基于BP神经网络的炉辊健康评估  

Health Assessment of Furnace Roll Based on BP Neural Network

作　　者：牛锐祥 NIU Rui-xiang(Cold-Rolled Silicon Steel Plant,Shanxi Taigang Stainless Steel Co.,Ltd.,Taiyuan 030003,China)

机构地区：[1]山西太钢不锈钢股份有限公司冷轧硅钢厂,山西太原030003

出　　处：《机械工程与自动化》2024年第1期146-148,共3页Mechanical Engineering & Automation

摘　　要：炉辊是连续退火机组的重要设备,其健康状态直接影响带钢表面质量,有效评估炉辊的健康状态对于炉辊的健康维护计划具有重要意义。从某钢厂炉辊的历史数据出发,针对炉辊工作特性,提出了炉辊健康状态监测方案,设计了基于BP神经网络的炉辊健康评估方法,实现了对炉辊正常、轻微退化、严重退化和故障4个等级健康状态的识别,达到了指导炉辊有效健康评估和更换维护的目的,具有良好工程实用性。Furnace roll is an important equipment in continuous annealing unit,and its health status directly affects strip surface quality.Effectively evaluating the health status of furnace roll is of great significance for the health maintenance plan of furnace roll.Based on the historical data of the furnace roll in a steel mill and the working characteristics of the furnace roll,a scheme for monitoring the health status of furnace roller is put forward,and the health assessment method of the furnace roll based on BP neural network was designed,which realized the identification of the four levels of health status of the furnace roll:normal,slight degradation,serious degradation and failure.It can achieve the purpose of guiding the effective health assessment and replacement maintenance of the furnace roll.It has good engineering practicability.

关 键 词：BP神经网络 炉辊 健康评估 

分 类 号：TF307[冶金工程—冶金机械及自动化] TP391[自动化与计算机技术—计算机应用技术]