The prediction of seedy grape drying rate using a neural network method

This paper presents an application which uses Feedforward Neural Networks (FNNs) to model the nonlinear behaviour of the drying of seedy grapes. First, a novel type of dryer for experimentally and mathematically evaluating the thin-layer drying kinetics of seedy grapes is developed. In the developed drying system, an expanded-surface solar air collector, a solar air collector with Phase-Change Material (PCM) and drying room with swirl element have been particularly included. Secondly, the drying rate is estimated as an exponential-type equation using non-linear regression analysis. Thirdly, the drying rate of seedy grapes is estimated using an FNN. Finally, the performance of the FNN model is compared with those of nonlinear and linear regression models by means of the root mean square errors, the mean absolute errors, and the correlation coefficient statistics. The results indicate that the FNN is more accurate and performed more consistently than alternative approaches employed in estimating drying rate.