mdh.sePublications
Change search
Link to record
Permanent link

Direct link
BETA
Sandberg, Alexander
Publications (1 of 1) Show all publications
Sandberg, A., Wallin, F., Li, H. & Maher, A. (2017). An analyze of long-term hourly district heat demand forecasting of a commercial building using neural networks. Energy Procedia, 3784-3790
Open this publication in new window or tab >>An analyze of long-term hourly district heat demand forecasting of a commercial building using neural networks
2017 (English)In: Energy Procedia, ISSN 1876-6102, E-ISSN 1876-6102, p. 3784-3790Article in journal (Refereed) Published
Abstract [en]

With the building sector standing for a major part of the world's energy usage it of utmost importance to develop new ways of reduce the consumption in the sector. This paper discusses the evolution of the regulations and policies of the Swedish electric and district heating metering markets followed by the development of a nonlinear autoregressive neural network with external input (NARX), with the purpose of performing heat demand forecasts for a commercial building in Sweden. The model contains 13 input parameters including; calendar, weather, energy and social behavior parameters. The result revealed that these input parameters can predict the building heat demand to 96% accuracy on an hourly basis for the period of a whole year. Further analysis of the result indicates that the current data resolution of the district heat measuring system limits the future possibilities for services compared to the electric metering system. This is something to consider when new regulation and policies is formulated in the future.

National Category
Energy Engineering
Identifiers
urn:nbn:se:mdh:diva-37548 (URN)10.1016/j.egypro.2017.03.884 (DOI)000404967903130 ()2-s2.0-85020704281 (Scopus ID)
Available from: 2017-12-22 Created: 2017-12-22 Last updated: 2018-07-25Bibliographically approved
Organisations

Search in DiVA

Show all publications