Customer Load Profile Clustering Using K-means Algorithm: A Case Study in an Electric Distribution Company in the Philippines Amidst the COVID-19 Pandemic

ABSTRACT

Most of the electric distribution companies in the Philippines are interested in analyzing customer load profile, they are concerned in classifying their customer's profile into different categories based on the energy consumption, Also the user's profile will help to understand how the consumption of energy may affect the electric distribution grid. In the current condition right now, facing the COVID-19 pandemic, most Filipinos are inclined to work at home, thus the consumption of energy increased. In this paper, residential data were collected in one of the electric distribution companies in the Philippines amidst the COVID-19 pandemic conditions. The data consist of 1,048,575 customer profiles from the year 2021. This study aims to use clustering methods such as the K-means algorithm in grouping customers' profiles and validate the suitable amount of clusters using the proposed method, such as the multi-criteria model and elbow method. Results show that 2 and 7 clusters, respectively, were fitted in the data. © 2022 IEEE.