Adaptive Neuro Fuzzy Inference System Modeling for Biodiesel Kinetics Study


Abstract

Transesterification of Jatropha curcas for biodiesel production is a kinetic control process, which is complex in nature and controlled by temperature, molar ratio, mixing intensity and catalyst process parameters. A precise choice of catalyst is required to improve the rate of transesterification and to simulate the kinetic study in a batch reactor. The present paper uses Adaptive Neuro-Fuzzy Inference   System (ANFIS) approach to model and simulate the butyl ester production using alkaline catalyst (NaOH). Amount of catalyst and time for reaction were taken as model’s input parameters. The model is the combination of fuzzy inference, artificial neural network, and set of fuzzy rules has been developed directly from experimental data. The proposed modeling approach is verified by comparing the expected results with the observed practical results obtained by conducting the batch reactor operation. The application of ANFIS test shows, that the amount of catalyst predicted by a proposed model is well in agreement with the experimental values at 0.5% level of significance.

Keywords: biodiesel, Neuro-Fuzzy, reactor, kinetic