Combination of ANNs and heuristic algorithms in modelling and optimizing of Fenton processes for industrial wastewater treatment

This study evaluated the COD removal performances of classical-Fenton and photo-Fenton Processes by different prediction models. To optimize both Fenton processes performed in batch reactors, effects H2O2 dose, Fe(II) H2O2/Fe(II) rate, contact time were determined as independent variables Besides response surface methodology, three neural networks used to more reliably effectively predict behaviors dependent at values relevant parameters. Multi-Layer Perceptron trained Levenberg–Marquardt (MLP-LM), Single Multiplicative Neuron models Particle Swarm Optimization algorithm (MLP-PSO; SMN-PSO) studied. Models’ Root-Mean-Square Error (RMSE) Mean Absolute Percent criteria. Regression analysis was applied determine performance best model. The results from criteria indicated that SMN-PSO model produced predictive almost all cases. Moreover, key process parameters applying genetic outputs. optimized conditions achieved optimum with over 99% desirability. dose 399.99 mg/L processes. found 726.18 894.07 efficiencies 86.50 87.49% for classical photo-Fenton, respectively. As a result, it will be possible simulate improve obtained data treatment wastewater similar characteristics without many experiments, which are difficult costly.