滚动轴承性能退化评估的自适应频带熵能比指标

Performance degradation assessment of rolling bearings based on the entropy energy ratio of adaptive frequency band

针对轴承的性能退化评估中常规时域统计参数如均方根值(RMS)对早期故障不敏感,而基于概率估计和空间距离的评估方法存在计算量大、模型较为复杂等问题。将能检测时间序列细微突变且计算效率高的排列熵(PE)引入滚动轴承性能退化评估;同时考虑到PE对振动信号幅值不敏感,从而无法准确地反映振动信号不同频带能量分布情况,在PE的基础上提出一种新的性能退化评估指标——熵能比(EER);通过计算离散实验数据和全寿命疲劳实验数据的RMS,PE和EER,对比各指标与故障程度的一致性以及对早期故障和故障程度变化的敏感性,结果显示EER相较于RMS和PE具有对早期故障更敏感且与轴承故障发展趋势一致性更好等优点。通过工程实际数据分析进一步验证了EER指标的有效性,该研究结果有望为性能退化程度量化评估提供一种性能优异的指标。

For bearing performance degradation assessment(PDA),time-domain statistics and similarity-based methods using probability or distance measures are frequently used.However,time-domain statistics parameters like root mean square(RMS)are insensitive to early defect,while similarity-based methods are of complex models and time consuming.As such,permutation entropy(PE)which can detect slight mutations in time series and has high computational efficiency,was introduced to bearing PDA.Though PE has its advantages,it doesn’t consider the absolute amplitude of vibrations and thus is unable to depict the signal’s energy distribution over frequencies.In the present work,a new index named entropy energy rate(EER)was proposed.Artificially induced defects and run-to-failure data set from rolling bearings were processed to demonstrate the advantages of EER over RMS and PE in terms of the consistence between the indices and fault severity levels,as well as the sensitivity of the indices to early fault and the change of fault degrees.Furthermore,results on field data verify the effectiveness of the EER index in bearing PDA,which means in turn EER is a potential PDA tool.