Development of an Intelligent Supervisor for the Steel Melting Process Control System

The aim of the work is to obtain recommendations for the selection of the chemical composition of the raw material mass, ensuring a reduction in the defects of high-strength steel smelted in an open-hearth furnace, as well as the creation of an intelligent supervisor of an automated steelmaking process control system. Machine learning models are built on real data from a metallurgical production enterprise. Using the method of mathematical modeling, graphs of the dependencies of the impact toughness of castings on the most significant parameters of the chemical composition of the raw material mass are constructed. Based on the modeling results, recommendations are proposed to increase the impact toughness of the metal to an acceptable level.