Neuro fuzzy and hybrid modeling of supercritical fluid extraction of Pimpinella Anisum L. seed

Abstract

In the current study, a Neuro-Fuzzy model has been developed to predict the mass of extract in the process of supercritical fluid extraction of Pimpinella anisum L. seed. The adaptive-network-based fuzzy inference system (ANFIS) technique was trained with the recorded data from kinetic experiments of the mentioned process at pressures of 8, 10, 14 and 18 MPa and constant temperature of 303.15 K which generated the membership function and rules that excellently expounded the input/output correlations in the process. Excellent prediction with Root Mean Square Error (RMSE) of 0.0235 was observed. In the next step of study, mass transfer coefficient in terms of Sherwood number was estimated by a neuro-fuzzy network. Then, the estimated mass transfer coefficient was introduced into the mathematical model. The proposed gray box (hybrid) model was validated with the experimental data. Results confirmed that equipping mathematical model with neuro-fuzzy network improved performance of the model significantly. Shokri et al. (2011) applied Artificial Neural Networks and mathematical modeling on this process, and reported the results of the proposed models. In the last part of this thesis, all four models (including two proposed models of this study) were compared. It was concluded that neuro-fuzzy and gray box models had the best performance.