Premis UPC de Sostenibilitat i CampusLab 2024

#PremiSostUPC2024 Edició 2024

Fase 1 de 1

Fase única 19/04/2024 - 21/06/2024

Fases del procés

Canvis a "Development of Adaptive Consumption Models for HVAC Systems using Neural Networks"

Mode de vista de comparació:

Vista HTML:

Títol

+{"ca"=>"Development of Adaptive Consumption Models for HVAC Systems using Neural Networks"}

Cos

+["
The escalating demand for cooling in buildings due to climate change and the imperative of thermal comfort necessitates the adoption of energy conservation measures. However, there is a research gap regarding comfort-based approaches for predicting energy consumption specifically in nursing homes considering real data, catering to vulnerable populations.

Enllaç a UPCommons

This master thesis addresses this void by developing neural network-based adaptive consumption models that assess the energy implications of HVAC systems during the cooling season, utilizing real-world data on energy consumption and environmental conditions. The models are elaborated and validated using a one-year data set comprising eight nursing homes situated in Mediterranean and Continental-Mediterranean climates. Through an iterative process, the neural network is trained, tested, and validated under different factors and variables using IBM SPSS Statistics and Matlab, finally incorporating cooling area, construction age, outdoor and indoor temperatures, and outdoor relative humidity as the model inputs, and cooling consumption as the output. The findings demonstrate that the use of a neural network-based consumption model results in excellent predictive capabilities (R2 = 0.95), significantly outperforming the previously elaborated linear modeling techniques under the same data set (R2 ≈ 0.65). Implementing adaptive thermal comfort control methods and using the neural network to evaluate cooling consumption, results show substantial energy savings achieved, outrunning the fixed set point temperature current strategy to acclimatize nursing homes. Specifically, the study reveals potential average energy savings of up to 23.4% (21.9% in the Mediterranean climate and 24.9% in the Continental-Mediterranean climate) for the analyzed facilities. Notably, construction age plays a decisive role in cooling consumption, with inefficiencies observed in older buildings. Predicting energy consumption using an adaptive comfort-based approach enhances energy efficiency in nursing homes, ensuring the well-being of vulnerable residents through optimal thermal comfort. These findings hold significant value for effective energy management in buildings under future climate change scenarios, warranting careful consideration by nursing home facility managers.
"]

+["The escalating demand for cooling in buildings due to climate change and the imperative of thermal comfort necessitates the adoption of energy conservation measures. However, there is a research gap regarding comfort-based approaches for predicting energy consumption specifically in nursing homes considering real data, catering to vulnerable populations.<blockquote><a href=\"https://upcommons.upc.edu/handle/2117/397562\" rel=\"noopener noreferrer\" target=\"_blank\">Enllaç a UPCommons</a></blockquote>This master thesis addresses this void by developing neural network-based adaptive consumption models that assess the energy implications of HVAC systems during the cooling season, utilizing real-world data on energy consumption and environmental conditions. The models are elaborated and validated using a one-year data set comprising eight nursing homes situated in Mediterranean and Continental-Mediterranean climates. Through an iterative process, the neural network is trained, tested, and validated under different factors and variables using IBM SPSS Statistics and Matlab, finally incorporating cooling area, construction age, outdoor and indoor temperatures, and outdoor relative humidity as the model inputs, and cooling consumption as the output. The findings demonstrate that the use of a neural network-based consumption model results in excellent predictive capabilities (R2 = 0.95), significantly outperforming the previously elaborated linear modeling techniques under the same data set (R2 ≈ 0.65). Implementing adaptive thermal comfort control methods and using the neural network to evaluate cooling consumption, results show substantial energy savings achieved, outrunning the fixed set point temperature current strategy to acclimatize nursing homes. Specifically, the study reveals potential average energy savings of up to 23.4% (21.9% in the Mediterranean climate and 24.9% in the Continental-Mediterranean climate) for the analyzed facilities. Notably, construction age plays a decisive role in cooling consumption, with inefficiencies observed in older buildings. Predicting energy consumption using an adaptive comfort-based approach enhances energy efficiency in nursing homes, ensuring the well-being of vulnerable residents through optimal thermal comfort. These findings hold significant value for effective energy management in buildings under future climate change scenarios, warranting careful consideration by nursing home facility managers."]

