Progresses and Challenges of Machine Learning Approaches in Thermochemical Processes for Bioenergy: A Review

• Published in: The Korean journal of chemical engineering Vol. 41; no. 7; pp. 1923 - 1953
• Main Authors: Ogunsola, Nafiu Olanrewaju, Oh, Seung Seok, Jeon, Pil Rip, Ling, Jester Lih Jie, Park, Hyun Jun, Park, Han Saem, Lee, Ha Eun, Sohn, Jung Min, Lee, See Hoon
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.07.2024; Springer Nature B.V; 한국화학공학회
• Subjects: Alternative energy sources; Biofuels; Biomass; Biomass energy production; Biotechnology; Catalysis; Chemistry; Chemistry and Materials Science; Industrial Chemistry/Chemical Engineering; Machine learning; Materials Science; Renewable energy; Review Article; Robustness (mathematics); 화학공학; Sustainable biomass utilization; Thermochemical conversion; Bioenergy; Machine learning; Artificial neural networks
• ISSN: 0256-1115; 1975-7220
• DOI: 10.1007/s11814-024-00181-7

Thermochemical conversions of nonedible biomass into energy are promising alternatives for ensuring a sustainable energy society. However, determining the optimum design and operating conditions of the processes remains a major challenge due to the laborious and costly experimental methods. Machine learning techniques are cost-effective and non-time consuming and have been widely utilized in thermochemical conversion process modelling with robust and accurate results and solutions. Nonetheless, no standard method has been proposed for applying ML models to biomass thermochemical processes. Consequently, the general development procedure for ML models with high accuracy and robustness remains unclear. This review provides a comprehensive review of machine learning techniques for predicting biofuel yield and composition. It is recommended that quality datasets be ensured to enable the development of more robust machine learning-aided models for practical engineering applications. Finally, solutions to the identified challenges and prospective future research directions on machine learning-based biomass thermochemical conversion processes are recommended to accelerate the optimization and large-scale deployment of these processes.