摘要
冷轧镀锌板比冷轧原板的使用寿命长,广泛用于汽车及家电制造、建筑材料等行业。在热镀锌工艺中,锌层厚度的控制是一个非常重要的生产环节。目前在国内多数镀锌产线中锌层厚度控制还需依赖人工控制,耗费大量精力的同时带来锌层厚度波动的质量风险。结合某钢厂镀锌线设备与工艺特点,建立了基于气刀循环神经网络的控制系统,将实际锌层与目标锌层的差值、气刀到带钢表面距离、气刀高度、稳定辊、带钢速度作为循环神经网络的输入,气刀压力作为循环神经网络的输出。利用循环神经网络的时序性特征实现锌层厚度的闭环控制。经实际生产数据验证,该系统在实现自动锌层厚度控制的同时降低锌耗,该算法可以较为广泛的适用于不同设备特征的气刀控制模型设计过程,具有较大的应用前景。
The service life of cold-rolled galvanized sheet is longer than that of cold-rolled raw sheet,it is widely used in industries such as automotive and home appliance manufacturing,building materials,etc.In the hot-dip galvanizing process,the control of the zinc layer thickness is a very important production process.At present,in most domestic galvanizing production lines,the control of zinc layer thickness still relies on manual control,which consumes a lot of energy and brings quality risks of zinc layer thickness fluctuations.This article establishes a control system of air knife recurrent neural network,based on the equipment and process characteristics of a galvanized line in a steel plan.It takes such as the difference between the actual zinc layer and the target zinc layer,the distance from the air knife to the strip surface,the height of the air knife,the stability roll and the strip speed as the input of the recurrent neural network,and the air knife pressure is used as the output of the network.The temporal characteristics of the recurrent neural network are used to realize the closed-loop control of zinc layer thickness.Actual production data shows that this system can realize automatic zinc layer thickness control while reducing zinc consumption.This algorithm can be widely applied to the design process of air knife control models with different equipment characteristics,and has great application prospects.
作者
杨凯
代龙飞
王枫
谢谦
YANG Kai;DAI Longfei;WANG Feng;XIE Qian(Anhui University of Technology,School of Metallurgical Engineering,Ma’anshan 243032,China)
出处
《重型机械》
2024年第2期35-39,共5页
Heavy Machinery
基金
国家自然科学基金青年基金(52104366)
安徽省工业智联网智能应用与安全工程实验室开放基金(IASII21-03)。
关键词
热镀锌
冷轧
气刀
循环神经网络
自动控制
hot-dip galvanizing
cold rolling
air knives
recurrent neural networks
automatic control
作者简介
杨凯(1999-),男,硕士研究生,主要研究方向为冷轧设备过程控制系统开发;通信作者:谢谦(1986-),男,博士,副教授,主要研究方向为冷轧设备过程控制系统开发。
