Cluster-based Energy Load Profiling on Residential Smart Buildings

摘要

Percentage of population living in cities is expected to reach 60% by 2030, accounting for 60% - 80% of world annual energy needs and making the impact of energy efficient solutions in cities quite significant for environmental protection and fighting climate change. The building sector uses about 40% of European energy and emits approximately 1/3 of greenhouse gas emissions. Black-box measurement based modeling methods allow the estimation of consumption in buildings relying on smart metering devices installed. Vast amount of data generated poses new challenges with reference to their handling and timely processing. The paper presents an approach related to building energy load profiling utilising profile compression and clustering. It discusses the application of different clustering algorithms through their experimental evaluation.