<![CDATA[PLN UP3 Binjai faces challenges in understanding customer characteristics in each rayon so the service strategy is not fully effective. This study aims to group electricity user areas based on customer type and installed power so that PLN can develop a more targeted and efficient service strategy. The method used in this study is clustering with the K-Means algorithm, with customer data for 2024 from 9 rayons analyzed using the MATLAB R2014b application. Grouping is done with 3, 4 and 5 clusters. The results of grouping 3 clusters produce a variance value of 1.0890, in grouping 4 clusters it produces a variance value of 0.8047 and in grouping 5 clusters it produces a variance value of 0.6093. The results of the Grouping that has been carried out show that 5 clusters provide the most optimal results with the smallest variance value of 0.6093 which shows that the distribution of data between clusters is more homogeneous and representative. Each cluster shows a different combination pattern of rayon, installed power and customer type, which can be used as a basis for developing a more effective service strategy. Keywords: Electricity Customers, Electricity Power, K-Means Clustering.]]>
Copyrights © 2025
