Using intelligent models for prediction of density of methyl tert-butyl ether (MTBE) as an aviation fuel additive

This study highlights the application of two intelligent models named radial basis function networks optimized by genetic algorithm (GA-RBF) and least square support vector machine optimized by coupled simulated annealing (CSA-LSSVM) for estimation of density of a fuel additive known as MTBE at different temperature and pressures. The predictions of two models were evaluated by using graphical and statistical methods. Results showed that although both models perform accurate results, the CSA-LSSVM model represented the most reliable and dependable predictions. Results of this work also showed that intelligent models are accurate predictive tools to be implemented for prediction of physical properties of different fuel additives including MTBE.

Keywords:

MTBE , Intelligent models , Density