Comparative adsorptive removal of Reactive Red 120 using RSM and ANFIS models in batch and packed bed column

The present study investigated the removal efficiency of Reactive Red 120 (RR120) using biochar produced from a seaweed Ulva reticulata using the predictive models namely Response Surface Methodology (RSM) and Adaptive Neuro-Fuzzy Inference System (ANFIS). The experiments were conducted in both batch and continuous processes, and the coded variables namely biochar dose, pH, Initial RR120 concentration, and temperature were studied for batch process, and the coded variables namely sorbent depth, solute flowrate. and initial RR120 concentration were studied for the continuous process. The correlation coefficient of the RSM and ANF1S for the batch process was obtained as 0.9977 and 0.9999. Similarly, for the continuous process, the correlation coefficient of 0.9979 and 0.9997 was obtained for RSM and ANFIS. Further statistical error analysis was conducted to find the goodness of the model with the experimental values. A comparison study was arrived based on the cluster analysis of experimental, RSM, and ANFIS models. The results concluded that the ANFIS model was superior to RSM in the prediction of the removal efficiency of the Reactive Red 120.

Automated summary

This study investigated the efficiency of removing Reactive Red 120 dye using seaweed biochar produced from Ulva reticulata. Predictive models, namely Response Surface Methodology (RSM) and Adaptive Neuro-Fuzzy Inference System (ANFIS), were used. The results showed that the ANFIS model outperformed the RSM model in predicting the removal efficiency of Reactive Red 120.