Prediction Air Temperature in Geothermal Heat Exchangers Using Pseudorandom Numbers: The New DARL Model
By: C. Ramírez-Dolores , J. C. Zamora-Luria , J. A. Altamirano-Acosta and more
The use of Earth-Air-Water Heat Exchangers (EAWHE) for sustainable air conditioning has not been widely studied. Due to their experimental nature, methods of characterizing internal thermal air distribution impose high dependence on instrumentation by sensors and entail data acquisition and computational costs. This document presents an alternative method that estimates air temperature distribution while minimizing the need for a dense network of sensors in the experimental system. The proposed model, DARL (Data of Air and Random Length), can predict the temperature of air circulating inside EAWHEs. DARL is a significant methodological advance that integrates experimental data from boundary conditions with simulations based on pseudo-random numbers (PRNs). These PRNs are generated using Fermat's prime numbers as seeds to initialize the generator. Ordinary linear regressions and robust statistical validations, including the Shapiro-Wilk test and root mean square error, have demonstrated that the model can estimate the thermal distribution of air at different lengths with a relative error of less than 6.2%. These results demonstrate the model's efficiency, predictive capacity, and potential to reduce dependence on sensors.
Similar Papers
Time-series Forecast for Indoor Zone Air Temperature with Long Horizons: A Case Study with Sensor-based Data from a Smart Building
Machine Learning (CS)
Predicts building temperature for better heating and cooling.
Deep Learning Atmospheric Models Reliably Simulate Out-of-Sample Land Heat and Cold Wave Frequencies
Atmospheric and Oceanic Physics
AI weather models predict extreme heat and cold.
Improving performance estimation of a PCM-integrated solar chimney through reduced-order based data assimilation
Numerical Analysis
Improves solar chimney airflow prediction using smart temperature guessing.