Implementación de un sistema basado en inteligencia artificial para la detección de regímenes anormales en transformadores de potencia.

The present work has as fundamental topic: “IMPLEMENTATION OF A SISTEM BASED ON ARTIFICIAL INTELLIGENCE FOR THE DETECTION OF ANORMAL REGIMES IN TRANSFORMES OF POWER”. Since there are normal regimes similar to anormal regimes and other less seriores faults, the conventional portections can operate incorrectly. The present work with investigation ends demonstrates that the Artificial Neuronal Net (ANN) created, trained and feigned correctly in Matlab they are able the regimes that show up in the Transformes of Power ti differ. It is achieved in this mediating application the use of LabVIEWa d a system of Acquisition of Data to reproduce the work of a ANN Multilayer Feedforward for this objetive. After having trained the net in Matlab it was implemented in LabVIEW. This Solution was proven their operation with real signs in having reeled primary and Secondary of transformers. The obtained Results Registered by means of dynamic exchange of data in Excel. The Structure is based on the introduction, there chapters, conclusions, recommendations, bibliography and appendixes. The first chapter establishe the tools that allow us to understand the conventional and actual methods that are use to protect the transformers of power against phenomenons that cause transitory regimes. Besides it is presented a Brief summary About the artificial networks neuronal and their foundations. The second chapter has as goal the structuration of the experimental part of a rele that protects a transformer of power against the most Typical regimes of breakdown in a power system. The third shows the Results of the experimentation of the Intelligent digital rele for the protection of transformers of power. The inteligente rele detects the anormal regimes of External and internal short circuits to the transformer of power and it does nor operate incorrectly during the process of magnetization, between other adventages compared with conventional dispositives that are used now.

• 1. Verfasser: Maza Pinza, Julio Geovanny (author)
• Format: bachelorThesis
• Sprache: spa
• Veröffentlicht: 2008
• Schlagworte: <INTELIGENCIA ARTIFICIAL><REGÍMENES ANORMALES><TRANSFORMADORES DE POTENCIA><MATLAB>
• Online Zugang: http://dspace.unl.edu.ec/jspui/handle/123456789/16716