Data-Driven Model for On-Line Steel Temperature Prediction to Optimize the Secondary Refining Processes

An on-line data-driven model was developed to predict the steel temperature at the ladle furnace (LF), vacuum degasser and continuous casting machine (CCM) stations. The learned model, trained on plant data automatically collected by the system, estimates the current temperature in the ladle along the secondary metallurgy processes. The model is also used to calculate the target temperature at the LF exit to reach the required CCM target, adapting to the current production conditions. Model results on a pilot data set show good predictive accuracy. By providing operators with the current process status and targets calculated dynamically, the system helps increase the consistency and efficiency of secondary refining processes.

Mr. Manuele Piazza | Danieli & C. Officine Meccaniche S.p.A.
 

Mr. Giulio Planu | Danieli Automation S.p.A.
 

Mr. Marco Ometto | Danieli Automation S.p.A.