Prediction of the Amount of Sulfur Adsorption Using a GA-RBF Model

This work highlights the application of one type of artificial neural network namely GA-RBF for prediction of sulfur adsorption in liquid phase of hydrocarbon solution of isotherm batch systems. The precision and reliability of developed model were studied by various graphical and statistical approaches.Initial sulfur concentration, utilized adsorbent weights, molecular weights and densities of solvent and solute, average size of adsorbent particle, Si/ Al ratio of adsorbent and temperature were used as input parameters of model and the model output was the amount of adsorption. Results show that the predictionsof GA-RBF model is precise and reliable. The overall values of R2 and AARD% for GA-RBF model were 0.998 and 2.21 that show precision and robustness of the applied model.

Keywords:

Sulfur adsorption , Model , GA-RBF , Prediction