基于BP神经网络的炉辊联轴器健康评估  

Health Assessment of Furnace Roller Coupling Based on BP Neural Network

作　　者：丁保宇 牛锐祥 Ding Baoyu;Niu Ruixiang(Silicon Steel Business Unit of Shanxi Taigang Stainless Steel Co.,Ltd.,Taiyuan Shanxi 030003,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030003

出　　处：《山西冶金》2024年第3期50-52,共3页Shanxi Metallurgy

摘　　要：针对炉辊联轴器故障诊断难的问题,通过分析某连续退火机组传动部多个测点的振动数据,提出了基于BP神经网络的炉辊联轴器故障诊断方法。该方法以一测点振动数据为输入,以健康标签三分类为输出,可准确识别联轴器正常、摩擦和故障的三种健康状态,具有较高的工程实用性。In response to the problem of difficult fault diagnosis for furnace roller couplings,a fault diagnosis method for furnace roller couplings based on BP neural network is proposed by analyzing the vibration data of multiple measurement points in the transmission part of a continuous annealing unit.This method takes vibration data from a measuring point as input and health label three classifications as output,which can accurately identify the three health states of coupling:normal,friction,and fault,and has high engineering practicality.

关 键 词：BP神经网络 联轴器 健康评估 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]