基于PEGIES和TQWT的滚动轴承故障特征提取

Fault feature extraction of rolling bearing based on PEGIES and TQWT

针对滚动轴承在强背景噪声下周期性冲击特征难以提取的问题,提出一种基于周期增强的包络谱基尼系数(periodically enhanced Gini index of the envelope spettrum,PEGIES)和可调品质因子小波变换(tunable Q-factor wavelet transform,TQWT)相结合的滚动轴承故障特征提取方法。首先,以PEGIES为TQWT分解效果的评判指标,事先设定品质因子Q和冗余度迭代范围区间,以中心频率比CFR为阈值指标及最大分解层数公式确定对应品质因子Q分解下的最佳分解层数J。通过网格搜索的方式确定最佳品质因子Q,得到其对应的TQWT分解重构子带,选取大于PEGIES均值的子带进行合并处理得到最佳分量,通过Hilbert包络解调得到故障特征。为验证方法的有效性,将XJTU-SY滚动轴承加速寿命实验数据集和DDS(drivetrain diagnostics simulator)实验台实测信号作为研究对象,结合仿真信号结果,与其他方法进行对比,证实本文所提方法能在一定程度上降低转频的干扰,具有更好的故障特征比,能实现更加准确的诊断。

For the problem that rolling bearings were difficult to extract periodic impact features under strong background noise,a rolling bearing fault feature extraction method based on the combination of periodically enhanced Gini index of the envelope spectrum(PEGIES)and TQWT was proposed.Firstly,PEGIES was used as the evaluation index of TQWT decomposition effect,and the range interval of quality factor Q and redundancy iteration was set in advance,and the optimal number of decomposition layers J under the decomposition of corresponding quality factor Q was determined by the threshold index of central frequency ratio CFR and the formula of maximum number of decomposition layers.PEGIES mean value of the sub-bands were selected for merging to obtain the best component,and the fault characteristics were obtained by Hilbert envelope demodulation.In order to verify the effectiveness of the method,the XJTU-SY rolling bearing accelerated life experimental data set and the DDS test bench measured signals were used as the research object,and the simulation signal results were combined with other methods to compare.The results show that the proposed method can reduce the interference of rotational frequency to a certain extent,and has a better fault characteristic ratio,which can achieve more accurate diagnosis.