Statistical modeling of copper losses in the silicate slag of the sulfide concentrate smelting process

This article presents the results of the statistical modeling of copper losses in the silicate slag of the sulfide concentrates smelting process. The aim of this study was to define the correlation dependence of the degree of copper losses in the silicate slag on the following parameters of technological processes: SiO₂, FeO, Fe₃O₄, CaO and Al₂O₃ content in the slag and copper content in the matte. Multiple linear regression analysis (MLRA), artificial neural networks (ANNs) and adaptive network based fuzzy inference system (ANFIS) were used as tools for mathematical analysis of the indicated problem. The best correlation coefficient (R² = 0.719) of the final model was obtained using the ANFIS modeling approach.