Control of a PCM ventilated facade using reinforcement learning techniques

Thumbnail
View/Open
Issue date
2015
Author
Gracia Cuesta, Alvaro de
Fernàndez Camon, César
Castell, Albert
Mateu Piñol, Carles
Cabeza, Luisa F.
Suggested citation
Gracia Cuesta, Alvaro de; Fernàndez Camon, César; Castell, Albert; Mateu Piñol, Carles; Cabeza, Luisa F.; . (2015) . Control of a PCM ventilated facade using reinforcement learning techniques. Energy and Buildings, 2015, vol. 106, p. 234-242. https://doi.org/10.1016/j.enbuild.2015.06.045.
Impact

Google Scholar logo  Google Scholar
Share
Export to Mendeley
Metadata
Show full item record
Abstract
Artificial intelligence techniques have been successfully applied to control dynamic systems looking for an optimal control. Among those techniques, reinforcement learning has been shown as particularly effective at reducing the dimensionality of some real problems and solving control problems by learning
from experience. The use of thermal energy storage active systems in the building sector is identified as suitable option to reduce their energy demand for heating and cooling. However, these systems might be expensive and require appropriate control strategies in order to improve the performance of the building. In this paper a ventilated facade with PCM is controlled using a reinforcement learning algorithm. The ventilated facade uses mechanical ventilation during nighttime to solidify the PCM and releases this cold stored to the inner environment during the peak demand period. It is crucial to decide correctly the schedule of charge and discharge process of the PCM according to the weather and indoor conditions. An experimentally validated numerical model is used to test the performance of the control algorithm under different weather conditions. Important improvements on the energy savings due to the use of control strategies were found and supported by the data under the different tested climatic conditions.
URI
http://hdl.handle.net/10459.1/58435
DOI
https://doi.org/10.1016/j.enbuild.2015.06.045
Is part of
Energy and Buildings, 2015, vol. 106, p. 234-242
Collections
cc-by-nc-nd, (c) Elsevier B.V., 2015
Except where otherwise noted, this item's license is described as cc-by-nc-nd, (c) Elsevier B.V., 2015