摘要
针对数控磨床主轴振动信号中噪声影响特征值提取的问题,提出了一种基于经验模态分解方法(EMD)与快速傅里叶变换(FFT)联合的滤波方案。首先利用经验模态法(EMD)将主轴振动信号分解成若干个满足内稟模态函数的IMF分量,再利用快速傅里叶变换对每个IMF分量做时频转换,可以得出IMF分量的频域图谱。对包含噪声的IMF分量进行阈值滤波,得到降噪后的IMF分量,将IMF分量重新合成为原信号。利用仿真信号对该方法实验,同时对数控机床主轴信号进行实际测试表明,该方法不仅仅能够很好地分解出主要频率成分,同时弥补了FFT变换无法分析非平稳信号的缺点,提高了EMD算法的精度,有效地滤除主轴信号中的噪声成分。
Aiming at extracting the characteristic value of noise in the vibration signal of the spindle of CNC grinding machine, a filtering scheme based on the combination of EMD and FFT was proposed. Firstly, the principal axis vibration signal is decomposed into several IMF components satisfying the intrinsic mode function by empirical mode method (EMD), and then the time-frequency conversion is performed on each IMF component by using the fast Fourier transform to obtain the frequency domain maps of IMF component. Threshold filtering is performed on the IMF component that contains noise, and the denoised IMF component is obtained, and the IMF component is recombined into the original signal. Using the simulation signal to test the method, the actual test on the spindle signal of the CNC machine tool shows that the method can not only decompose the main frequency components well, but also make up for the shortcomings of the FFT transform that can not analyze the non-stationary signal, and improve the EMD algorithm. The noise component in the filtered spindle signal is effectively eliminated.
作者
甘屹
李平阳
孙福佳
Gan Yi;Li Pingyang;Sun Fujia(School of Mechanical Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)
出处
《农业装备与车辆工程》
2019年第8期82-85,90,共5页
Agricultural Equipment & Vehicle Engineering
关键词
EMD
FFT
IMF分量
阈值滤波
empirical mode decomposition
fast fourier transform
IMF component
threshold filter
