Composition Engineering on the Local Structure and Viscosity of the CaO-SiO2-Al2O3-P2O5-FeO Slag by Machine Learning Methods

Due to the high cost and low accuracy of high-temperature tests, viscosity data for multicomponent slag systems is difficult be obtained precisely. Therefore, it important fulfill database via reasonable methods with lower costs. In this study, a prediction method based on machine learning was proposed CaO-SiO2-FeO-Al2O3-P2O5 quinary system. To provide valid model, predicted by molecular dynamic multiple semi-empirical models were compared verify applicability these Different also developed. The results showed that from gradient boosting decision tree most accurate Based method, color-map concerning numerical effect Al2O3 P2O5 contents provided, which provides assistance composition engineering certain demand design.