The quantitative model of hearth activity based on data mining and characteristics of deadman

Abstract

In order to obtain the quantitative model of hearth activity, the key control parameters of hearth activity were mined based on big data technology, the distribution characteristics of slag-iron-coke in deadman and the change of voidage in deadman were analyzed, and then the quantitative model of hearth activity was constructed. The results show that a temperature ratio was selected as the characterization parameter of hearth activity combined with some manifestations when the hearth activity became poor, through data sensitivity analysis. Blast speed, average particle size of coke, binary basicity of slag and Ti in iron were preliminarily selected as the key control parameters based on data correlation. The distribution of slag-iron-coke in hearth showed that the average particle size, CSR of coke and the fluidity of slag iron were very important for hearth activity. The coke particle size and the blast momentum should be given priority consideration according to the voidage of deadman in hearth. Finally, the quantitative model of hearth activity was constructed, the accuracy of the model was verified through the data of a commercial BF, so the hearth activity of any BF can be calculated by substituting the known production data of BF.