DONG QIU et al: A STUDY ON PREDICTION OF PHOSPHORUS CONTENT FOR AOD FURNACE

AOD (Argon Oxygen Decarburization) smelting low carbon ferrochrome is extremely complex and multiphase, which is carried out under high temperature chemical process of nonlinear physical. The discretion of the phosphorus’ impurity content is an important factor to affect the quality of ferrochrome products. In order to predict the quality of the product control, it is an optimization method taking the grey incidence degree between the network’s output and input as an objective. The RBF neural network predictive control theory is taken as the foundation and then used as an identification prediction model of AOD furnace system. At the same time, the modified particle swarm optimization (PSO) is used as an optimization algorithm of the network predictive controller to get the online prediction phosphorus content control. The simulation results show that the accuracy of optimization control model is 95.4% with error rate of ±0.003%, which realizes the effective and optimal control of phosphorus content. The established prediction model is an optimal process control model, which provides an important theoretical support for improving the quality of low carbon ferrochromium.