TY - JOUR
T1 - Performance evaluation of artificial neural networks in sustainable modelling biodiesel synthesis
AU - Treve, Mark
AU - Patra, Indrajit
AU - Prabu, P.
AU - Rama Sree, S.
AU - Keerthi Kumar, N.
AU - Methkal Abd Algani, Yousef
AU - Kiran Bala, B.
AU - Balaji, S.
N1 - Publisher Copyright:
© 2022 Elsevier Ltd
PY - 2022/8
Y1 - 2022/8
N2 - Biodiesel is a characteristic and inexhaustible homegrown fuel removed from creature fats or vegetable oil and liquor through a transesterification response. The exploration work means to assess the exhibition of biodiesel blend. In this paper, biodiesel was displayed and improved by utilizing a hereditary calculation and Artificial Neural Network (ANN). In AI, hereditary calculations and counterfeit neural organizations assume a significant part in displaying biodiesel blend. To upgrade an excellent arrangement hereditary calculation was created. The mix of ANN and Genetic Algorithm gives the ideal condition as the temperature of methanol molar proportion, impetus fixation. It tentatively decides the exhibition trademark like the Coefficient of determination and Absolute Average deviation (AAD). It predicts the Fatty Acid Methyl Ester (FAME) model productively than Response Surface Methodology (RSM). The exhibition examination is reenacted and hypothetical outcomes are recorded then it is contrasted with constant information to decide the exactness of ANN.
AB - Biodiesel is a characteristic and inexhaustible homegrown fuel removed from creature fats or vegetable oil and liquor through a transesterification response. The exploration work means to assess the exhibition of biodiesel blend. In this paper, biodiesel was displayed and improved by utilizing a hereditary calculation and Artificial Neural Network (ANN). In AI, hereditary calculations and counterfeit neural organizations assume a significant part in displaying biodiesel blend. To upgrade an excellent arrangement hereditary calculation was created. The mix of ANN and Genetic Algorithm gives the ideal condition as the temperature of methanol molar proportion, impetus fixation. It tentatively decides the exhibition trademark like the Coefficient of determination and Absolute Average deviation (AAD). It predicts the Fatty Acid Methyl Ester (FAME) model productively than Response Surface Methodology (RSM). The exhibition examination is reenacted and hypothetical outcomes are recorded then it is contrasted with constant information to decide the exactness of ANN.
KW - Absolute average deviation (AAD)
KW - Biodiesel
KW - Catalyst concentration
KW - Fatty acid methyl ester (FAME)
KW - Genetic algorithm
KW - Machine learning (ML)
KW - Response surface methodology (RSM)
KW - Transesterification
UR - http://www.scopus.com/inward/record.url?scp=85125008191&partnerID=8YFLogxK
U2 - 10.1016/j.seta.2022.102098
DO - 10.1016/j.seta.2022.102098
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:85125008191
SN - 2213-1388
VL - 52
JO - Sustainable Energy Technologies and Assessments
JF - Sustainable Energy Technologies and Assessments
M1 - 102098
ER -