TY - JOUR AU - Cantera, C. AU - Jiménez, J. AU - Varela, I. AU - Formoso, A. PY - 2002/08/30 Y2 - 2024/03/28 TI - Hot metal temperature prediction by neural networks in the blast furnace JF - Revista de Metalurgia JA - Rev. metal. VL - 38 IS - 4 SE - Articles DO - 10.3989/revmetalm.2002.v38.i4.406 UR - https://revistademetalurgia.revistas.csic.es/index.php/revistademetalurgia/article/view/406 SP - 243-248 AB - Based on a simplified model, the underlying temperature criteria is proposed as a method to study the temperature trends in a blast furnace. As, an application, a neural network able to forecast hot metal temperatures from 2 to 16 h in advance (with decreasing precision) has been built. This neural network has been designed to work at real time in a production plant. ER -