滚动轴承故障诊断中的分形

Fractal in Fault Diagnosis of Rolling Bearings

简述了滚动轴承运转中时域序列所表现出的自相似性，重构了振动信号对应的嵌入相空间的序列，计算得出分维数，实验结果表明，振动时域序列的分维数在不同工况、不同间隙下是有差别的，可以做为识别轴承故障的特征量。

Self-affine of time series which exists in vibration signals of rolling bearings is investigated. The vibration signals' imbedding space is rebuilt, and each the fractal dimension is calculated. Some experimental results are presentted. One technique based on the detecting fractal dimensions of time series in different working status is developed for recongnizing the rolling bearings' fault characteristics.