An Intelligent Regression-Based Approach for Predicting a Geothermal Heat Exchanger’s Behavior in a Bioclimatic House Context

Growing dependence on fossil fuels is one of the critical factors accelerating climate change, a global concern that can destabilize ecosystems and economies worldwide. In this context, renewable energy is emerging as a sustainable and environmentally responsible alternative. Among the options, geot...

Full description

Saved in:
Bibliographic Details
Published in:Energies (Basel) 2024-06, Vol.17 (11), p.2706
Main Authors: Díaz-Longueira, Antonio, Rubiños, Manuel, Arcano-Bea, Paula, Calvo-Rolle, Jose Luis, Quintián, Héctor, Zayas-Gato, Francisco
Format: Article
Language:English
Subjects:
Alternative energy sources
Artificial intelligence
Biomass
Efficiency
Electrical installations
Electricity
Electricity distribution
energy efficiency
Energy resources
Equipment and supplies
Fossil fuels
geothermal heat exchanger
Geothermal power
Heat exchangers
Heating
Machine learning
MLP
prediction
R&D
random forest
Renewable resources
Research & development
Sensors
Supplies
SVR
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Growing dependence on fossil fuels is one of the critical factors accelerating climate change, a global concern that can destabilize ecosystems and economies worldwide. In this context, renewable energy is emerging as a sustainable and environmentally responsible alternative. Among the options, geothermal energy stands out for its ability to provide heat and electricity consistently and efficiently, offering a feasible solution to reduce the carbon footprint and promote more sustainable development in a globalized economy. In this work, a machine learning approach is proposed to predict the behavior of a horizontal heat exchanger from a bioclimatic house. First, a correlation analysis was conducted for optimal feature selection. Then, several regression techniques were applied to predict the output temperature of the geothermal exchanger. Satisfactory prediction results were obtained in different scenarios over the whole dataset. Also, a significant correlation between several sensors was concluded.
ISSN:1996-1073
1996-1073
DOI:10.3390/en17112706