NSChE Journal Articles
*The article download page is restricted to registered members only.
ARTIFICIAL NEURAL NETWORKS MODELING APPROACH FOR PREDICTION OF ADSORPTION OF PAHS IN SLURRY REACTOR
C. N. Owabor and *B.V.Ayodele Department of Chemical Engineering, University of Benin, Benin City, Nigeria.
Article Abstract
Artificial Neural Network (ANN) model for the prediction of adsorption of Poly Aromatic Hydrocarbons (PAHs) in a slurry reactor was developed based on the multi-layer perceptron (MLP) feed forward network that was trained using the Levenberg-Marquadt algorithm. Experimental data were employed to design the feed forward neural networks modeling process and a total of 14 data points were used for training (8) and testing (6), respectively. The parameter estimation resulted in the following values for root mean square error (RMSE) - 1.274, 2.046 and 6.418, regression coefficient (r2) - 0.989, 0.912 and 0.999, and mean absolute deviation - 1.073, 1.677 and 0.325 for naphthalene, anthracene and pyrene respectively. Predicted values of the adsorption kinetic data were obtained from the ANNs and compared with the experimental values.
Click here to download full article *Access restricted to members only
<<<<<<<Back to journal listing
 
(c) 2010 Nigerian Society of Chemical Engineers ---------------- Contact Webmaster