Feed-forward in temperature control of buildings
Artikel i vetenskaplig tidskrift, 2005

In most temperature control systems for buildings and rooms, the control strategy is based on feedback of the measured indoor temperature. In this paper, however, control strategies with both feedback and feed-forward are investigated and discussed. In particular, we discuss feed-forward from internal disturbances, such as lighting and electrical machines. By means of a number of computer simulations based on an identified state-space model of a small building, it is shown that such a control strategy reduce the energy use as well as improve the control performance (smaller temperature variations). Furthermore, it is shown that dynamic feed-forward gives better performance than static feed-forward and that static feed-forward gives better performance than no feed-forward at all. It is also shown how the improvement of using feed-forward controller strategies in the building studied depends on the frequencies of the disturbances. © 2004 Elsevier B.V. All rights reserved.

neural network

predictive climate control

intelligent building

Författare

Bertil Thomas

Chalmers, Signaler och system

Mohsen Soleimani-Mohseni

Chalmers, Energi och miljö, Installationsteknik

Per Fahlén

Chalmers, Energi och miljö, Installationsteknik

Energy and Buildings

0378-7788 (ISSN)

Vol. 37 755-761

Ämneskategorier

Samhällsbyggnadsteknik

DOI

10.1016/j.enbuild.2004.10.002