改进EMD算法在滚动轴承振动信号分析中的应用

Application of Improved EMD Algorithm to Vibration Signal Analysis of Rolling Bearing

针对传统的经验模态分解处理非平稳、非线性振动信号存在的模态混淆问题,提出了一种改良的经验模态分解处理滚动轴承的振动信号。针对采集振动信号过程中存在多频率信号、噪声和间歇信号的特点,构建了高低频率信号和带有噪声及间歇信号的混合信号,并对仿真信号进行传统的经验模态分解和改进经验模态分解。结果表明,使用改进经验模态分解对实际采集的轴承振动信号进行分析,能够较好地改善模态混淆现象。

In view of the modal confusion of non-stationary and nonlinear vibration signals treated by traditional empirical mode decomposition(EMD),an improved empirical mode decomposition(EMD)is proposed to deal with the vibration signals of rolling bearings.Firstly,according to the characteristics of multi-frequency signal,noise and intermittent signal which appear in the process of collecting vibration signal,the high-low frequency signal and the mixed signal with the noise and intermittent signals are constructed,and then the comparison between the traditional empirical mode decomposition and the improved empirical mode decomposition of the simulation signal is done.Finally,the analysis on the actual bearing vibration signal is performed by the means of the improved empirical mode decomposition and therefore the modal confusion phenomenon gets improved.