Research Progress of Intelligent Ore Blending Model
Yifan Li,Bin Wang,Zixing Zhou,Aimin Yang,Yunjie Bai
DOI: https://doi.org/10.3390/met13020379
IF: 2.695
2023-02-16
Metals
Abstract:The iron and steel industry has made an important contribution to China's economic development, and sinter accounts for 70–80% of the blast furnace feed charge. However, the average grade of domestic iron ore is low, and imported iron ore is easily affected by transportation and price. The intelligent ore blending model with an intelligent algorithm as the core is studied. It has a decisive influence on the development of China's steel industry. This paper first analyzes the current situation of iron ore resources, the theory of sintering ore blending, and the difficulties faced by sintering ore blending. Then, the research status of the neural network algorithms, genetic algorithms, and particle swarm optimization algorithms in the intelligent ore blending model is analyzed. On the basis of the neural network algorithm, genetic algorithm and particle swarm algorithm, linear programming method, stepwise regression analysis method, and partial differential equation are adopted. It can optimize the algorithm and make the model achieve better results, but it is difficult to adapt to the current complex situation of sintering ore blending. From the sintering mechanism, sintering foundation characteristics, liquid phase formation capacity of the sinter, and the influencing factors of sinter quality were studied, it can carry out intelligent ore blending more accurately and efficiently. Finally, the research of intelligent sintering ore blending model has been prospected. On the basis of sintering mechanism research, combined with an improved intelligent algorithm. An intelligent ore blending model with raw material parameters, equipment parameters, and operating parameters as input and physical and metallurgical properties of the sinter as output is proposed.
materials science, multidisciplinary,metallurgy & metallurgical engineering