Artificial Neural Network Approach for Modeling of Cr (VI) Adsorption from Waste Water by Lewatit MP64 and Dowex 1x8

In this study, an artificial neural network model was developed to estimate the removal efficiency of Cr (VI) ion from waste water by Lewatit MP64 and Dowex 1x8 resins. For this purpose, 36 experimental data obtained in a laboratory batch study. In the developed model, contact time, adsorbent dosage, pH and concentration were used as the input parameters, and removal efficiency for Lewatit MP64 and Dowex 1x8 were also used as output parameters. The model performances were determined by the mean square error and the coefficient of determination. The model using the Levenberg-Marquardt backpropagation algorithm (TrainLM) was found the best prediction. This model also has a hidden layer and 15 neurons (4-15-1). The coefficient of determination between experimental and estimates was found to be 0.99 removal efficiency for Lewatit MP64 and 0.92 for Dowex 1x8. The results show that removal efficiency can be predicted successfully with artificial neural networks.