摘要
精准预测连铸坯裂纹、中心偏析等缺陷并在连铸坯下线清理与热装热送间做出决策对于稳定连铸生产、提高连铸坯质量具有重要意义。然而,实际生产中影响连铸坯质量的因素众多,连铸生产存在着不可预见的扰动性,且生产参数间具有较强的非线性和耦合性,这使得连铸坯裂纹、中心偏析等缺陷的精准预测极具挑战。随着连铸自动化和计算机技术的不断发展,人工智能逐渐得到重视,其中机器学习因其强大的非线性逼近能力而逐步应用于连铸生产。本文着重从机器学习、专家系统方面总结了国内外在连铸坯质量预测、诊断方面的研究进展,分析比较了各类方法的优缺点,并对连铸坯质量预测进行了展望。
Accurately predicting the defects such as cracks and center segregation of continuous casting strand and making a choice between offline cleaning and hot delivery are of great significance to stabilize the continuous casting production and improve the production quality.However,there are many factors affecting the quality of continuous casting strand in actual production.There are unpredictable disturbances in continuous casting production,and there is strong nonlinearity and coupling between production parameters,which makes the accurate prediction of the defects such as crack and central segregation of continuous casting strand very challenging.With the development of the continuous casting automation and computer technology,artificial intelligence has been paid more and more attention,among which machine learning has been gradually applied in the continuous casting production because of its strong nonlinear approximation ability.The research progress of the strand quality prediction at home and abroad are summarized from the aspects of the machine learning and expert system,and the advantages and disadvantages of various methods are analyzed and compared.Meanwhile,the quality prediction of continuous casting strand is prospected.
作者
邹雷雷
黄俊雄
李权辉
张江山
刘青
ZOU Lei-lei;HUANG Jun-xiong;LI Quan-hui;ZHANG Jiang-shan;LIU Qing(State Key Laboratory of Advanced Metallurgy,University of Science and Technology Beijing,Beijing 100083,China;Research Institute,Nanjing Iron and Steel Co.,Ltd.,Nanjing 210035,Jiangsu,China)
出处
《连铸》
2022年第2期2-9,共8页
Continuous Casting
基金
国家自然科学基金资助项目(U21A20112)
钢铁冶金新技术国家重点实验室自主课题资助项目(41619004)。
关键词
连铸
铸坯裂纹
中心偏析
质量预测
机器学习
专家系统
continuous casting
billet crack
center segregation
quality prediction
machine learning
expert system
作者简介
邹雷雷(1984-),男,博士生,E-mail:zoulei153@163.com;通讯作者:刘青(1967-),男,博士,教授,E-mail:qliu@ustb.edu.cn。
