Blast Furnace Thermal State Prediction Model And Application

Thermal state is a fundamental characterization of blast furnace smelting, and thermal state predicting model is one of basic investigations for realizing intelligence control.Various controlling models related to blast furnace at home and abroad are briefly described; On the basis of computer-based information system of large blast furnaces in Angang, through some countermeasures, such as, combined with expert knowledge of blast furnace, applying forecasting model which combined recurrent neural network with time difference method, an instant calculation method of theoretical combustion temperature for raceway in blast furnace is developed, which may be used for heat and the changing tendency prediction.It is proved by actual application in some blast furnaces of Angang, that the heat prediction model can both precisely forecast Silicon content of hot metal, and calculate instant adiabatic flame temperature. It is convenient for working out routine adjusting parameters and in-situ application.