Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
RIYANTO, YOGA AGUS
Asri, Yessy
Kuswardani, Dwina
Subject
Teknik Informatika
Datestamp
2022-09-26 07:55:43
Abstract :
Unreasonable use of electric power due to non-technical losses resulting in financial losses for the power company. AMR (Automatic Meter Reading) is used to read or retrieve electrical energy measurement data from each consumer locally and remotely. This study uses the K-means clustering algorithm method with data from the UCI Repository on energy consumption for 30 days with a total of 100 subscribers. To test the optimality of a set of clusters, the DBI (Davies-Bouldin index) is the minimum value. In the whole of cluster 2, it becomes the most optimal cluster with a DBI value = 0.1551 and the purity value reaches 0.95. So the K-Means algorithm produces two energy use models which are used as benchmarks to determine the fair use and the unnatural use of electrical energy. This control aims to obtain the results of the analysis to act in accordance with the SOP in order to be able to increase the TMP (Service Quality Level).