基于分形和小波包理论的滚动轴承故障诊断

为了提高滚动轴承故障分形诊断的准确性,利用仿真信号对不同数据长度和不同信噪比下信号的盒维数和关联维数的差异进行对比,发现两种分形维对不同信号具有不同适应性;利用基于小波包分解能量图的特征信号强化技术,突出含噪轴承振动信号的故障信息特征,并对消噪前后振动信号盒维数进行计算和对比。分析结果表明,分形盒维数比关联维数更适用于分析含噪较重的信号;滚动轴承故障振动信号盒维数小于正常信号盒维数;相比原始信号,经小波包提取后不同类型故障振动信号的盒维数区分更为明显,诊断结果更加准确直观。

Fault Diagnosis of Rolling Bearings Based on Fractal and Wavelet Packets Theory

To improve the accuracy of fractal fault diagnosis of rolling bearings,the differences between box dimensions and correlation dimensions under different length of data and different SNR are compared using simulation signals.It is found that the two kinds of fractal dimensions have different adaptability to different signals.By using the characteristic signal strengthening technology based on the energy diagram of wavelet packet decomposition,the fault information characteristics of vibration signals of nois...