K-Means Clustering of Electricity Consumption from IoT Data: A Case Study in Electrical Engineering Department Building, ITN Malang

The electricity consumption profile represents the occupant’s behavior in consuming electricity. It provides valuable information in the energy management system. This paper presents a work for clustering electricity consumption using the K-means clustering technique. Instead of a whole building, our approach deals with grouping the electricity consumption in each room, namely the Automation Laboratory and Department Office of the Electrical Engineering Department of ITN Malang. The electricity dataset is obtained from the IoT data of the Building Automation System (BAS) installed in the rooms, which records each room’s lamps and outlet consumption separately. The clustering results show that individual rooms’ load profile provides a more detailed profile than the aggregated ones. Further, the developed IoT-based BAS system offers a valuable approach to electricity consumption clustering.