Modelling Caffeine and Paracetamol Removal from Synthetic Wastewater Using Nanofiltration Membranes: A Comparative Study of Artificial Neural Networks and Response Surface Methodology

Abstract

The integration of computational intelligence techniques into pharmaceutical wastewater treatment offers promising opportunities to improve process efficiency and minimize operational costs. This study compares the predictive capabilities of Response Surface Methodology (RSM) and Artificial Neural Network (ANN) models in forecasting the rejection efficiencies of caffeine and paracetamol using AFC 40 and AFC 80 nanofiltration (NF) membranes. Experiments were conducted under varying operating conditions, including transmembrane pressure, feed concentration, and flow rate. The predictive performance of both models was evaluated using statistical metrics such as the Coefficient of Determination (R2), Root Mean Square Error (RMSE), Marquardt’s Percentage Squared Error Deviation (MPSED), Hybrid fractional error function (HYBRID), and Average Absolute Deviation (AAD). Both models demonstrated strong predictive accuracy, with R2 values of 0.9867 and 0.9832 for RSM and ANN, respectively, in AFC 40 membranes, and 0.9769 and 0.9922 in AFC 80 membranes. While both approaches closely matched the experimental results, the ANN model consistently yielded lower error values and higher R2 values, indicating superior predictive performance. These findings support the application of ANNs as a robust modelling tool in optimizing NF membrane processes for pharmaceutical removal.