OVMD-MPE群稀疏全变分去噪算法研究  

Reasearch on OVMD-MPE Group Sparsity Total Variational Denoising Algorithm

作　　者：陈维兴[1] 孙习习 CHEN Wei-xing;SUN Xi-xi(Civil Aviation University of China,Tianjin 300300,China)

机构地区：[1]中国民航大学,天津300300

出　　处：《计量学报》2022年第1期48-56,共9页Acta Metrologica Sinica

基　　金：国家自然科学基金民航联合研究基金(U1433107);中央高校基本科研业务中国民航大学专项基金(3122017041,3122018D009)。

摘　　要：轴承振动数据在采集过程中易受噪声干扰,无法有效突出微弱局部故障脉冲,从而影响轴承故障诊断效率。针对这一问题,提出了一种OVMD-MPE的群稀疏全变分去噪算法。首先,利用变分模态分解分解信号,再利用蚱蜢优化算法获得变分模态分解的最优参数;然后,计算各模态分量的经验模态分解,分离出噪声主导分量和有用分量;最后,通过群稀疏全变分去噪算法对噪声主导分量滤波,并将滤波后分量和有用分量合并重构去噪信号。实验结果表明:与传统的去噪方法相比,模拟重构信号的平均信噪比提高了约3.3 dB,轴承数据故障准确率提高至98.9%。Bearing vibration data is susceptible to noise interference during the acquisition process and cannot effectively highlight weak local fault pulses,thereby affecting the efficiency of bearing fault diagnosis.To solve this problem,an ovmd-mpe group sparse total variational denoising algorithm(OVMD-MPE-GSTVD)is proposed.Firstly,variational model decomposition is used to decompose the signal,and then the optimal parameters of variational model decomposition are obtained by grasshopper optimization algorithm.Then,calculate the empirical model decomposition of each modal component to separate the noise dominant component and the useful component.Finally,the dominant component of the noise is filtered by group sparse total variational denoising algorithm,and the filtered component and useful component are combined to reconstruct the denoising signal.The experimental results show that compared with the traditional denoising method,the average signal-to-noise ratio of the simulated reconstructed signal is improved by about 3.3 dB,the bearing data fault accuracy is increased to 98.9%.

关 键 词：计量学 故障诊断 滚动轴承 变分模态分解 多尺度排列熵 群稀疏全变分去噪 

分 类 号：TB936[一般工业技术—计量学] TB973[机械工程—测试计量技术及仪器]