“人工+智能”复合回归管理方法在高炉生产的应用  

Application of"Artificial+Intelligence"Compound Regression Management Method in Blast Furnace Production

作　　者：赵丽华 王志刚 ZHAO Lihua;WANG Zhigang(Laiwu Branch of Shandong Iron and Steel Co.,Ltd.,Laiwu 271104,China)

机构地区：[1]山东钢铁股份有限公司莱芜分公司,山东莱芜271104

出　　处：《山东冶金》2024年第4期54-56,60,共4页Shandong Metallurgy

摘　　要：通过归纳一线人工经验,确定关键参数和权重系数,形成“人工+智能”复合最佳回归公式;通过电脑编辑智能程序,实时采集各关键参数,精准预测绩效指标,进而指导实际生产管理。模型铁水硅成分预测准确度90%以上,改进后铁水平均硅降低约0.04%,实现了高炉生产高效运行,经济效益显著。By summarizing the first-line manual experience,determine key parameters and weight coefficients,form the"artificial+intelligence"compound optimal regression formula;by editing smart programs on computers,real-time collection of key parameters,accurately predict performance indicators,and then guide the actual production management.The model can predict the silica content of blast furnace melted iron with more than 90%accuracy.After the model is implemented,the average silica content of blast furnace melted iron is reduced by about 0.04%,and the high efficiency operation of blast furnace is realized with remarkable economic benefits.

关 键 词：高炉铁水 硅含量 复合回归 关键参数 预测准确度 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]