An application of ANN Modeling on the Biosorption of Arsenic

Abstract

To solve the water pollution problem by arsenic contamination resulting from human technological activities has for long presented a challenge. Biosorbents shows better efficiency and multiple reuses to increase their economic attractiveness. The Artificial Neural Networks (ANNs) technology is a robust artificial intelligence technology that can handle the complex and dynamic nature of adsorption processes. A single layer ANN model was developed to predict the sorption efficiency of arsenic species from water bodies using Leucaena leucocephala seed powder. Batch experiments resulted into standardization of optimum conditions: biosorbent dosage (4.0 g), arsenic concentration (25 mg/L), volume (200 mL) at pH 7.5 for As (III) and 2.5 for As (V). Back-Propagation and Lenvenberg-Marquardt techniques are used to train various neural network architectures and the accuracy of the obtained models have been examined by using test data set. The optimal neural network architectures of this process contain 14 neurons with minimum mean squared error. The Levenberg–Marquardt algorithm was found best of BP algorithms with a minimum mean squared error for training and cross validation as 1.63694 × 10−5 and 0.000157767 respectively.