Development of an information and control system architecture for energy-efficient iron ore processing

Abstract

Analysis of recent research and publications. The high energy intensity of iron ore processing necessitates the implementation of control systems based on streaming data analysis. Recent studies confirm the effectiveness of integrating IIoT platforms, edge computing, and cloud services. Traditional hierarchical architectures are limited by weak horizontal integration and a lack of predictive control tools. Designing multi-level systems to process data arrays from on-stream quality analyzers while accounting for the RAMI 4.0 reference model and strict cybersecurity requirements requires a comprehensive approach.

Purpose of the research. To substantiate a comprehensive multi-level architecture of an information and control system aimed at minimizing the specific energy consumption of beneficiation processes through multi-criteria data processing in accordance with the ISA-88, ISA-95, and ISA/IEC 62443 industry standards.

Presentation of the main research material. The proposed architecture is based on the RAMI 4.0 concept. At the field level, on-stream analyzers are integrated via Asset Administration Shells to guarantee semantic interoperability. The implemented Edge level provides primary aggregation, normalization, and local analytics of technological data, critically reducing the load on industrial networks and transmission latency. The cloud level operates with specialized time-series databases and digital twins of process cells. Energy consumption optimization is achieved by combining physics-based models and machine learning methods within a unified Model Predictive Control loop. The system's cybersecurity is implemented through spatial distribution into secure zones using hardware data diodes.

Conclusions. The developed architecture constitutes a scientific and practical foundation for the digital transformation of the mineral processing industry. The application of hybrid predictive models and edge analytics lays the technological groundwork for reducing specific electricity consumption by 8–15%, increasing the iron content in the concentrate by 0.5–1.5%, and minimizing unscheduled downtime. The prospect for further research is the verification of the digital twin mathematical models on real industrial datasets.