摘要
相关熵为高斯、非高斯噪声处理的一种有效方法,针对强高斯噪声和非高斯噪声干扰下齿轮故障诊断问题,提出了一种基于相关熵和双谱的齿轮故障诊断方法。该方法综合利用高斯核函数和不完全Cholesky分解算法计算信号的相关熵,然后再计算相关熵的双谱,根据相关熵的双谱特征识别齿轮故障。通过不完全Cholesky分解算法计算信号的相关熵,不仅大大压缩了数据量,突出了齿轮故障特征,而且提高了计算效率。通过仿真和齿轮磨损故障振动信号分析结果表明:强背景噪声会造成传统双谱故障诊断方法失效,而基于相关熵和双谱分析的齿轮故障诊断方法,能在强噪声干扰背景中提取齿轮的故障特征准确识别齿轮故障,其性能优于传统双谱和小波变换域双谱,为一种有效的齿轮故障诊断方法。
Correntropy is an effective method to deal with Gaussian and non-Gaussian noise.Aiming at the problem of gear fault diagnosis under the interference of strong Gaussian noise and non-Gaussian noise,a gear fault diagnosis method based on correntropy and bi-spectrum is put forward.Gaussian kernel function and incomplete Cholesky decomposition algorithm are used to calculate the correntropy of the vibration signal,the bi-spectrum of the correntropy is calculated,and the gear fault is identified according to the bi-spectrum characteristics of the correntropy.The incomplete Cholesky decomposition based correntropy algorithm not only greatly reduces the amount of data,highlights the fault characteristics of the gear,but also improves the calculation efficiency.The results of simulation and vibration signal analysis of gear wear fault show that the strong background noise will cause the failure of the traditional bi-spectrum fault diagnosis method,while the gear fault diagnosis method based on correntropy and bi-spectrum analysis can extract the fault features of gear in the background of strong noise interference,accurately identify gear fault,and its performance is better than that of traditional bi-spectrum and wavelet transform domain bi-spectrum,which is an effective method for gear fault diagnosis.
作者
李辉
郝如江
LI Hui;HAO Ru-jiang(School of Mechanical Engineering,Tianjin University of Technology and Education,Tianjin 300222,China;School of Mechanical Engineering,Shijiazhuang Tiedao University,Shijiazhuang 050043,China)
出处
《振动工程学报》
EI
CSCD
北大核心
2021年第5期1076-1084,共9页
Journal of Vibration Engineering
基金
国家自然科学基金资助项目(51375319)
河北省自然科学基金资助项目(E2013421005)。
关键词
故障诊断
齿轮
信号处理
相关熵
双谱
fault diagnosis
gear
signal processing
correntropy
bi-spectrum
