An investigation of indoor thermal environments and thermal comfort in naturally ventilated educational buildings

Indoor thermal conditions are essential in educational buildings. The health and well-being of students can be affected by a poor thermal environment, which also has a clear impact on building energy consumption. In this context, the indoor thermal environment in naturally ventilated university clas...

Descripción completa

Detalles Bibliográficos
Autores: Hoz Torres, María Luisa de la, Aguilar Aguilera, Antonio Jesús, Ruiz Padillo, Diego Pablo, Martínez Aires, María Dolores
Tipo de recurso: artículo
Fecha de publicación:2024
País:España
Institución:Consejo General de la Arquitectura Técnica de España (CGATE)
Repositorio:RIARTE
OAI Identifier:oai:www.riarte.es:20.500.12251/3729
Acceso en línea:http://hdl.handle.net/20.500.12251/3729
https://doi.org/10.1016/j.jobe.2024.108677
Access Level:acceso abierto
Palabra clave:Centros educativos
Universidad
Cuestionario
Confort térmico
Sensorización
Ventilación natural
Calidad percibida
Algoritmos
Redes neuronales
3305.26 Edificios Públicos
1203.06 Sistemas Automatizados de Control d
3308.04 Ingeniería de la Contaminación
Descripción
Sumario:Indoor thermal conditions are essential in educational buildings. The health and well-being of students can be affected by a poor thermal environment, which also has a clear impact on building energy consumption. In this context, the indoor thermal environment in naturally ventilated university classrooms is explored in this study during a complete academic year. A monitoring campaign and a questionnaire survey were conducted simultaneously in higher education buildings in Spain. A total of 2115 sets of data were collected. Thermal sensation prediction indices (predicted mean vote, extended predicted mean vote and adaptive predicted mean vote) were applied to evaluate student’s thermal perception and their prediction accuracy was assessed. Additionally, two machine-learning models, based on Artificial neural network (ANN) and random forest (RF) algorithms, were formulated to predict occupants’ thermal sensation. The obtained results evidenced that the proposed ANN and RF models outperform traditional indices. Finally, it is also proposed an adaptive thermal comfort model. The results obtained suggest that students have a greater adaptive capacity to changes in environmental conditions than suggested by the ASHRAE-55 adaptive model and that they preferred an environment with lower temperatures than those suggested by the EN-16798 adaptive model.