Simulation of Supercritical Fluid Extraction (SFE) of oil from pomegranate seeds using Supercritical Carbon Dioxide (SC-CO2) was investigated to study the influence of process parameters on the extraction rate and oil composition. Yield predictions of the pomegranate seed oil production outbreaks from input vectors of our experiments may help us to interpret well to find optimized operating condition to achieve high yields. Intelligent systems are perfect tools for prediction in such systems with numerous effective factors that each one can change the experimental output answers maybe against expect. Our effort in this paper is to find a route and effective way to evaluate and predict the yield of seed oil production at different operational conditions without paying extra costs and spending more times by using intelligent systems. Several approaches to predict the pomegranate seed oil extraction with fuzzy sets; neural and adaptive fuzzy neural systems are analyzed and tested. Prediction strategies tested in the paper include the fuzzy C-means (FCM) clustering, the common neural networks (NN) and application of fuzzy neural networks. The results indicate the superiority of the adaptive fuzzy neural networks method over common neural network and fuzzy clustering approaches. The experimental results demonstrate that the proposed fuzzy neural network algorithm is able to reveal a better performance than conventional back propagation NN and FCM algorithms.

Keywords