[This article belongs to Volume - 55, Issue - 12]
Gongcheng Kexue Yu Jishu/Advanced Engineering Science
Journal ID : AES-09-01-2024-665

Title : Machine Learning-driven Predictive Modeling of Biodiesel Characteristics with Heterogeneous Catalysts: A Review
vartika gupta, Kishan Pal Singh,

Abstract : It is generally acknowledged that energy is a basic requirement for humankind's daily survival. However, due to the steadily increasing need for energy, the researcher's focus has switched to renewable energy sources in an effort to reduce reliance on fossil fuels. Additionally, these fossil fuels' detrimental effects on the biosphere contribute to global warming. In that circumstance, biodiesels are the most useful type of green energy. Typically, a transesterification process using an appropriate catalyst is used to create biodiesel. Depending on the fatty acids, the catalyst can be homogeneous or heterogeneous; in this study, heterogeneous catalyst was utilised since it is reusable and simple to separate. A precise modelling tool, such as machine learning (ML) and artificial intelligence, is needed for biodiesel production since it is a complex process that deals with non-linear interactions between input and output variables. Machine learning is helpful for modelling the transesterification process, physico-thermal properties, and combustion engines that burn biodiesel because it produces superior results with high precision and is stimulated by the human brain's automatic learning. This review article discusses how machine learning technology is being used to forecast the characteristics of biodiesel, including its cetane number, viscosity, heating value, cold flow properties, oxidation stability, density, flash point, and iodine value.