The strength prediction model of iron ore sinter based on an artificial neural network

ABSTRACT

The iron ore sinter is still the main raw material for the blast furnace ironmaking process, its properties, such as strength and reducibility, are of vital importance to the productivity and the smooth operation of the blast furnace. In the present study, one model based on an artificial neural network (ANN) was established to predict the sinter strength. The ANN model was trained with the sample set, which was generated from the credible data from the published papers. The comparison between the direct prediction model and the indirect prediction model with the amount of liquid phase and spinal phase calculated with thermodynamic theory as the middle layer was carried out. The results show that the indirect ANN model gave much higher accurate prediction results than that of the direct one without the middle layer.The parametric study with the validated model shows that the sinter strength increased first with increasing the SiO₂ to 5.4% and then decreased with further increasing the SiO₂.

KEYWORDS:

• sinter
• artificial neural network
• indirect model
• FactSage
• liquid phase
• spinal phase
• strength
• univariate analysis

Acknowledgements

The authors would like to express their gratitude for the financial support of the National Natural Science Foundation of China (grant nos. 51974048 and 51974053), Natural Science Foundation of Chongqing (grant no. cstc2021jcyj-msxmX0882), the Innovation research group of universities in Chongqing (CXQT21030) and Chongqing University of Science and Technology Graduate Research Innovation Project (grant no. YKJCX2120225).

Disclosure statement

No potential conflict of interest was reported by the author(s).The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Additional information

Funding

The authors would like to express their gratitude for the financial support of the National Natural Science Foundation of China [grant nos. 51974048 and 51974053], Natural Science Foundation of Chongqing [grant no. cstc2021jcyj-msxmX0882], the Innovation research group of universities in Chongqing [CXQT21030] and Chongqing University of Science and Technology Graduate Research Innovation Project [grant no. YKJCX2120225].