Optimasi Rute Penggantian Kwhmeter Pascabayar dengan K-Means dan Algoritma Genetika

Abstract

This research is driven by the need of PLN Distribution Main Unit to mitigate the extreme risk of high account balances on the 21st of each month, which can disrupt cash flow and financial performance. One mitigation strategy is replacing conventional kWh meters with Advanced Metering Infrastructure (AMI), or two-way meters, which allow remote disconnection. The purpose of this study is to prioritize customers for meter replacement based on limited material availability and to optimize replacement routes under restricted team resources. The method used involves the K-means algorithm for customer segmentation, validated by silhouette score, which resulted in six optimal clusters. Genetic algorithm was then applied to determine the most efficient routes based on distance and time. The results showed that the optimal route had a minimum total travel distance of 1,434.29 km and a total time of 18,868.58 minutes, using a population size of 50 chromosomes and 500 generations. This study is expected to offer an effective solution for improving operational efficiency and risk mitigation at PLN.