Use of artificial neural networks and genetic algorithms for prediction of sorption of an azo-metal complex dye onto lentil straw

Abstract

Artificial neural network (ANN), pseudo second-order kinetic, and gene expression programming (GEP) models were constructed to predict removal efficiency of Lanaset Red G (LR G) using lentil straw (LS) based on 1152 experimental sets. The sorption process was dependent on adsorbent particle size, pH, initial dye concentration, and contact time. These variables were used as input to construct a neural network for prediction of dye uptake as output. ANN was an excellent model because of the lowest error and the highest coefficient values. ANN indicated that initial dye concentration had the strongest effect on dye uptake, followed by pH. The GEP model successfully described the sorption kinetic process as function of adsorbent particle size, pH, initial dye concentration, and contact time in a single equation. Low cost adsorbent, LS, had a great potential to remove LR G as an eco-friendly process, which was well described by GEP and ANN.