基于参数优化特征模态分解的强背景噪声下滚动轴承故障诊断

Fault diagnosis of rolling bearing under strong background noise based on POFMD

为准确提取被强背景噪声掩盖的滚动轴承故障信息,提出一种参数优化特征模态分解(parameter-optimized feature mode decomposition,POFMD)方法。首先,为解决特征模态分解(feature mode decomposition,FMD)方法的输入参数依赖人工经验选取的问题,以平方包络谱峭度(kurtosis of the square envelope spectrum,KSES)为权值,结合平方包络谱基尼系数(Gini index of the square envelope spectrum,GISES)构建加权平方包络谱基尼系数(weighted Gini index of the square envelope spectrum,WGISES)作为目标函数,通过优化算法确定FMD的最优参数组合;其次,为解决FMD的主模态分量难以选取的问题,通过计算所分解模态分量的KSES值选取主模态分量;最后,通过包络谱分析实现故障诊断。经仿真信号和实测信号分析,验证了POFMD在强背景噪声下滚动轴承故障诊断中的有效性。与变分模态分解、最大相关峭度解卷积和谱峭度相比,POFMD有更优越的故障特征提取性能。

Here,to accurately extract fault information of rolling bearing covered by strong background noise,a parameter optimized feature mode decomposition(POFMD)method was proposed.Firstly,to solve the problem of manually selecting input parameters for feature mode decomposition(FMD)method,taking kurtosis of square envelope spectrum(KSES)as the weight,Gini index of square envelope spectrum(GISES)was combined to construct the weighted Gini index of square envelope spectrum(WGISES)as the objective function,the optimization algorithm was used to determine the optimal parametric combination for FMD.Secondly,to solve the problem of difficult selection of main modal components of FMD,its main modal components were selected by calculating KSES value of the decomposed modal components.Finally,fault diagnosis was realized with envelope spectrum analysis.The effectiveness of POFMD in diagnosing rolling bearing faults under strong background noise was verified with analyses of simulated and actually measured signals.It was shown that compared with variational mode decomposition(VMD),maximum correlation kurtosis deconvolution(MCKD)and spectral kurtosis(SK),POFMD can have more superior fault feature extraction performance.