Abstract:
This study explores the rehydration kinetics of dry dates using a solar water hydration system
under real climatic conditions in Ghardaïa. It focuses on evaluating the thermal response of
dates in three regions:El Oued, Guerrara, and Beriane, at hydration temperatures ranging
from 39°C to 55°C.
Mathematical modeling of the rehydration process using Peleg, Weibull, and exponential
models yielded high coefficients of determination (R2 > 0.95), confirming their suitability for
accurately predicting water absorption kinetics. The effective diffusion coefficients varied
considerably with temperature, reflecting the combined effects of water viscosity and
molecular mobility.
In addition, an artificial neural network (ANN) model was developed and trained using eight
days of meteorological data (ambient temperature, solar irradiation, wind speed, and relative
humidity) to predict water temperature inside the solar heater. The ANN model demonstrated
excellent predictive performance (R2 = 0.968; MAE = 3.140; RMSE = 4.988), with predicted
temperatures closely matching measured values. The high accuracy of the ANN model
underscores its potential for real-time temperature prediction, enabling better control and
management of the solar rehydration process.
Overall, this study emphasizes the critical role of precise temperature management in
optimizing the rehydration of dates and highlights the integration of ANN models as
powerful tools for enhancing process efficiency and sustainability in arid regions.
