Prediction of the Degree of Metallization of Hot Briquetted Iron

The development of a predictive model for the degree of metallization of hot briquetted iron is one of the urgent tasks of metallurgical science. The article discusses global and domestic experience in analyzing data on the operation of HBI plants and the possibility of predicting the quality of manufactured products. It is noted that the Midrex iron ore recovery technology has received the most widespread use and study in world practice. The authors of the article propose an algorithm for predicting the degree of metallization of a restored product based on clustering input raw materials according to their metallurgical and chemical properties and constructing a regression model based on a set of technological parameters of the Midrex process, which have a direct impact on product quality. The article provides a comparative analysis of regression analysis methods using machine learning, as a result of which random forest was selected as the most optimal method in terms of the root mean square error of prediction (RMSE). The developed algorithm can be used in production to solve problems of operational control of the iron recovery process in order to improve the quality of manufactured products.