摘要
热轧钢板的头尾形状蕴藏了大量钢板的轧制信息。宝钢5 m厚板在通过图像边缘识别技术获取钢板的平面形状数据后,尝试对钢板头尾形状的二维图像数据进行进一步的加工,提取了多种一维的特征数据,实现了对头尾形状的特征描述,并通过厚板PIDAS大数据平台建立了头尾形状特征数据库,为产品质量提升、工序生产优化提供生产数据支撑。头尾形状特征应用于剪切指导、组坯设计优化、PVPC控制参数优化等多个应用场景,充分发挥了其数据价值。
The plane shape of the head and tail of the hot-rolled heavy plates contains a large amount of rolling information.After obtaining the plane shape data of the plates through image edge recognition technology for Baosteel Heavy Plate Mill,attempted to extract a variety of one-dimensional feature data from the two-dimensional plane shape data of the head and tail of the plates and established database for these shape features in PIDAS,a big data platform for Baosteel Heavy Plate Mill,to provide actual result data for product quality improvement and process optimization.The head and tail shape features are used in multiple application scenarios such as cutting guidance,yield design optimization,PVPC control optimization,etc.
作者
贾俊彪
孔伟
刘晔
JIA Junbiao;KONG Wei;LIU Ye(Research Institute,Baoshan Iron & Steel Co. ,Ltd. , Shanghai 201999, China)
出处
《宝钢技术》
CAS
2021年第1期29-33,共5页
Baosteel Technology