基于HDLMD和JRD距离的电机轴承故障信号分解及性能评估  

Bearing Fault Signal Decomposition and Performance Evaluation Based on HDLMD and JRD Distance

作　　者：刘超 LIU Chao(Liyuan Hospital Affiliated to Tongji Medical College,Huazhong University of Science and Technology,Wuhan 430077,China)

机构地区：[1]华中科技大学同济医学院附属梨园医院,武汉430077

出　　处：《自动化与仪表》2023年第7期95-99,共5页Automation & Instrumentation

摘　　要：为了提高电机轴承运行寿命评估能力,采用HDLMD与JRD距离分析方法相结合的方式,设计得到了一种轴承性能测试方案,构建形成了更完善的DLMD理论分析系统。通过Renyi熵准确评价振动信号处于各个退化状态的复杂性,完成轴承性能退化的分析功能。研究结果表明:轴承振动信号在各故障状态下形成的时域特征也存在较大差异。HDLMD对样本信号进行分解后形成了5种尺度对应的PF分量。NASA全寿命周期波形表明此时信号受到强烈冲击,推断轴承存在故障问题。HDLMD分解得到JRD距离具备更小波动程度。采用本文设计HDLMD与JRD相结合的方法可以实现状态的精确识别。In order to improve the operating life evaluation ability of motor bearings,a bearing performance test scheme was designed by combining HDLMD and JRD distance analysis method,and a more perfect DLMD theoretical analysis system was constructed.Renyi entropy was used to accurately evaluate the complexity of vibration signals in each degradation state,and the analysis function of bearing performance degradation was completed.The results show that the time-domain characteristics of bearing vibration signals formed in different fault states are also quite different.After the sample signal is decomposed by HDLMD,PF components corresponding to five scales are formed.NASA life-cycle waveform indicates that the signal is strongly impacted at this time,and it is inferred that the bearing has a fault problem.HDLMD decomposition obtained JRD distance has a smaller degree of fluctuation.The method of combining HDLMD and JRD designed in this paper can realize accurate state recognition.

关 键 词：轴承性能评估 微分局部均值分解 拉普拉斯分值 RENYI熵 JRD距离 

分 类 号：TH137[机械工程—机械制造及自动化]