Artificial neural network for modelling the removal of pollutants A review

摘要

Modeling of pollutant degradation using artificial neural networks (ANN) has been done well. The techniques used to model degradation vary. This literature review was done to examine the development of the use of ANN modeling from year to year. It will provide an overview of predictive studies from a degradation treatment condition that will produce optimal conditions. These conditions will be supported by experimental data so that costs and time can be reduced at laboratory scale. Some relevant techniques include separation methods, coagulation, advanced oxidation processes, and chemical oxidation. The algorithmic approaches used are ANN-LM, ANN-BP, ANN-BP (SCG), and ANN-BFGS. Modelling using ANN has very high potential for further development. The perfomance indicator of an ANN method is a strong coefficient of determination (R2), with good RMSE, MAPE, and MSE values.