference system and an expert system based on colored Petri nets, to identify events by electrical fault, selecting the useful information and establishing automatically and in real time the previous state of the STE components (System of Transport of Energy), from SOE registries ofSCADA system is presented. The former is part of a project of ISA, the National University of Colombia and COLCIENCIAS whose intention is both designing and implementing an intelligent computational tool that makes the automatic diagnosis of faults (abnormal events) in a STE. The fuzzy inference model is a Mamdani type, with a parameter of output "possibility of event" that is found by means of the defuzzyfication by the centroid method. The system was proven successfully in real SOE registries with thousands of signals of the STE in Colombia from the company Interconexion Electrica S.A. E.S.P. - SA.