Identificación de fallas en sistemas eléctricos de potencia basado en el reconocimiento de patrones

Abstract

This article identifies electrical failures in the transmission lines of a Power System using a Machine Learning algorithm, using the algorithm of the closest neighbors that will be implemented in Matlab and PSCAD. Machine learning is increasingly applied to electrical systems because they help the study of electrical signals and the development of various applications which are going to become indispensable today. In this way, the power systems with advanced communication equipment are sought to deliver a better quality of energy ensuring that in a disconnection due to failure this sea is resolved in the shortest possible time and the identifier the type of fault causes the necessary corrective measures to be taken to keeping the system stable, knowing what type of fault is the one that occurs is not only important at a high power system level, it is also an industrial level to always control the quality of energy. Failure patterns are acquired from an initial system from which current identification patterns are taken to subsequently simulate cases of transmission line failures.