基于灰色关联分析和超限学习机的高炉铁水硅含量预测  

Prediction of silicon content in hot metal of blast furnace based on grey correlation analysis and extreme learning machine

作　　者：崔博 陈伟 王宝祥 武鹏飞 陈颖 CUI Bo;CHEN Wei;WANG Baoxiang;WU Pengfei;CHEN Ying(College of Artificial Intelligence,North China University of Science and Technology,Tangshan 063210,China;College of Metallurgy and Energy,North China University of Science and Technology,Tangshan 063210,China;Hebei Province High Quality Steel Continuous Casting Engineering Technology Research Center,Tangshan 063000,China)

机构地区：[1]华北理工大学人工智能学院,河北唐山063210 [2]华北理工大学冶金与能源学院,河北唐山063210 [3]河北省高品质钢连铸工程技术研究中心,河北唐山063000

出　　处：《冶金自动化》2022年第1期54-62,共9页Metallurgical Industry Automation

基　　金：河北省自然科学基金资助项目(E2012209025)。

摘　　要：炉温的实时预测技术对高炉生产稳定顺行具有重要意义,在高炉炼铁过程中,通常间接用铁水硅含量的变化来表示高炉炉温的变化。针对硅含量预测效率和精度不足的问题,建立了铁水硅含量预测模型。以现场数据为样本数据,采用灰色关联分析(grey correlation analysis, GCA)获得与硅含量相关度较高的生产指标,以相关指标为输入、硅含量为输出,构建超限学习机(extreme learning machine, ELM)算法模型,对模型进行训练。现场数据计算表明,该模型的预报命中率达87%(误差不小于0.10),实现了高炉铁水硅含量的准确预报。The real-time prediction technology of furnace temperature is of great significance to the stable and smooth operation of blast furnace production.In the process of blast furnace ironmaking, the change of blast furnace temperature is usually expressed indirectly by the change of hot metal silicon content.Aiming at the problem of insufficient efficiency and accuracy of silicon content prediction, a hot metal silicon content prediction model is established.Taking the field data as the sample data, the production index with high correlation with silicon content is obtained by grey correlation analysis(GCA).Taking the relevant index as the input and silicon content as the output, the extreme learning machine(ELM) algorithm model is constructed to train the model.The field data calculation shows that the prediction hit rate of the model is 87%(the error is not more than 0.10) and the accurate prediction of silicon content in hot metal of blast furnace is realized.

关 键 词：高炉 铁水硅含量 算法模型 灰色关联分析 超限学习机 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] TF53[自动化与计算机技术—控制科学与工程]