Ultrasound assisted biodiesel production from sesame (Sesamum indicum L.) oil using barium hydroxide as a heterogeneous catalyst: Comparative assessment of prediction abilities between response surface methodology (RSM) and artificial neural network (ANN)

Abstract

The present study estimates the prediction capability of response surface methodology (RSM) and artificial neural network (ANN) models for biodiesel synthesis from sesame (Sesamum indicum L.) oil under ultrasonication (20kHz and 1.2kW) using barium hydroxide as a basic heterogeneous catalyst. RSM based on a five level, four factor central composite design, was employed to obtain the best possible combination of catalyst concentration, methanol to oil molar ratio, temperature and reaction time for maximum FAME content. Experimental data were evaluated by applying RSM integrating with desirability function approach. The importance of each independent variable on the response was investigated by using sensitivity analysis. The optimum conditions were found to be catalyst concentration (1.79wt%), methanol to oil molar ratio (6.69:1), temperature (31.92°C), and reaction time (40.30min). For these conditions, experimental FAME content of 98.6% was obtained, which was in reasonable agreement with predicted one. The sensitivity analysis confirmed that catalyst concentration was the main factors affecting the FAME content with the relative importance of 36.93%. The lower values of correlation coefficient (R2=0.781), root mean square error (RMSE=4.81), standard error of prediction (SEP=6.03) and relative percent deviation (RPD=4.92) for ANN compared to those R2 (0.596), RMSE (6.79), SEP (8.54) and RPD (6.48) for RSM proved better prediction capability of ANN in predicting the FAME content.