Aggregated forecasting of the load control responses using a hybrid model that combines a physically based model with machine learning

Pekka Koponen, Tuukka Salmi, Corentin Evens, Suvi Takala, Antti Hyttinen, Christina Brester, Mikko Kolehmainen, Harri Niska

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

Abstract

The value of active demand in the electricity and ancillary service markets depends very much on the predictability of its aggregated control responses. In this work, we study electrically heated small houses that have electrical heating with heat storage tanks and remote control via a smart metering system. We integrate a simple physically based model to a machine learning forecasting method thus combining the strengths of the component methods. Now a Stacked Boosters Network, a new deep learning method, is applied and briefly compared with a support vector regression, an earlier machine learning model. The simple physically based model component models the thermal dynamics of the heat storage tank and the outdoor dependent heat demand of the house. Varying types of heuristic market based dynamic load control were applied during the
field trials that comprised an identification period (31 May 2012 – 31 May 2013) and a verification period (1 January 2015 – 31 December 2019). Each of the 727 houses was hourly metered and aggregated to two groups. We short–term forecast the power of these dynamically controlled groups. We summarise the results. The hybrid method outperformed its component methods.
Original languageEnglish
Title of host publicationProceedings of the CIRED 2020 Berlin Workshop
PublisherInternational Conference and Exhibition on Electricity Distribution CIRED
Number of pages4
Publication statusAccepted/In press - 2020
MoE publication typeA4 Article in a conference publication
EventCIRED 2020 Workshop Online, Berlin 22-23 September: How to Implement Flexibility in the Distribution System? - Online, Berlin, Germany
Duration: 22 Sep 202023 Sep 2020
https://www.cired2020berlin.org/

Publication series

SeriesCIRED: Conference Proceedings
ISSN2032-9644

Conference

ConferenceCIRED 2020 Workshop Online, Berlin 22-23 September
Abbreviated titleCIRED 2020
CountryGermany
CityBerlin
Period22/09/2023/09/20
Internet address

Fingerprint Dive into the research topics of 'Aggregated forecasting of the load control responses using a hybrid model that combines a physically based model with machine learning'. Together they form a unique fingerprint.

  • Cite this

    Koponen, P., Salmi, T., Evens, C., Takala, S., Hyttinen, A., Brester, C., Kolehmainen, M., & Niska, H. (Accepted/In press). Aggregated forecasting of the load control responses using a hybrid model that combines a physically based model with machine learning. In Proceedings of the CIRED 2020 Berlin Workshop [105] International Conference and Exhibition on Electricity Distribution CIRED. CIRED: Conference Proceedings