Two new models of direct luminous efficacy under clear sky conditions for daylighting in Burgos, Spain
The use of daylight in buildings contributes to energy savings while significantly improving visual comfort and well-being. It is therefore very important to be able to quantify illuminance to take advantage of daylight. Although several models have been proposed in recent years to determine global...
| Autores: | , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2024 |
| País: | España |
| Institución: | Universidad Pública de Navarra |
| Repositorio: | Academica-e. Repositorio Institucional de la Universidad Pública de Navarra |
| OAI Identifier: | oai:academica-e.unavarra.es:2454/52554 |
| Acceso en línea: | https://hdl.handle.net/2454/52554 |
| Access Level: | acceso abierto |
| Palabra clave: | Direct illuminance Luminous efficacy models Modeling |
| Sumario: | The use of daylight in buildings contributes to energy savings while significantly improving visual comfort and well-being. It is therefore very important to be able to quantify illuminance to take advantage of daylight. Although several models have been proposed in recent years to determine global and diffuse illuminance, the same may not be said of direct solar illuminance, which situates this study in an area of noteworthy scientific and technological interest. Two luminous efficacy models for clear sky conditions are proposed and the results of benchmarking with previous models from the literature are presented. Data collected in Burgos (Spain) were analyzed. Specifically, eight previous models for the prediction of direct illuminance were compared with our two new models. The two new models predicted illuminance more accurately than most of the classic models. Specifically, running the models on the training data yielded Root Mean Square Error (RMSE) values of 2.58 % and 2.76 % for the first and the second model, respectively. Likewise, the test data yielded RMSE values of 3.31 % and 3.49 %, and the Mean Bias Error values with the training data were 0.06 % and 0.11 %, respectively. The models achieved high accuracy levels with both the training and the test data sets. |
|---|