Eintrag weiter verarbeiten
Heat load estimation using Artificial Neural Network
Gespeichert in:
| Veröffentlicht in: | Energy reports 6(2020), 2 vom: Feb., Seite 742-747 |
|---|---|
| Personen und Körperschaften: | , , |
| Titel: | Heat load estimation using Artificial Neural Network/ Apisit Panyafong, Nattawut Neamsorn, Chatchawan Chaichana |
| Format: | E-Book-Kapitel |
| Sprache: | Englisch |
| veröffentlicht: |
2020
|
| Gesamtaufnahme: |
: Energy reports, 6(2020), 2 vom: Feb., Seite 742-747
, volume:6 |
| Schlagwörter: | |
| Quelle: | Verbunddaten SWB Lizenzfreie Online-Ressourcen |
| Zusammenfassung: | Electricity demand from air conditioners are increasing every year due to increasing ambient air temperature. Several air-conditioned rooms in Thailand currently use low-efficient fixed-speed air conditioners. Prediction of heat load entering air-conditioned rooms may lead to better control of these air conditioners. In this paper, the Artificial Neural Network (ANN) model was developed, trained, and verified. Data from experiments were collected and used in the model. Sensitivity analysis showed that Tamb, Troom, and Gsolar are the major factors affecting heat load. The 3 factors were identified as the main input to the ANN model. There are 17 different model settings used in the study. The settings covered different model configurations such as the transfer function of the output layer, the transfer function of the hidden layer, the number of hidden layers and the number of neurons. It was found that an increasing number of neurons can increase the performance of the ANN model. Nonetheless, it is not conclusive that the increasing number of hidden layers can increase the performance of the model. Finally, it was found that choosing "tansig" as the transfer function of the output layer and "tansig" and "logsig" as transfer function of hidden layers offers the best model performance. |
|---|---|
| ISSN: |
2352-4847
|
| DOI: | 10.1016/j.egyr.2019.11.149 |
| Zugang: | Open Access |