基于FWECS-CYCBD的轴承故障特征提取研究
Research on Bearing Fault Feature Extraction Based on FWECS-CYCBD
褚惟 1刘韬 1刘畅1
作者信息
- 1. 昆明理工大学机电工程学院 昆明,650500
- 折叠
摘要
针对最大二阶循环平稳盲解卷积(maximum second-order cyclostationary blind deconvolution,简称CYCBD)特征提取中循环频率和滤波带宽难确定的问题,引入频率加权能量相关谱(frequency weighted energy correlation spectrum,简称FWECS)来改进CYCBD,实现了低信噪比条件下的滚动轴承故障特征提取.首先,通过FWECS获取周期冲击频率,构造循环频率集;其次,利用最大加权谐波显著性指标设计了一种等步长搜索策略,自适应选取滤波器长度;最后,基于优选的循环频率和滤波带宽进行CYCBD解卷积.轴承仿真和实验数据表明:在循环频率等先验信息未知的情况下,FWECS-CYCBD对故障信号中的微弱冲击特征更敏感;与最小熵解卷积、改进最大相关峭度解卷积和自适应最大二阶循环平稳盲解卷积等方法相比,所提方法在低信噪比条件下能较好地提取轴承故障特征频率信息.
Abstract
The cycle frequency and filter bandwidth are difficult to determine in the maximum second-order cy-clostationary blind deconvolution(CYCBD)feature extraction.In this study,the frequency weighted energy correlation spectrum(FWECS)is introduced to improve the CYCBD and achieves the bearing fault feature ex-traction under low signal-to-noise ratio conditions.This method firstly obtains the periodic impact frequency by FWECS and constructs the cyclic frequency set.Secondly,an equal-step search strategy is designed to adap-tively select the filter length using the maximum weighted harmonic significant index.Finally,the CYCBD is performed based on the optimized cyclic frequency and filter bandwidth.Bearing simulation and experimental data verification show that FWECS-CYCBD is more sensitive to the weak impact features in the fault signal un-der the circumstance that the priori information such as the cyclic frequency is unknown.Compared with meth-ods such as minimum entropy deconvolution,improved maximum correlation kurtosis deconvolution and adap-tive maximum second-order cyclostationary blind deconvolution,the proposed method is able to extract the fre-quency information of bearing fault features under low signal-to-noise ratio conditions.
关键词
滚动轴承/故障诊断/特征提取/最大二阶循环平稳盲解卷积/频率加权能量相关谱/加权谐波显著性指数Key words
rolling bearing/fault diagnosis/feature extraction/the maximum second-order cyclostationary blind deconvolution(CYCBD)/frequency weighted energy correlation spectrum(FWECS)/weighted harmonic saliency index引用本文复制引用
基金项目
云南省科技厅重大科技专项资助项目(202102AC080002)
国家自然科学基金资助项目(52065030)
出版年
2024
NETL
NSTL
万方数据