基于多元回归法的酸洗线和镀锌线用轴承维修策略  

Maintenance Strategy for Pickling Line and Galvanizing Line Bearings Based on Multiple Regression Method

作　　者：王桂梅[1] 李艳强[1] 李永利 宋辉 

机构地区：[1]河北工程大学机电学院,河北邯郸056038 [2]务腾咨询(上海)有限公司北京分公司,北京100025 [3]河北钢铁集团,河北邯郸056038

出　　处：《轴承》2014年第10期48-50,共3页Bearing

基　　金：河北省自然科学基金项目(E2012402027)

摘　　要：针对传统维修导致设备"带病工作"和"无病诊断"的现象,以视情维修为基础,设计了一种基于多元回归法的轴承温度预测方法,该方法能够对采集的数据进行合理筛选,有效提高模型的预测精度,减少因轴承温度偏高而产生的维修费用,有效降低维修成本和设备故障率。Aiming at the phenomenon about “work with fault”and “diagnosis without fault”of equipment caused by traditional maintenance,based on on-condition maintenance,a temperature predicting method for bearings is designed based on multivariate regression method.The method is able to select collected data reasonably,effectively improve the prediction accuracy of the model,reduce maintenance costs due to the higher bearing temperature and effectively reduce maintenance costs and equipment failure rate.

关 键 词：滚动轴承 多元回归法 温度 预测模型 

分 类 号：TH133.33[机械工程—机械制造及自动化] TH17