The optimization of process variables in electromagnetic metallurgy and continuous casting technology is an important means for enterprises to improve the quality of continuous casting billets.However,the explanation of the mechanism of continuous casting process and the adjustment methods of process variables by enterprises are seriously lagging behind production.Therefore,this article proposes the integration of multiple technologies such as numerical simulation,computer vision,and big data mining to achieve visualization of multiple physical fields in the continuous casting process,automatic identification and intelligent detection of continuous casting billet quality,and coupling optimization of electromagnetic metallurgy technology and continuous casting process variables.The electromagnetic metallurgy digital system has the preliminary ability to visualize multiple physical fields such as flow field,temperature field,liquid phase volume fraction,solute field,inclusion distribution,and surface stress of solidified billet shell in the crystallizer,as well as to digitally identify central defects in the casting billet.Among them,the recognition effect of the Unet model on central defects can reach over 90％,and the use of binarization can effectively evaluate the level of central looseness.It can be predicted that the digital system of electromagnetic metallurgy is of great significance for promoting the intelligent development of steel enterprises.
关键词
大数据/人工智能/电磁冶金/物理场可视化/数据挖掘/智能检测/数字化系统
Key words
big data/artificial intelligence/electromagnetic metallurgy/physical field visualization/data mining/intelligent detection/digital system
维普
万方数据