基于ITD与排列熵的齿轮故障特征提取方法研究

Research on Gear Fault Feature Extraction Based on ITD and Sample Entropy

针对齿轮故障信号的非线性、非平稳特征,采用本征时间尺度分解(ITD)结合排列熵的方法进行故障特征提取。采用ITD分解方法将原始信号分解得到一系列的PR旋转分量。通过剔出无意义的PR旋转分量,筛选出反映真实状态信息的分量,然后计算筛选出的PR旋转分量的排列熵。不同故障信号的PR旋转分量的排列熵大小不一,规律可寻,据此可以将排列熵的值作为元素构造故障特征向量。通过实验模拟齿轮正常、齿根裂纹、断齿和缺齿这4种状态,证明ITD-排列熵有很好的分类效果。

For the non-linear and the non-stalionary characteristics of gear faults signal, adopting the intrinsic time-scale decomposition （ITD） combined with the sample entropy method to extract fault features. With the ITD method, the original signal is deeomposed into a series of rotation components （PR）. Then by eliminating the meaningless components so that the components including real status information couhl be selected to calculate permutation entropy. The permutation entropy changed regularly with different fauh signals＇ PRs, and accordingly the sample entropy could be used as elements of fauh feature vector. Through experiments simulated under gear normal, tooth root cracked, tooth broken and missing teeth conditions, it is proved that the ITD-permutation entropy had good classification results.