Intelligent Systems for the Detection of Internal Faults in Power Transmission Transformers

Ivan N. da Silva and Rogerio A. Flauzino


Partial discharge, Fault identification, Fault location, Communication infrastructure, Planning, Operation


This paper presents some methodologies based on expert systems, which are intended to identify and locate internal faults in transmission transformers, as well as to provide an accurate diagnosis (predictive, preventive and corrective) so that proper maintenance is performed. In fact, the main difficulty of using conventional methods, based on analysis of acoustic emission or dissolved gases, lays on how to relate the measured variable with the existence of an internal fault in the transformer. Such situation has made difficult to design optimized systems, because it prevents the efficient localization and identification of possible defects in due time. In addition, there are many cases where the equipment must be turned off for such tests to be carried out. Thus, this paper proposes the necessary procedures for the design of intelligent automated systems that are based on techniques of artificial neural networks and fuzzy inference systems. Based on information of acoustic emission signals, concentration of gases present in the mineral insulating oil and electrical measurements, intelligent systems are able to provide, as final result, the identification, characterization and location of any electrical faults occurring in transformers.

Important Links:

Go Back