Clusterization of customer energy usage to detect power shrinkage in an effort to increase the efficiency of electric energy consumption

Abstract

<span>Automatic meter reading (AMR) is a reading system result the measurement of electrical energy consumen, both locally and remotely. The problems faced is the high non-technical shrinkage of AMR customers due to installation, maintenance errors as well as dishonest actions some consumers, this has a major influence on electrical power losses. PT. PLN Disjaya currently faces difficulties having to choose which customers should be checked first, so the field can only find a little damage. The K-means method based on historical electric power usage and determine the most optimal number of groups the davies-bouldin index (DBI) method. Based on the results of testing with 2-6 sets of clusters, the cluster set results are the most optimal is set cluster 4 because it has the smallest DBI value 0.893. The set of 4 clusters has the best performance in data grouping of historical power usage of AMR customers the business class, each centroid of each cluster is used as an attribute and value of the AMR customer power usage business chart. The testing phase is customers who categorized as customers with un-normal usage electricity power. The test is, by determining the distance data testing each centroid in the cluster 4 set.</span>

Identifikasi Profil Konsumsi Enegri Listrik untuk Meningkatkan Pendapatan dengan Klustering

“…Data hasil uji berada pada kelompok cluster yang memiliki jarak terkecil terhadap titik centroid, sehingga membentuk cluster. Setiap cluster memiliki karakteristik yang berbedabeda [21]. Keempat variabel uji memberikan informasi karakteristik pemakaian energi listrik dan potensi pendapatan penjualan tenaga listrik.…”