A model predictive approach to Blast Furnace operational management Automation

Blast furnace operational management automation using modelling and real-time predictive solutions for the object control are considered. Main features of the proposed control are: using of an operational data mining software to identify effective clusters of the furnace regime parameters values; real-time software for identification of the furnace cohesion zone parameters for the operational management correction; dynamics forecasting of the furnace thermal state indicators when charge load and blast parameters change. Usage of the software permits to achieve effective values of the furnace regime parameters with high productivity and reduced coke consumption. It is effective in conditions of the significant charge parameters changes, due to using of source materials from different suppliers. Therewith, forecasting of parameters dynamics allows supervisor to stabilize the blast furnace process in the effective regime. The system is based a joint development of the South Ural State University (National Research University), "Polytech-Automatica" Research & Production Ltd. (Chelyabinsk) and "AKOMM" Ltd. (Moscow).