Adaptive in-situ forecasting for demand-side management in low voltage power grids
Editors: Christian Backe, Miguel Bande Firvida
In Adaptive in-situ forecasting for demand-side management in low voltage power grids, (SEST-2021), 06.9.-08.9.2021, Vaasa, IEEE, pages 1-6, 2021. ÎEEE. ISBN: 978-1-7281-7660-4.
This paper describes a concept and results for adaptive in-situ forecasting for demand-side management (DSM) in low voltage power grids. It is motivated by shortcomings of the predominant development strategy for prediction models: When these are trained as static models, i.e. offline on previously acquired data, they cannot adapt quickly to new environments, and they require centralized data collection, creating privacy issues that may possibly decrease the willingness of consumers to make their demand flexibilities available. The forecasting models presented in this paper were trained and optimized on load balance data generated in a custom simulation environment for distributed DSM. The concept will be applied in the SENSE Smart Grid Laboratory of TU Berlin in the context of the project FUSE (FUture Smart Energy).
Machine Learning, Online Learning, Forecasting, Embedded Systems, Demand-Side Management