基于1DCNN的张力辊速度超差检测  

Detection of Speed Deviation of Tension Roller Based on 1DCNN

作　　者：刘真甫 牛锐祥 Liu Zhenfu;Niu Ruixiang(Silicon Steel Business Unit of Shanxi Taigang Stainless Steel Co.,Ltd.,Taiyuan Shanxi 030002,China)

机构地区：[1]山西太钢不锈钢股份有限公司硅钢事业部,山西太原030002

出　　处：《山西冶金》2024年第5期22-24,共3页Shanxi Metallurgy

摘　　要：速度超差是冷轧连续退火机组张力辊组最常见的问题,甚至造成诸多废降次产品,影响产品质量管控。以某连续退火机组多次发生速度超差异常的出口段张力辊为例,为提高张力辊速度超差检测效率,提出基于一维卷积神经网络(1DCNN)的张力辊速度超差故障检测方法,实验验证表明,该方法具有较高的检测效率,可准确判断速度超差发生的时间和位置,为张力辊的周期性维护计划提供指导,同时为张力辊设备的健康管理提供新方法。Speed deviation is the most common problem in the tension roller group of cold rolling continuous annealing units,and even causes many scrap times,affecting product quality control.Taking the exit section tension roller of a continuous annealing unit that has experienced speed deviation anomalies multiple times as an example,in order to improve the efficiency of tension roller speed deviation detection,a speed deviation fault detection method based on one-dimensional convolutional neural network(1DCNN)is proposed.The experiment shows that this method has high detection efficiency.It can accurately determine the time and location of speed deviation,provide guidance for the periodic maintenance plan of tension rollers,and provide new methods for the health management of tension roller equipment.

关 键 词：1DCNN 张力辊 速度超差 故障检测 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置] TG333.17[自动化与计算机技术—控制科学与工程]