Development of an intellectual information system for managing the operation of a biogas plant

The paper presents the architecture of an intelligent information system for managing the operation of a biogas plant using mathematical modeling methods and neural network algorithms. The technological parameters influencing the efficiency of anaerobic digestion of organic waste, including the chemical composition of raw materials and process conditions, are analyzed. The influence of input parameters on biogas output and methane content is investigated based on historical and laboratory data. It is revealed that the use of intelligent methods makes it possible to increase the accuracy of forecasting in comparison with traditional approaches. A functional modeling of the system using the IDEF0 methodology was carried out, reflecting the key processes: data collection, processing, analysis, and forecast generation. The logical representation of the data is defined and the database structure is constructed in the PostgreSQL relational data management system. It is established that the chosen architecture of the multilayer perceptron provides the necessary balance between the complexity of the model and its generalizing ability. The structure of an intelligent information system for managing the operation of a biogas plant has been formed, adapted to the specifics of such complexes of various scales. It is proposed to use normalization, regularization, and adaptive parameter tuning methods to increase the model's resistance to overfitting. The model is supposed to be integrated into an intelligent information system implemented as a web interface with the ability to visualize parameters in real time, adjust thresholds and output recommendations for adjusting the operating mode of the biogas plant.