Applications of artificial neural networksin electric arc furnaces and continuous casting facilities

Various complex calculations and problems may arise during the control of production or services performed in industrial plants by automation systems using classical logical algorithms. It is important to model the system in order to solve these problems. System modeling is to obtain an unknown system model based on the data obtained with various mathematical equations. In this study, Artificial Neural Networks in different architectures in order to provide more stable conditions and to provide more stable conditions for the chronic problems experienced in electric arc furnace scrap melting plant which is important in iron and steel sector and continuous casting machines where molten liquid steel in meltshop is converted into SLAB semi-product. data modeling and estimation characteristics of Artificial Neural Networks were compared. In Artificial Neural Networks, electric arc furnace electrode control modeling has achieved over 95% success, while SLAB continuous casting machine has achieved over 98% success in mold breakout prediction. Artificial neural network-based control software is more ergonomic for the user than conventional logical software that requires continuous calibration and parameter change. Artificial Neural Network-based software has been able to make successful predictions on patterns of breakout that cannot be detected by conventional logical software.