摘要
滚动轴承故障信号常包含着大量的噪声,并以调制的形式存在,其故障特征信息提取困难;同时,采用快速经验小波变换(FEWT)分解故障信号时,又存在故障特征被削弱的问题。为此,将FEWT与快速独立分量分析(FastICA)的优点相结合,在此基础上提出了一种基于FEWT-FastICA的滚动轴承故障特征识别方法。首先,利用FEWT算法对轴承故障信号进行了分解,得到了一组固有模态分量(IMF);根据峭度准则,将峭度值大于3的IMF分量重构为振动冲击信号,峭度值小于3的IMF分量重构为虚拟通道信号;然后,将重构后的信号输入FastICA算法,进行信号的降噪解混,得到信号的最佳估计信号,对最佳估计信号进行了包络谱分析,完成了对滚动轴承的故障诊断;最后,为了验证FEWT-FastICA算法的有效性,采用仿真信号及真实轴承故障信号分别进行了实验验证;同时,为了验证FEWT-FastICA算法的优越性,将其与FEWT进行了对比分析。研究结果表明:该方法能有效地提取故障特征信息,比FEWT方法所得结果的信噪比提升了1.55倍,为轴承故障诊断提供了一种新方法。
Rolling bearing fault signals often contain a lot of noise and exist in the form of modulation,so it is difficult to extract fault feature information.At the same time,when fast empirical wavelet transform(FEWT)is used to decompose the fault signal,the fault feature is weakened.Therefore,combining with the advantages of FEWT and fast independent component analysis(FastICA),a fault feature recognition method of rolling bearing based on FEWT-FastICA was proposed.Firstly,the FEWT algorithm was used to decompose the bearing fault signals to obtain a set of intrinsic modal components(IMF).According to the kurtosis criterion,IMF components with kurtosis greater than 3 were reconstructed as vibration impact signals,while IMF components with kurtosis less than 3 were reconstructed as virtual channel signals.Then,the reconstructed signal was denoised and unmixed by FastICA algorithm,and the best estimated signal was obtained.The envelope spectrum analysis of the best estimated signal was carried out to complete the fault diagnosis and analysis of rolling bearing.Finally,in order to verify the effectiveness of FEWT-FastICA algorithm,simulation signals and real bearing fault signals were used for experimental verification.At the same time,in order to verify the superiority of FEWT-FastICA algorithm,it was compared with FEWT.The research results show that the FEWT-FastICA method can effectively extract fault feature information,and the signal-to-noise ratio of the results obtained by FEWT method is 1.55 times higher than that obtained by FEWT method,which provides a new method for bearing fault diagnosis.
作者
黄致远
颜丙生
刘兆亮
HUANG Zhi-yuan;YAN Bing-sheng;LIU Zhao-liang(School of Mechanical and Electrical Engineering,Henan University of Technology,Zhengzhou 450001,China)
出处
《机电工程》
CAS
北大核心
2023年第4期509-515,共7页
Journal of Mechanical & Electrical Engineering
基金
国家自然科学基金资助项目(U1604134)
河南省科技攻关项目(212102210327)
河南工业大学创新基金支持计划专项资助项目(2020ZKCJ28)。
关键词
轴承故障诊断
快速经验小波变换
快速独立分量分析
降噪解混
故障特征提取
信噪比
bearing fault diagnosis
fast empirical wavelet transform(FEWT)
fast independent component analysis(FastICA)
denoised and unmixed
fault feature extract
signal-to-noise ratio
