Utilización de técnicas de agrupamiento para la obtención de curvas de carga representativas de tipo: residencial, comercial e industrial : Utilización de la técnica de agrupamiento fuzzy k-means para la obtención de curvas de carga representativas de tipo: residencial, comercial e industrial.

Abstract

The following curricular integration work has as its main objective to find a representative load curve through the fuzzy k-means (FKM) clustering technique for three types of consumers: industrial, commercial, and residential; for this purpose, an Excel database containing the power measurements of the three consumption groups in one-hour intervals throughout the day will be used. This database must be treated to obtain more reliable results, since distorted measurements hinder the algorithm. Matlab software is used to develop the FKM algorithm, so it is necessary to import the database to create sub-matrices containing the power data classified from Mondays to Sundays for the three types of consumers; thus, three load curves are obtained for each customer, one representing the work week (Monday to Friday), another for Saturdays and a last one for Sundays. With the curves for each consumption group, the respective analysis will be carried out and the consumer's behavior throughout the day and week will be explained. Finally, with the curves already obtained, the Mean Index Adequacy (MIA) and Cluster Dispersion Index (CDI) are calculated, which allow us to know the state of the curves by measuring the performance of the algorithm and how efficient it is in grouping the data and finding a center among all these curves.