Performance evaluation of adaptive neuro-fuzzy inference system and response surface methodology in modeling biodiesel synthesis from jatropha-algae oil

Biodiesel production from different feedstocks is an effective method of resolving problems related to the fuel crisis and environmental issues. In this study, an adaptive neuro-fuzzy inference system (ANFIS) and the response surface methodology based Box-Behnken experimental design were used to model the parameters of biodiesel production for a jatropha-algae oil blend, including the molar ratio, temperature, reaction time, and catalyst concentration. A significant regression model with an R-2 value of 0.9867 was obtained under a molar ratio of 6-12, KOH of 0-2% w/w, time of 60-180min, and temperature of 35-55 degrees C using response surface methodology (RSM). The ANFIS model was used to individually correlate the output variable (biodiesel yield) with four input variables. An R-2 value of 0.9998 was obtained in the training. The results demonstrated that the developed models adequately represented the processes they described.