Analysis of Electric Energy Consumption Profiles Using a Machine Learning Approach: A Paraguayan Case Study

• DOI: 10.3390/electronics11020267
• Publication Date: 2022-01-14T17:34:04Z
• Authors: Félix Morales; Miguel García-Torres; Gustavo Velázquez; Federico Daumas-L...; Pedro E. Gardel-S...; Francisco Gómez-Vela; Federico Divina; José Luis Vázquez...; Carlos Sauer Ayala; Diego P. Pinto-Roa; Julio César Mello...; David Becerra-Alonso
• Keywords: Hierarchical clustering; Feature (linguistics); Data set; Linkage (software)

Correctly defining and grouping electrical feeders is of great importance for system operators. In this paper, we compare two different clustering techniques, K-means hierarchical agglomerative clustering, applied to real data from the east region Paraguay. The raw were pre-processed, resulting in four sets, namely, (i) a weekly feeder demand, (ii) monthly (iii) statistical feature set extracted original (iv) seasonal daily consumption obtained considering characteristics Paraguayan load curve. Considering algorithms, distance metrics five linkage criteria total 36 models with Silhouette, Davies–Bouldin Calinski–Harabasz index scores was assessed. algorithms sets showed best performance validation configuration six clusters.