TY - JOUR
T1 - Recent advances and future prospects of thermochemical biofuel conversion processes with machine learning
AU - Jeon, Pil Rip
AU - Moon, Jong Ho
AU - Ogunsola, Nafiu Olanrewaju
AU - Lee, See Hoon
AU - Ling, Jester Lih Jie
AU - You, Siming
AU - Park, Young Kwon
N1 - Publisher Copyright:
© 2023 Elsevier B.V.
PY - 2023/9/1
Y1 - 2023/9/1
N2 - Biofuels have been widely recognized as potential solutions to addressing the climate crisis and strengthening energy security and sustainability. However, techno-economic and environmental challenges for the production of biofuels remain and complicated conversion processes and factors, such as materials and process design, need to be taken into consideration for solving the challenges, which is not easy. Machine Learning (ML) has been combined with the theories of thermochemical biofuel conversion processes to achieve accurate and efficient biofuel process modelling. In this review, existing ML applications to predict biofuel yield and composition are critically reviewed. The details of the input and output variables of the developed models for thermochemical biofuel conversion processes were summarized, and their development procedures were compared. Techno-economic analysis results incorporating ML applications in biofuels were also reviewed. Although developed models in literature showed good performance for their targets, respectively, they can hardly be applied to other feedstocks or operating conditions. To overcome the challenge and develop universal model, perspective approaches were suggested in this study. It was suggested that it is essential to develop systematic datasets to support more comprehensive machine learning-based modelling towards practical applications. Potential prospective research and development directions on machine learning-based thermochemical biofuel conversion process modeling were recommended, so that it can assist in the commercialization and optimization of various biofuel conversions leading to a sustainable and circular society.
AB - Biofuels have been widely recognized as potential solutions to addressing the climate crisis and strengthening energy security and sustainability. However, techno-economic and environmental challenges for the production of biofuels remain and complicated conversion processes and factors, such as materials and process design, need to be taken into consideration for solving the challenges, which is not easy. Machine Learning (ML) has been combined with the theories of thermochemical biofuel conversion processes to achieve accurate and efficient biofuel process modelling. In this review, existing ML applications to predict biofuel yield and composition are critically reviewed. The details of the input and output variables of the developed models for thermochemical biofuel conversion processes were summarized, and their development procedures were compared. Techno-economic analysis results incorporating ML applications in biofuels were also reviewed. Although developed models in literature showed good performance for their targets, respectively, they can hardly be applied to other feedstocks or operating conditions. To overcome the challenge and develop universal model, perspective approaches were suggested in this study. It was suggested that it is essential to develop systematic datasets to support more comprehensive machine learning-based modelling towards practical applications. Potential prospective research and development directions on machine learning-based thermochemical biofuel conversion process modeling were recommended, so that it can assist in the commercialization and optimization of various biofuel conversions leading to a sustainable and circular society.
KW - Biofuel conversion
KW - Techno-economic analysis
KW - Theory-integrated machine learning
KW - Thermochemical conversion processes
UR - http://www.scopus.com/inward/record.url?scp=85164368988&partnerID=8YFLogxK
U2 - 10.1016/j.cej.2023.144503
DO - 10.1016/j.cej.2023.144503
M3 - Review article
AN - SCOPUS:85164368988
SN - 1385-8947
VL - 471
JO - Chemical Engineering Journal
JF - Chemical Engineering Journal
M1 - 144503
ER -