Obtención de Curvas de Carga para los diferentes tipos de usuarios Residencial, Comercial e Industrial a través del algoritmo de agrupamiento k-means: utilización de la técnica de agrupamiento k-means para la obtención de curvas de carga representativas de consumidores tipo: residencial, comercial e industrial.

Abstract

The present work on curricular integration will focus on obtaining representative load curves through the use of the k-means clustering algorithm. This algorithm will be applied to the measurements taken from the different types of consumers, in order to define the daily load curves, as well as the calculation of the CDI (Clustering Dispersion Indicator) and MIA (Mean Index Adequacy) indicators. The data are obtained from the measurements made by the quality equipment installed at the heads of the feeders corresponding to the users connected in medium voltage and from the distribution transformers. These data are provided by the company CNEL-EP-Unidad de Negocios Azogues. Representative load curves will run through two main stages. The first involves processing the information to segment it by types of customers and in turn in a subclassification by days and time between taking measurements. The second stage is the development of data mining specifically of the k-means grouping method, this method makes use of the cluster technique in order to obtain representative load curves for the different types of residential, commercial and industrial users. This is done through the k-means clustering algorithm implemented in Matlab software, which will process the segmented information and result in representative load curves. Additionally, the CDI and MIA indicators are obtained in order to verify the veracity of the use of this method to obtain representative load curves.