SGMD-MOMEDA滚动轴承故障特征提取方法研究

RESEARCH ON ROLLING BEARING FAULT FEATURE EXTRACTION METHOD WITH SGMD-MOMEDA

针对滚动轴承的振动信号因非线性、非平稳且信噪比低而造成故障特征难以提取的问题,基于辛几何模态分解(Symplectic Geometry Mode Decomposition, SGMD)和多点最优最小熵解卷积调整(Multipoint Optimal Minimum Entropy Deconvolution Adjusted, MOMEDA)理论,提出了SGMD-MOMEDA故障提取方法。首先,使用SGMD对故障信号进行分解,得到一列的辛几何分量(Symplectic Geometry Components, SGC);其次,依据相关性准则选取SGC进行信号重构,并确定MOMEDA分解参数;最后,使用MOMEDA方法对重构信号进行处理以提高信噪比,并利用包络谱分析对处理后的信号提取故障特征。仿真和实验结果表明,该方法能够准确地提取滚动轴承的故障频率,且与经验模态分解(Empirical Mode Decomposition, EMD)方法的对比结果显示了SGMD方法作为预处理其分解结果更加准确,在故障诊断领域具有较大的应用价值。

Aiming at the problem that the vibration signal of rolling bearing is difficult to extract due to the characteristics of non-linear, non-stationary and low signal-to-noise ratio, a new fault extraction method based on symplectic geometry mode decomposition(SGMD) and multipoint optimal minimum entropy deconvolution adjusted(MOMEDA) theory is proposed. Firstly, a list of symplectic geometry components(SGCs) are obtained with SGMD decomposing the fault signal;secondly, SGCs are selected for signal reconstruction according to the correlation criterion, then, MOMEDA decomposition parameters are determined;finally, the reconstructed signal is processed with MOMEDA for enhancing the signal-to-nosise ratio, and envelope spectrum analysis is utilized to extract fault features. Simulated and experimental results verify that SGMD-MOMEDA can accurately extract the fault frequency of rolling bearings, and the comparison with the Empirical Mode Decomposition(EMD) shows that the SGMD is more accurate when reconstructing signals. This method has certain application value in the field of fault diagnosis.