Anzeige
Erschienen in:
2020 | OriginalPaper | Buchkapitel
Fitting Analysis for Flow Coefficient Predictive Model of Residential Buildings with High Airtightness
verfasst von : Wenqian Zhou, Xiangli Li, Lin Duanmu
Erschienen in: Proceedings of the 11th International Symposium on Heating, Ventilation and Air Conditioning (ISHVAC 2019)
Verlag: Springer Singapore
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
The flow coefficient is an important parameter of air infiltration, affected by many factors, such as building parameters, wind speed and temperature difference between indoor and outdoor. And its modeling is used to adopt the method, multiple linear regression, ignoring the problem caused by characteristic of small sample size. Stepwise regression, least squares and partial least squares (PLSs) with corresponding independent variable screening methods are adopted to obtain the predictive models which are compared with F-test, goodness of fit and CV (RMSE). The result shows that the goodness of fit of apartment’s model and villa’s model received by PLS is increased by 25.9% and 2.2%, respectively, compared with least squares and the PLS models’ CV (RMSE) which are 9.4% and 18.3%. Considering the essential characteristics and fitting results of the three methods, it can be concluded that PLS which is suitable for small sample size is the preferred choice for the establishment of flow coefficient model.