Page 101 - C:\Users\am_se\OneDrive - Higher Education Commission\Desktop\FlipBook\

International Conference on

Recent Trends in Environmental Sustainability

ESCON22/ETERM/40
Modeling of lead phytoextraction from the soil using artificial neural networks (ANN)
and genetic algorithm (GA)

Usman Rauf Kamboh1, Maria Manzoor2,3* Ubaid Ullah1, Iram Gul4, Muhammad Arshad2*
1Department of Computational Sciences, The University of Faisalabad, Pakistan
2Institute of Environmental Sciences and Engineering, School of Civil and Environmental
Engineering, National University of Sciences and Technology, Islamabad, 44000, Pakistan
3Department of Environmental Sciences, University of Okara, Okara, 56300, Pakistan
4Department of Earth and Environmental Sciences, Hazara University, Mansehra, Pakistan

Correspondence: marea.manzoor@gmail.com; marshad@iese.nust.edu.pk
Abstract

Lead (Pb) is the well-known for containment of soil surfaces. In the last few decades,
phytoremediation is the most ideal technology to extract Pb from the soil, involving numerous
chemical reactions and cost analysis. In this study, Pb extraction from the soil by Pelargonium
hortorum has been shown and optimized by applying “Genetic Algorithm” (GA) technique for
“Response Surface Methodology” (RSM) and “Artificial Neural Network” (ANN). In the
modern era of modelling the extraction of metals pollutant, ANN is best fit because of its
efficiency and affordance. To determine the significance of the proposed solution, pot culture
experiments were done for optimizing Pb extraction competency from the Pb spiked (0 mg kg-
1, 500 mg kg-1, 1000 mg kg-1 and 1500 mg kg-1) soil by P. hortorum, applied with citric acid
(5 and 10 kg) and M. paraoxydance (1 and 1.5 OD). Plants were harvested at 30, 60 and 90
day’s intervals. Plant dry biomass and Pb uptake were determined from harvested plants. The
maximum Pb extraction efficiency of 86.0% was achieved with 500 mg/Kg soil Pb mm for 60
days which then reached to 81.49% with 400 mm OD solution. Furthermore, RSM based on
Box–Behnken design (BBD) and ANN-based Levenberg-Marquardt Algorithm (LMA) were
applied to model Pb extraction from soil. The predicted values from RSM and LMA were close
to 36.0% and 86.05%, respectively. The comprehensive evaluation of findings encouraged the
accuracy, reliability and efficiency of ANN for the optimization process. Therefore,
experimental results showed that ANN is an accurate technique to find the optimal chemical
soil washing parameters to remediate heavy metal polluted soil using environmental
ethanolamine. Furthermore, the proposed method is environment friendly and potentially cost-
effective.

Keywords: Genetic Algorithms, Pb-contaminated, Artificial Neural Network, Box–Behnken
design, Levenberg–Marquardt (LM)

Department of Environmental Sciences, COMSATS University Islamabad, Vehari Campus

68