Prediction of process and product parameters in an orange juice spray dryer using artificial neural networks

Abstract

In this study, the effects of feed flow rate, inlet-air temperature, and atomizer speed, in an orange juice semi-industrial spray dryer, were studied on seven performance indices, namely: residual moisture content of orange juice powder, particles size, bulk density, average time of wet ability, insoluble solids, outlet air temperature and dryer yield. A supervised artificial neural network (ANN) trained by back propagation algorithms was developed to predict seven performance indices based on the three input variables. The numbers of patterns used in this study were 80, used for training, verification, and testing the neural networks. After evaluating a large number of trials with various ANN architectures, the optimal model was a four-layered back-propagation ANN, with 14 and 10 neurons in the first and the second hidden layers, respectively. The ANN technology had been shown to be a useful tool to investigate, approximate and predict the physical properties of orange juice powder as well as process parameters of spray dryers. The final selected ANN model was able to predict the seven output parameters with RMSE lower than 0.042, R 2 higher than 0.93, and T value higher than 0.97. The results confirmed that the properly trained ANN model was able to produce simultaneously seven outputs, unlike traditional models where one mathematical model was required for each output. Radial Basis Function neural networks were not able well to learn the relationship between the input and output parameters. ANN parameters had a significant effect on learning ability of the ANN models.