Machine Learning-Based Tap Temperature Prediction and Control for Optimized Power Consumption in Stainless Electric Arc Furnaces (EAF) of Steel Plants

The steel industry has been forced to switch from the traditional blast furnace electric arc (EAF) process reduce carbon emissions. However, EAF still relies entirely on operators’ proficiency determine electrical power input. This study aims enhance efficiency of by predicting tap temperature in real time through a data-driven approach and applying system that automatically sets input amount production site. We developed prediction model (TTPM) with machine learning (ML)-based support vector regression (SVR) algorithm. operation data stainless EAF, where actual work was carried out, were extracted, models using six ML algorithms trained. validation results show an SVR radial basis function (RBF) algorithm resulted best performance root mean square error (RMSE) 20.14. performed better than others for features such as noise. As result five-month analysis operating TTPM deviation decreased 17% average consumption 282 kWh/heat compared depended operator’s skill. In economic evaluation facility investment, feasibility found be sufficient, internal rate return (IRR) 35.8%. Applying successfully it ten months verified system’s reliability. terms increasing proportion used decarbonize industry, is expected various studies will conducted more actively improve future. contributes improvement companies’ manufacturing competitiveness neutrality achieving energy improvements associated process.