Meta-tuning and fast optimization of machine learning models for dynamic methane prediction in anaerobic digestion

• Published in: Bioresource technology Vol. 432; p. 132654
• Main Authors: Meola, Alberto, Wolf, Klara, Weinrich, Sören
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.09.2025
• Subjects: Algorithms; anaerobic digestion; Anaerobiosis; Artificial intelligence; Bayes Theorem; Bayesian theory; Biofuels; biogas; Biogas technology; data collection; Data processing; income; Machine Learning; methane; Methane - biosynthesis; Methane - metabolism; methane production; Neural Networks, Computer; Parameter estimation; prediction; Process modelling; Process modelling; Data processing; Parameter estimation; Biogas technology; Artificial intelligence
• ISSN: 0960-8524; 1873-2976; 1873-2976
• DOI: 10.1016/j.biortech.2025.132654

[Display omitted] •Bayesian Search is recommended for general biogas prediction routine optimization.•Simple optimization scenarios can be optimized with a 50-step optimization process.•Complex scenarios including neural networks require more effective optimization.•Meta-tuning has a positive influence on prediction results in complex scenarios. This study evaluates the performance of several optimization algorithms for tuning a data preparation and hyperparameter optimization pipeline applied to machine and deep learning models predicting methane production. Bayesian ridge regression and recurrent neural networks were applied to steady-state and dynamic datasets. Results show that 50 optimization steps are sufficient for optimal performance in simpler cases (62.8 % model accuracy). For complex scenarios, such as recurrent neural networks on dynamic datasets, extended optimization processes improve accuracy. Among the tested algorithms, Bayesian Search performed well without meta-tuning. However, meta-tuned Genetic Algorithm performed better (94.4 % vs 99.2 % baseline). Meta-tuning improves tuning parameter selection and model precision. Differential Evolution and Particle Swarm Optimization with time-varying acceleration also performed well, particularly in steady-state. These findings highlight the need to match optimization to dataset and model complexity, with meta-tuning offering advantages in challenging cases. Improved accuracy can increase revenue in flexible biogas operations.