连铸结晶器粘结性漏钢预报系统研究与设计  

Research and Design of Sticker Breakout Prediction System in Continuous Casting Process

作　　者：周小凤[1] 肖俊生[2] 王志春[2] ZHOU Xiaofeng;XIAO Junsheng;WANG Zhichun(Department of Electrical Engineering, Baotou Vocational Technical College, Baotou 014035, China;School ofInformation Engineering, Inner Mongolia University of Science & Technology, Baotou 014010, China)

机构地区：[1]包头职业技术学院电气工程系,内蒙古包头014035 [2]内蒙古科技大学信息工程学院,内蒙古包头014010

出　　处：《铸造技术》2019年第5期496-499,共4页Foundry Technology

基　　金：国家自然科学基金资助项目(61463041)

摘　　要：介绍了粘结性漏钢的形成过程,对比分析了正常工况和粘结漏钢形成过程中结晶器壁的温度变化特征。通过BP神经网络建立了漏钢预报温度识别模型,用某钢厂200组典型历史温度数据对其进行训练;采用虚拟仪器平台搭建了漏钢预报实验系统并进行了模拟实验。结果表明,该方法预报实时、准确,具有一定的应用价值。The forming process of sticking breakout was introduced, and the temperature variation characteristics of mold wall in normal working condition and bonding process were compared and analyzed. BP neural network was used to establish the temperature identification model for molten steel leakage prediction. 200 groups of typical historical temperature data of a steel plant were used for training. The simulation experiment was carried out by using the virtual instrument platform to build the molten steel leakage prediction experiment system. The results show that the method is practical, accurate and certain application value.

关 键 词：连铸 粘结性漏钢 漏钢预测 BP神经网络 

分 类 号：TF777.1[冶金工程—钢铁冶金]