Partager

Estimation de consommation d’électricité et analyse sensible des variables d’un système de pompe à chaleur hydrothermique dans un immeuble de bureaux.

Electricity consumption prediction and variables sensitivity analysis of water source heat pump system in office building.

Numéro : pap. L14

Auteurs : ZHANG Y., ZHOU X., CHEN Z.

Résumé

Heating, ventilation, and air conditioning(HVAC) system is one of the major energy-consuming devices in buildings,and the energy consumption prediction of HVAC system has great significance for building energy conservation. This paper proposed a data-based method to predict the electricity consumption of water source heat pump(WSHP) system in an office building. The prediction models are established by support vector regression (SVR) algorithm, and trained by the local meteorological data. Furthermore, the root mean square error (RMSE) and the relative square error (RSE) are employed to evaluate the prediction performance, and the relative importance of each input variable is obtained for sensitivity analysis. Results show that only three variables,illumination intensity, time type and day type,have positive impacts on the performance of the prediction model, but the rest five variables,outdoor air temperature, outdoor air relative humidity, wind speed, wind direction and indoor air temperature,have negative impacts on it. When the negative variables are removed, the RMSE fall sto 4.6044 from 7.8227and the RSE fall sto 0.1494 from 0.4313during testing course.

Documents disponibles

Format PDF

Pages : 10

Disponible

  • Prix public

    20 €

  • Prix membre*

    15 €

* meilleur tarif applicable selon le type d'adhésion (voir le détail des avantages des adhésions individuelles et collectives)

Détails

  • Titre original : Electricity consumption prediction and variables sensitivity analysis of water source heat pump system in office building.
  • Identifiant de la fiche : 30022042
  • Langues : Anglais
  • Source : 8th international conference on compressors and refrigeration, 2017.
  • Date d'édition : 20/07/2017

Liens


Voir d'autres communications du même compte rendu (97)
Voir le compte rendu de la conférence

Indexation