Estimation capabilities of biodiesel production from algae oil blend using adaptive neuro-fuzzy inference system (ANFIS)

Biodiesel produced from different raw materials is the most effective way to solve problems related to the fuel crisis and environmental problems. In the present study, the methodology of the adaptive neuro-fuzzy inference system (ANFIS) for the modeling and estimation of a process applied to the production of biodiesel from the blend of algae oil. The Gaussian membership function was applied and studied. The results of ANFIS are compared with the actual results obtained through the experiment using the mean root-mean-square error (RMSE) and the determination coefficient (R-square). The results show an improvement in the prediction, accuracy, and capacity of the ANFIS technique for estimation. The statistical characteristics of RMSE were 0.2179 and R-squared was 0.9998 obtained in training. As the ANFIS offers a good estimation, the modeling quality can be applied to the biodiesel production process as well.