Partager

Étude d’un système à débit de frigorigène variable (VRF) en mode refroidissement au moyen d’essais sur le terrain et de modélisation basée sur l’apprentissage automatique.

Investigation of VRF System under Cooling Mode through Field Testing and Machine Learning-based Modeling. 

Numéro : No 196

Auteurs : WAN H., CAO T., HWANG Y., BAE H., OH S.

Résumé

Variable Refrigerant Flow (VRF) system is widely known for its flexibility and energy efficiency. Nevertheless, most of the research has focused on experiments in laboratory. Limited papers discussed the field test case. To enhance operational efficiency, we investigated details of cooling operational behavior through field testing and developed a machine learning-based model for a better control algorithm. We conducted the test in a university office building. The power consumption under cooling mode was investigated. In addition, a Support Vector Regression algorithm-based model has been built to predict the power consumption of the VRF system. The penalty parameters in the SVR model have been optimized to decrease the generalization error of the model. We found that the compressor frequency and condensing temperature are the two most important parameters for the accurate. The mean relative error of the model was 7% for outdoor unit power consumption when we compared experimental data and modeling results. The modeling results showed that if the compressor frequency, outdoor fan speed, and condensing temperature were the same, the power consumption should be the same.

Documents disponibles

Format PDF

Pages : 9 p.

Disponible

Gratuit

Télécharger

Détails

  • Titre original : Investigation of VRF System under Cooling Mode through Field Testing and Machine Learning-based Modeling. 
  • Identifiant de la fiche : 30030034
  • Langues : Anglais
  • Sujet : Technologie
  • Source : 13th IEA Heat Pump Conference 2021: Heat Pumps – Mission for the Green World. Conference proceedings [full papers]
  • Date d'édition : 31/08/2021

Liens


Voir d'autres communications du même compte rendu (198)
Voir le compte rendu de la conférence

Indexation