A framework for prediction and segmentation of daily energy load profiles of building clusters using machine learning

Research output: Other contributionNet publication - Internet publicationCommunication

27 Downloads (Pure)

Abstract

In the future smart grid and smart city context it is a necessary step to predict load profiles and categorize them for energy system management although the focus of existing studies on the energy profile classification and customer segmentation is still mainly on historical data analysis. In this paper, we propose a framework to identify and classify daily energy load profiles, and further categorize the forecasted future load profiles. The contributions of this study lie in: (i) A two-step clustering method for load profiling and categorization preserving both consumption magnitude and the shape of the load profile. (ii) A comprehensive classification procedure enabling the classification of forecasted daily heating profiles for the day ahead demand side management.
Original languageEnglish
Publication date2020
PublisherTechnical University of Denmark
Number of pages2
Publication statusPublished - 2020

Fingerprint Dive into the research topics of 'A framework for prediction and segmentation of daily energy load profiles of building clusters using machine learning'. Together they form a unique fingerprint.

Cite this