Machine Learning Algorithms for Temperature Management in the Anaerobic Digestion Process

Machine learning for high solid anaerobic digestion: Performance prediction and optimization

Biogas production through anaerobic digestion (AD) is one of the complex non-linear biological processes, wherein understanding its dynamics plays a crucial role towards process control and optimization. In this work, a machine learning based biogas predictive model was developed for high solid systems using algorithms, including SVM, ET, DT, GPR, and KNN and two different datasets (Dataset-1:10, Dataset-2:5 inputs). Support Vector Machine had the highest accuracy (R2) of all the algorithms at 91 % (Dataset-1) and 87 % (Dataset-2), respectively. The statistical analysis showed that there was no significant difference (p = 0.377) across the datasets, wherein with less inputs, accurate results could be predicted. In case of biogas yield, the critical factors which affect the model predictions include loading rate and retention time. The developed high solid machine learning model shows the possibility of integrating Artificial Intelligence to optimize and control AD process, thus contributing to a generic model for enhancing the overall performance of the biogas plant.

Intelligent approaches for sustainable management and valorisation of food waste

Food waste (FW) is a severe environmental and social concern that today's civilization is facing. Therefore, it is necessary to have an efficient and sustainable solution for managing FW bioprocessing. Emerging technologies like the Internet of Things (IoT), Artificial Intelligence (AI), and Machine Learning (ML) are critical to achieving this, in which IoT sensors' data is analyzed using AI and ML techniques, enabling real-time decision-making and process optimization. This work describes recent developments in valorizing FW using novel tactics such as the IoT, AI, and ML. It could be concluded that combining IoT, AI, and ML approaches could enhance bioprocess monitoring and management for generating value-added products and chemicals from FW, contributing to improving environmental sustainability and food security. Generally, a comprehensive strategy of applying intelligent techniques in conjunction with government backing can minimize FW and maximize the role of FW in the circular economy toward a more sustainable future.

Metaheuristic optimization of data preparation and machine learning hyperparameters for prediction of dynamic methane production

Machine learning algorithms provide detailed description of the anaerobic digestion process, but the impact of data preparation procedures and hyperparameter optimization has rarely been investigated. A genetic algorithm was developed for optimizing data preparation and model hyperparameters to simulate dynamic methane production from steady-state anaerobic digestion of agricultural residues at full-scale. A long short-term memory neural network was used as prediction model. Results indicate that batch size, learning rate and number of neurons are the most important model parameters for accurate description of methane production rates, whereas combination of hyperparameter and data preparation optimization shows best model efficiencies, with a root mean square scaled error of 76.5 %. Mass of solid feed, time and mass of volatile solids are the most relevant input features. This study provides fundamental steps for optimal prediction of dynamic biomethane production, as a reliable basis for improving bioconversion efficiency during anaerobic digestion of agricultural residues.

Review of explainable machine learning for anaerobic digestion

Anaerobic digestion (AD) is a promising technology for recovering value-added resources from organic waste, thus achieving sustainable waste management. The performance of AD is dictated by a variety of factors including system design and operating conditions. This necessitates developing suitable modelling and optimization tools to quantify its off-design performance, where the application of machine learning (ML) and soft computing approaches have received increasing attention. Here, we succinctly reviewed the latest progress in black-box ML approaches for AD modelling with a thrust on global and local model interpretability metrics (e.g., Shapley values, partial dependence analysis, permutation feature importance). Categorical applications of the ML and soft computing approaches such as what-if scenario analysis, fault detection in AD systems, long-term operation prediction, and integration of ML with life cycle assessment are discussed. Finally, the research gaps and scopes for future work are summarized.