复杂传递路径下轴承滚动体故障特征提取

为消除复杂传递路径对轴承滚动体振动信号的影响并提高故障特征提取的能力，研究了基于变分模态分解(VMD)、优化最大相关峭度解卷积(MCKD)和1.5维谱的轴承滚动体故障特征提取问题；分析了轴承滚动体原始振动信号特点、早期故障信号的特性以及复杂传递路径对振动信号的影响，运用VMD将原始振动信号分解为一系列本征模态函数(IMFs),提出了转频分量剔除方法，通过峭度准则优选2个峭度较大的IMFs分量进行重构；基于网格搜索法研究了MCKD算法参数优化方法，用以增强重构信号的周期性故障特征，消除复杂传递路径对轴承滚动体故障信号的影响；利用1.5维谱分析重构信号，建立了复杂传递路径下轴承滚动体故障特征提取新方法，实现了轴承滚动体故障的准确诊断；为了证明方法的有效性，选取美国凯斯西储大学轴承SKF6205基座滚动体数据进行试验验证与分析。试验结果表明：网格搜索法获得了MCKD算法的最优滤波长度与冲击周期参数(365、85),优化MCKD算法增强了重构信号的故障特征，减少了无关频率分量，明显降低了其他成分的干扰；提出的故障特征提取方法在0、735和1 470 W负载条件下均提取到了轴承滚动体的故障特征频...

Fault feature extraction of bearing rolling elements under complex transmission path

In order to eliminate the influence of complex transmission path on the vibration signal of bearing rolling elements and improve the ability of fault feature extraction, the problem of fault feature extraction of bearing rolling elements based on the variational mode decomposition (VMD), optimized maximum correlated kurtosis deconvolution (MCKD) and 1.5-dimensional spectrum was studied. The characteristics of original vibration signal of bearing rolling element and early fault signal and the influence of the complex transmission path on the vibration signal were analyzed. The VMD was employed to decompose the original vibration signal into a series of intrinsic mode functions (IMFs), and thus the frequency conversion component elimination method was proposed. Two IMFs components with large kurtosis values were selected for the signal reconstruction according to the kurtosis criterion. Based on the grid search method, the parameter optimization method of MCKD algorithm was studied to enhance the periodic fault characteristics of reconstructed signals and eliminate the influence of complex transmission path on the fault signal of bearing rolling elements. The 1.5-dimensional spectrum was used to analyze the reconstructed signals, and a new fault feature extraction method of bearing rolling elements under the complex transmission path was established, thus realizing their accurate fault diagnosis. In order to prove the effectiveness of the method, the data on the base roller of the bearing SKF6205 from Case Western Reserve University were selected for the was experimental verification and analysis. Experimental results show that the grid search method leads to the optimal parameter values (365, 85) of the filter length and the impact period in the MCKD algorithm. The optimized MCKD algorithm enhances the fault characteristics of the reconstructed signals, reduces the irrelevant frequency components, and significantly decreases the interference of other components. The proposed fault feature extraction method can effectively extract the fault feature frequency (140.6 Hz), double frequency (281.3 Hz), triple frequency (421.9 Hz) and all other components under the loads of 0, 735 and 1 470 W, and there is no influence from load conditions, thus eliminating the influence of complex transmission path on the fault feature extraction of bearing rollers. It follows that the proposed method can effectively solve the problems of the fault feature extraction and diagnosis of bearing rollers under the complex transmission path.