Household Energy Consumption Clustering Using ‘k-means' Algorithm: A Bangladesh Case Study

Previous research emphasizes the need for effective power management techniques to optimize energy utilization and GHG emissions, and control peak demand. In Bangladesh, the residential sector accounts for half of the total energy consumption. Thus, understanding the power consumption behavior of residential customers to design effective demand-side management techniques is crucial. This study investigates the application of the k-means algorithm to analyze real-time household electricity consumption data in Bangladesh. The k-means algorithm, known for its simplicity and efficiency, is applied to a dataset collected from smart meters, revealing four distinct clusters representing low, moderate, low to moderate, and high consumption households. This clustering provides valuable insights for designing targeted demand-side management schemes such as tariff-based strategies. However, the study acknowledges limitations, such as a small dataset of 14 households and a single region, emphasizing the necessity for future research with larger datasets and regional variability to enhance the algorithm's effectiveness in policy-related decision-making for sustainable energy management in Bangladesh.