论文部分内容阅读
采用GA-BP神经网络模型对熔渣组元活度进行预测,通过对不同温度条件下不同组元渣系活度值的验证,证明了GA-BP渣系活度预测模型有较好的预测精度。在此基础上建立了奥氏体不锈钢、铁素体不锈钢冶炼过程中钢液脱氧热力学模型。热力学模型表明,钢液中铬质量分数越高,脱氧越困难;奥氏体不锈钢铝脱氧条件下,镍质量分数越高,脱氧能力越差;任何情况下降低熔渣中脱氧产物的活度都有利于降低平衡条件下钢液中溶解氧质量分数。
GA-BP neural network model was used to predict the activity of slag components. Through the validation of slag activity values of different components under different temperature conditions, it was proved that GA-BP slag activity prediction model has a good prediction Accuracy. Based on this, a thermodynamic model of deoxidation of molten steel during austenitic stainless steel and ferritic stainless steel smelting was established. The thermodynamic model shows that the higher the chromium content in the molten steel, the more difficult it is to deoxidize. The higher the nickel content, the worse the deoxidation ability under the austenitic stainless steel aluminum deoxidation condition and the lower the deoxidation product activity in the slag Help to reduce the equilibrium conditions of molten steel dissolved oxygen mass fraction.