摘要
诊断滚动轴承变工况故障时通常采用硬件或信号预处理方法处理特征频率模糊或特征参数扰动问题,诊断环节复杂且不易实施,因而提出直接获取工况不敏感特征参数的诊断思路。首先推导滚动轴承振动信号的双谱表达式,发现工况参数仅影响双谱幅值,双谱分布具有工况鲁棒优点。再通过仿真和实测信号探究双谱对滚动轴承振动特性表征,并用不同工况、不同故障状态下的双谱验证了理论分析结果。最后进行变工况故障诊断实验,3种故障程度下的5折交叉验证结果表明,使用任意一种工况数据训练模型都能有效辩识4种不同工况故障类型。工况不敏感特征参数诊断方法具有简单易实施优点,无额外的变工况处理环节,在工程实践中具有良好的应用前景。
In the fault diagnosis of rolling element bearings under variable operation conditions,particular hardware or signal reprocessing is necessary to overcome the frequency ambiguity or feature inconsistency.It results in the diagnosis complicated and hard to implement.Therefore,this paper proposes a novel diagnosis approach that directly extracts the features insensitive parameters in the operation condition.Firstly,the bi-spectrum expression of the vibration signals of the rolling bearings is formulated.The result shows that the condition parameter only affects the bi-spectrum magnitude,and the bi-spectrum distribution is robust to the condition variation.Next,through simulation and signal measurement,the vibration characteristic of the rolling bearing element expressed by the bi-spectrum is explored,and the result of theoretical analysis is verified by the bi-spectrum under different working conditions and fault status.Finally,an experiment for fault diagnosis of variable operation condition is implemented to validate the proposed approach.The 5-fold cross-verification in three fault severities shows that the model in any operation conditions can effectively identify the fault types under four kinds of operation conditions.The proposed method does not need the additional signal processing steps,has the advantage of simplicity and ease of implementation,and has a good application prospect in engineering practice.
作者
张锐戈
肖荣辉
高忠坚
ZHANG Ruige;XIAO Ronghui;GAO Zhongjian(School of Mechanical and Electronic Engineering,Sanming University,Sanming 365004,Fujian,China;Key Laboratory of Intelligent Equipment of Fujian Higher Education Institutions,Sanming 365004,Fujian,China)
出处
《噪声与振动控制》
CSCD
北大核心
2021年第6期97-104,共8页
Noise and Vibration Control
基金
福建省教育厅自然科学基金重点资助项目(JZ160476)
三明市引导性科技计划资助项目(2019-G-7)。
关键词
故障诊断
滚动轴承
变工况
双谱
fault diagnosis
rolling element bearing
variable operation condition
bi-spectrum
