Improved estimation of Cetane number of fatty acid methyl esters (FAMEs) based biodiesels using TLBO-NN and PSO-NN models

• Published in: Fuel (Guildford) 2018-11, Vol.232, p.620-631
• Main Authors: Baghban, Alireza, Kardani, Mohammad Navid, Mohammadi, Amir H.
• Format: Article
• Language: English
• Subjects: Algorithms; ANN algorithm; Artificial neural networks; Biodiesel; Biodiesel fuels; Cetane number; Diesel fuels; Esters; Evolutionary algorithm; FAME; Fatty acids; Genetic algorithms; Ignition; Learning theory; Neural networks; Particle swarm optimization
• ISSN: 0016-2361; 1873-7153
• DOI: 10.1016/j.fuel.2018.05.166

[Display omitted] •Cetane number (CN) of biodiesel based on fatty acid methyl esters (FAMEs) composition is modeled.•PSO-ANN and TLBO-ANN are developed for modeling CN.•A number of 232 fuel samples derived from the literature was used for the models development.•Different evaluative factors prove satisfactory performances of the proposed ANN models. Cetane number (CN) is one of the key factors of biodiesels and other diesel fuels. It is an indicator of ignition speed and required compression for ignition. CN estimation of biodiesel based on fatty acid methyl esters (FAME) composition was the main goal of this work. Application of artificial neural network (ANN) combined with particle swarm optimization (PSO) and teaching-learning based optimization (TLBO) is discussed in this communication. A number of 232 fuel samples was derived from the literature as the raw data for the models development. Different evaluative factors prove the satisfactory performance of the proposed ANN models. The obtained values of R-squared and mean square of errors are 0.973 & 3.538 and 0.951 & 6.324 for the proposed TLBO-ANN and PSO-ANN, respectively. Based on the outcome of this study, ANN coupled with PSO and TLBO algorithms can be suitable tools, especially TLBO algorithm to estimate CN of biodiesels.