摘要
针对齿轮箱故障信号微弱且易受周围噪声影响的问题,提出了一种基于变分模态分解(VMD)的独立分量(ICA)算法。该方法首先将采集的信号进行MCKD降噪,将降噪后的信号利用VMD算法分解为多个不同的本征模态分量(IMF),然后依据快速谱峭度图和相关系数选取有效的IMF分量进行重构信号,对于重构信号利用FastICA再次进行降噪处理,根据FastICA降噪后得到的故障特征分量,可以有效地识别故障。结果表明:该方法可以更清晰、准确地提取出故障特征频率和找出故障发生的位置。
Aiming at the problem that the gearbox fault signal is weak and susceptible to the surrounding noise,a fault diagnosis method based on the combination of Variational Modal Decomposition(VMD)and Independent Component Analysis(ICA)is proposed.First,the MCKD is used to reduce the noise of the collected signal,and the signal after noise reduction is decomposed into several different intrinsic modal components(IMFs)by using VMD method.Then,according to the fast spectral cliff graph and correlation coefficient,the effective IMF components are selected to reconstruct the signal,and the FastICA is used to reduce the noise again for the reconstructed signal.Depend on the fault characteristic component obtained after the noise reduction of FastICA,the fault can be identified effectively.The results show that this method can extract the fault characteristic frequency more clearly and accurately and find out where the fault occurs.
作者
吴鲁明
郝如江
陆一鹤
Wu Luming;Hao Rujiang;Lu Yihe(School of Mechanical Engineering,Shijiazhuang Tiedao University,Shijiazhuang 050043,China)
出处
《石家庄铁道大学学报（自然科学版）》
2020年第3期14-20,共7页
Journal of Shijiazhuang Tiedao University(Natural Science Edition)
基金
国家自然科学基金(51375319)。
