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.
View Article and Find Full Text PDFRigorous process models provide a reliable basis for model-based monitoring and control of anaerobic digestion plants. Due to the complex model structure and non-linear system characteristics, the established Anaerobic Digestion Model No. 1 (ADM1) is rarely applied in industrial plant operation.
View Article and Find Full Text PDFDue to a limited number of available measurements on agricultural biogas plants, established process models, such as the Anaerobic Digestion Model No. 1 (ADM1), are rarely applied in practise. To provide a reliable basis for model-based monitoring and control, different model simplifications of the ADM1 were implemented for process simulation of semi-continuous anaerobic digestion experiments using agricultural substrates (maize silage, sugar beet silage, rye grain and cattle manure) and industrial residues (grain stillage).
View Article and Find Full Text PDFDifferent model structures were compared to simulate the characteristic process variables of the anaerobic digestion of maize, sugar beet and grain silage. Depending on the type and number of the required components, it can be shown that in comparison to the complex Anaerobic Digestion Model No. 1 (ADM1) different simplified model structures can describe the gas production rate, ammonia nitrogen and acetate concentration or pH value equally well.
View Article and Find Full Text PDF