ANFIS modeling a decontamination process from water with MATLAB

In this paper, it was applied an Adaptive Neuro-Fuzzy Inference System (ANFIS) for modeling remove of an organic contamination from water. Limited experimental data was used for training and testing ANFIS configuration with the help of hybrid learning algorithm consisting of back propagation and least-squares estimation. Time and ozone flow rate were the input of model, whereas the output variable was degradation efficiency. The predicted values were found to be in good agreement with the experimental values with less root mean square error (RMSE=0.019916) and high determination coefficient (R2=0.9999).