Modeling and simulation of arsenic (V) extraction from water by emulsion liquid membrane process using artificial neural networks

This study surveys the modeling and simulation of arsenic (v) extration through emulsion liquid membrane process from an aqueous medium using artificial neural networks (ANNs) andregression models. Water in oil emulsions were prepared by means of ultrasound and prepared emulsions then were used in a emulsion liquid membrane process for extraction of arsenic (v) ions from an aquesus medium. The effects of different concentration of Span 80 and mixture of Span80 and Tween 20 emulsifiers on the extraction of arsenic were investigated. ANN modeling wasdone using the multi-layer perceptron (MLP) network with Levenberg-Marquardt training algorithm. Furthermore, 3D regression was also used to model the process. Concentration ofarsenic in feed was obtained as function of emulsifier concentration and time. Results show good agreement between the simulation results and experimental data. To evaluate models performance, some statistical parameters were computed. Results show significantly better prediction of MLP networks than regression models and MLP networks can apply as a powerful tool for reliableprediction of the emulsion liquid membrane process.

کلیدواژه ها:

Modeling ، emulsion liquid membrane ، arsenic ، extraction ، MLP network ، regression