Insights into Methylene Blue Removal: A Comparative Study of Soft Computing Predictive Models

This research explored the adsorption of methylene blue using various adsorbents under varying experimental conditions. The process was modeled and analyzed utilizing advanced computational approaches, including Adaptive Neuro-Fuzzy Inference Systems (ANFIS), artificial neural networks with MultiLayer Perceptron (MLP) architectures, and Radial Basis Function (RBF) networks. Data were collected through a comprehensive literature review, followed by the selection of pertinent studies and the systematic extraction and structuring of information on methylene blue removal by eight adsorbents across diverse environmental parameters. The study detailed the application of these computational models, emphasizing data normalization, splitting the data into separate sets for training and testing, and assessing model performance using key metrics such as efficiency (EF), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE). The primary goal of this study was to identify the most accurate predictive model for assessing methylene blue removal efficiency. Additionally, the investigation focused on evaluating the influence of key hyperparameters and the model's ability to generalize across diverse datasets. Additionally, a sensitivity analysis was conducted to identify critical environmental variables that influence adsorption efficiency, providing valuable insights for future research in this domain.

کلیدواژه ها:

Methylene blue removal ، adsorbent ، Soft computing ، optimizing ، Sensitivity analysis