Artificial Neural Network Approach for Modeling Cu(II) and Cr(VI) Ion Biosorption from Synthetic Textile Effluent Using Three Low-Cost Biosorbents

Abstract

In wastewater treatment, scientific and practical models utilizing numerical computational techniques such as artificial neural network (ANN) can significantly help to improve the process as a whole through biosorption system. The present work develops an ANN model to forecast the sorption efficiency of Cu(II) and Cr(VI) heavy metal ions by extracting experimental data from biosorption study. The biosorbents used in the present study are Sorghum bicolor roots (SR), Tamarindus indica bark (TB), and Aloe vera pulp (AVP). Six models were simulated: Cu(II) and Cr(VI) by SR; Cu(II) and Cr(VI) by TB; Cu(II) and Cr(VI) by AVP. The input parameters of the models were shaking speed, initial concentration, absorbent dose, temperature, time, and pH. In this study, we apply the most ideal ANN model, the efficiency of which is determined by mean square error (MSE) and coefficient of determination (R2). These models worked with two transfer functions: tangent sigmoid (input layer to hidden layer) and linear transfer function (hidden layer to output layer). These models were trained, using 75% of the dataset, till the minimum root mean square error (RMSE) was observed and then tested using the remaining 25%. Each model has a different number of neurons in hidden layers that were registered on the basis of minimum RMSE. Thus, a simple BP algorithm has proved as a significant mathematical and computing model for the various biosorption processes.