Modeling of cetane number of biodiesel from fatty acid methyl ester (FAME) information using GA-, PSO-, and HGAPSO- LSSVM models

One of the major properties of biodiesel fuels is cetane number (CN) which expresses the ignition characteristics and quality of motor power. The main idea of this work was proposing an accurate model for estimation of the cetane number of biodiesel in terms of fatty acid methyl esters composition. In doing so, least-square support vector machine (LSSVM) approach was coupled with Genetic algorithm (GA), particle swarm optimization (PSO) and hybrid of GA and PSO (HGAPSO) algorithms and a total number of 232 samples of fuels were extracted from literature. The coefficient of determination (R2), mean relative errors (MREs), mean squared errors (MSEs) and standard deviations (STD) were calculated for evaluation of the models. The R2 values in the testing phase for LSSVM-GA, LSSVM-PSO, and LSSVM-HGAPSO were estimated by 0.965, 0.966 and 0.978 respectively. The statistical and graphical analyses showed that the LSSVM algorithm coupled with GA, PSO or HGAPSO algorithms can be used as an accurate model for estimation of the cetane number of the biodiesel fuels.