Neural network model of burden layer formation dynamics in the blast furnace

A model of the thickness of burden layers in the ironmaking blast furnace i s presented. Local layer thickness estimates are calculated on the basis of signals from stockrods that measure the burden (stock) level in the furnac e. These estimates are used in developing a model for the relation between the layer thickness and variables such as stock level and movable armor set tings. Because of the nonlinear dependence of the variables, the models are based on feedforward or recurrent neural networks. The network size is car efully selected based an a cross-validation procedure. The resulting neural model is first studied by analyzing its predictions for different inputs. By further introducing a simplified scheme for considering the practical co nstraints of the charging process, an autonomous model, where the neural ne twork plays an important rots, is formed. This hybrid model is applied to y ield insight into the dynamics of the layer formation process; the effect o f movable armor settings, stock level and burden descent rate are analyzed and compared with practical experience.