摘要
针对未知分布数据序列稳健问题,提出改进Huber M稳健化处理方法处理滚动轴承振动数据。基于Huber M估计法,融合了中位数估计的优点,根据时间变量t将实验数据进行离散,构建数学模型,得到一种改进的Huber M方法。应用该方法对圆锥滚子轴承振动数据分析,确定稳健数据的边界值及显著水平。研究结果表明:在0~0.1显著性水平范围内,经过改进Huber M稳健化处理方法的滚动轴承振动数据稳健指标、离散指标均优于原数据指标,有效提高了滚动轴承数据的稳健性。
Aiming to robust problems of unknown distribution data series,the improved Huber M method was proposed to analyze vibration data of rolling bearings.Based on the Huber M estimation method,which was combined with the advantages of median estimation,the experimental data was discretized according to the time variable t.The mathematical model was constructed,and an Huber M method was improved.This method was applied to analyze the vibration data of tapered roller bearing to determine the boundary value and significance level of robust data.Results show that the robust index and discrete index of rolling bearing vibration data with improved Huber M robust processing method are better than the original data index in the range of 0~0.1 significance level.Therefore the method increases robust and discrete characteristics of rolling bearings.
作者
党凤魁
徐永智
丁慧玲
刘建
DANG Fengkui;XU Yongzhi;DING Huiling;LIU Jian(Agricultural Equipment Engineering School,Henan University of science&Technology,Luoyang 471023,China;National United Engineering Laboratory for Advanced Bearing Tribology,Henan University of science&Technology,Luoyang 471023,China;Sanmenxia Polytechnic,Sanmenxia 472000,China)
出处
《河南科技大学学报（自然科学版）》
CAS
北大核心
2022年第4期12-18,M0003,共8页
Journal of Henan University of Science And Technology:Natural Science
基金
国家重点研发计划基金项目(2018YFB2000300)。
关键词
滚动轴承
稳健化处理
乏信息
未知分布
显著水平
rolling bearing
robust processing
poor information
unknown distribution
significance level
