ARTIFICIAL NEURAL NETWORK FOR OPTIMIZING THE CONVECTIVE DRYING PROCESS OF PEAR SLICES
REŢEA NEURALĂ ARTIFICIALĂ PENTRU OPTIMIZAREA PROCESUL DE USCARE CONVECTIVĂ A FELIILOR DE PERE
DOI : https://doi.org/10.35633/inmateh-79-75
Authors
Abstract
This study develops an artificial neural network (ANN) model to optimize the convective drying process of pear slices. Experimental data from drying at 50°C, 60°C, and 70°C were used to train an ANN with three hidden layers. The model predicted drying rate with high accuracy (98.823% validation fidelity), capturing the nonlinear relationship between moisture content, time, and drying rate. Results demonstrated that higher temperatures accelerated drying but required careful control to maintain quality. The ANN effectively identified optimal drying parameters, balancing energy efficiency and product quality preservation, providing a valuable tool for industrial drying process optimization.
Abstract in Romanian



