Clasificación de fallas en líneas de transmisión usando señales dispersas basado en el reconocimiento de patrones para mejorar la seguridad en los sistemas eléctricos de potencia

Abstract

In this document, the classification of types of failures in transmission lines will be carried out, which exist within electrical power systems, the classification will help to have a better way of identifying the different types of failures that often occur. The study was carried out with the help of the DigSilent PowerFactory program that helped to generate the necessary data for each type of failure. Through the methodology proposed in this document, it was possible to classify the different types of failures with the help of compressed sensing that allows the reconstruction of the signals obtained from the simulation carried out and helps with the recovery of one-dimensional signals considering a minimum amount of data obtained from the original signal. The recognition of patterns is part of the classification of types of failures, since it helps to identify the type of failure that occurs on the transmission lines, and with the implementation of the decision tree algorithm, it will be possible to determine the state of failure or non-failure that occurs on each of the lines. All this so that there is an efficient, safe performance and with an immediate classification against the different types of failures that occur.