Esborrats

Addicions

+["
The escalating demand for cooling in buildings due to climate change and the imperative of thermal comfort necessitates the adoption of energy conservation measures. However, there is a research gap regarding comfort-based approaches for predicting energy consumption specifically in nursing homes considering real data, catering to vulnerable populations.

Enllaç a UPCommons

This master thesis addresses this void by developing neural network-based adaptive consumption models that assess the energy implications of HVAC systems during the cooling season, utilizing real-world data on energy consumption and environmental conditions. The models are elaborated and validated using a one-year data set comprising eight nursing homes situated in Mediterranean and Continental-Mediterranean climates. Through an iterative process, the neural network is trained, tested, and validated under different factors and variables using IBM SPSS Statistics and Matlab, finally incorporating cooling area, construction age, outdoor and indoor temperatures, and outdoor relative humidity as the model inputs, and cooling consumption as the output. The findings demonstrate that the use of a neural network-based consumption model results in excellent predictive capabilities (R2 = 0.95), significantly outperforming the previously elaborated linear modeling techniques under the same data set (R2 ≈ 0.65). Implementing adaptive thermal comfort control methods and using the neural network to evaluate cooling consumption, results show substantial energy savings achieved, outrunning the fixed set point temperature current strategy to acclimatize nursing homes. Specifically, the study reveals potential average energy savings of up to 23.4% (21.9% in the Mediterranean climate and 24.9% in the Continental-Mediterranean climate) for the analyzed facilities. Notably, construction age plays a decisive role in cooling consumption, with inefficiencies observed in older buildings. Predicting energy consumption using an adaptive comfort-based approach enhances energy efficiency in nursing homes, ensuring the well-being of vulnerable residents through optimal thermal comfort. These findings hold significant value for effective energy management in buildings under future climate change scenarios, warranting careful consideration by nursing home facility managers.
"]

Esborrats

Addicions

+["The escalating demand for cooling in buildings due to climate change and the imperative of thermal comfort necessitates the adoption of energy conservation measures. However, there is a research gap regarding comfort-based approaches for predicting energy consumption specifically in nursing homes considering real data, catering to vulnerable populations.<blockquote><a href=\"https://upcommons.upc.edu/handle/2117/397562\" rel=\"noopener noreferrer\" target=\"_blank\">Enllaç a UPCommons</a></blockquote>This master thesis addresses this void by developing neural network-based adaptive consumption models that assess the energy implications of HVAC systems during the cooling season, utilizing real-world data on energy consumption and environmental conditions. The models are elaborated and validated using a one-year data set comprising eight nursing homes situated in Mediterranean and Continental-Mediterranean climates. Through an iterative process, the neural network is trained, tested, and validated under different factors and variables using IBM SPSS Statistics and Matlab, finally incorporating cooling area, construction age, outdoor and indoor temperatures, and outdoor relative humidity as the model inputs, and cooling consumption as the output. The findings demonstrate that the use of a neural network-based consumption model results in excellent predictive capabilities (R2 = 0.95), significantly outperforming the previously elaborated linear modeling techniques under the same data set (R2 ≈ 0.65). Implementing adaptive thermal comfort control methods and using the neural network to evaluate cooling consumption, results show substantial energy savings achieved, outrunning the fixed set point temperature current strategy to acclimatize nursing homes. Specifically, the study reveals potential average energy savings of up to 23.4% (21.9% in the Mediterranean climate and 24.9% in the Continental-Mediterranean climate) for the analyzed facilities. Notably, construction age plays a decisive role in cooling consumption, with inefficiencies observed in older buildings. Predicting energy consumption using an adaptive comfort-based approach enhances energy efficiency in nursing homes, ensuring the well-being of vulnerable residents through optimal thermal comfort. These findings hold significant value for effective energy management in buildings under future climate change scenarios, warranting careful consideration by nursing home facility managers."]

Número de versió 1 de 1 Mostra totes les versions Torna a la proposta

Autoria de la versió

Nexus24 UPC

Versió creada el 13/06/2024 13:21