Removal of chromium from industrial wastewater by magnetic flocculation treatment: Experimental studies and PSO-BP modelling

A simple and effective method for chromium removal in wastewater by using magnetic nanoparticles (MPs, Fe3O4) in the gradient magnetic field was studied in this paper. The optimum process conditions were determined by testing the pH, added MPs dosage, stirring time and standing time. The maximum chromium adsorption efficiency (E) obtained by experiment was 99.33%. The back-propagation (BP) algorithm optimized by particle swarm optimization (PSO) was used to model chromium adsorption. The topology of BP was determined to be 4-8-1. The optimization process converged quickly with a high target accuracy, and the best training performance was 0.0098601 at 13 epochs for BP algorithm. The Pearson correlation coefficient (P) between the experimental data and the data predicted by PSO-BP model was 0.99914, with the mean absolute error (MAE) being 0.1974, and the maximum E predicted by PSO-BP model was 99.88%.

Automated summary

This paper studied a simple and effective method for chromium removal in wastewater using magnetic nanoparticles in a gradient magnetic field. The optimal process conditions were determined, and a PSO-BP model was established to predict chromium adsorption efficiency.