Меню
Vol. 2 No. 31 (2026) / 2026-06-30 / Number of views: 147
Authors
Keywords
Link to DOI:
How to quote
The article presents the development of a software and hardware complex (PAK) for intelligent control of the flotation process at a processing plant for conditions of variability in the properties of ore raw materials and non-standard technological modes. The relevance of the study is primarily due to the fact that the currently used traditional control systems at processing plants do not effectively maintain optimal flotation parameters with changes in the granulometric and chemical composition of the pulp, including energy efficiency and environmental safety of production.
The purpose of the research is to develop a management system architecture that combines machine vision, multisensory monitoring, and hybrid artificial intelligence models that support and automate management decision-making. The object of the study is the technological contour of the flotation of the Nurkazgan processing plant. The work uses methods of computer vision, deep machine learning (CNN, LSTM, Temporal Transformer), neural networks (PINNs) and methods of multi-criteria mode optimization.
The qualitative results consist in the transition from limited monitoring to an integrated digital control loop, with the ability to take into account the temporal dynamics of the process, applicability to changes in ore properties, integration with existing automated process control systems and SCADA. Quantitative estimates based on data from the pre-design survey and test modeling create prerequisites for reducing equipment failures by up to 10-15%, reducing downtime by up to 15%, reducing reagent consumption by 5-10%, and increasing the extraction of useful components from 1-3%.
The practical significance of the research lies in the possibility of implementing and scaling the developed PAK at other enterprises of the mining and metallurgical complex of the Republic of Kazakhstan.