Partager

Régulation intégrée du système de refroidissement et des ouvertures de surface utilisant les réseaux neuronaux artificiels.

Integrated control of the cooling system and surface openings using the artificial neural networks.

Auteurs : MOON J. W.

Type d'article : Article

Résumé

This study aimed at suggesting an indoor temperature control method that can provide a comfortable thermal environment through the integrated control of the cooling system and the surface openings. Four control logic were developed, employing different application levels of rules and artificial neural network models. Rule-based control methods represented the conventional approach while ANN-based methods were applied for the predictive and adaptive controls. Comparative performance tests for the conventional- and ANN-based methods were numerically conducted for the double-skin-facade building, using the MATLAB (Matrix Laboratory) and TRNSYS (Transient Systems Simulation) software, after proving the validity by comparing the simulation and field measurement results.
Analysis revealed that the ANN-based controls of the cooling system and surface openings improved the indoor temperature conditions with increased comfortable temperature periods and decreased standard deviation of the indoor temperature from the center of the comfortable range. In addition, the proposed ANN-based logic effectively reduced the number of operating condition changes of the cooling system and surface openings, which can prevent system failure. The ANN-based logic, however, did not show superiority in energy efficiency over the conventional logic. Instead, they have increased the amount of heat removal by the cooling system. From the analysis, it can be concluded that the ANN-based temperature control logic was able to keep the indoor temperature more comfortably and stably within the comfortable range due to its predictive and adaptive features.

Détails

  • Titre original : Integrated control of the cooling system and surface openings using the artificial neural networks.
  • Identifiant de la fiche : 30014862
  • Langues : Anglais
  • Source : Applied Thermal Engineering - vol. 78
  • Date d'édition : 03/2015
  • DOI : http://dx.doi.org/10.1016/j.applthermaleng.2014.12.058

Liens


Voir d'autres articles du même numéro (19)
Voir la source

Indexation