Abdolreza Moghadassi, Sayed Mohsen Hosseini, Fahime Parvizian, Ibrahim Al-Hajri, and Mehdi Talebbeigi

pp. 531 - 538

Abstract

The extraction of essential oils using compressed carbon dioxide is a modern technique offering significant advantages
over more conventional methods, especially in particular applications. The prediction of extraction efficiency is a powerful tool
for designing and optimizing the process. The current work proposed a new method based on the artificial neural network
(ANN) for the estimation of the extraction efficiency of the essential oil oregano bract. In addition, the work used the backpropagation learning algorithm, incorporating different training methods. The required data were collected; pre-treating was
used for ANN training. The accuracy and trend stability of the trained networks were verified according to their ability to predict
unseen data. The Levenberg-Marquardt algorithm has been found to be the most suitable algorithm, with the appropriate
number of neurons (i.e., ten neurons) in the hidden layer and a minimum average absolute relative error (i.e., 0.019164). In
addition, some excellent predictions with maximum error of 0.039313 were observed. The results demonstrated the ANN’s
capability to predict the measured data. The ANN model performance was also compared to a suitable mathematical model,
thereby confirming the superiority of the ANN model.