改进多尺度符号动力学信息熵及其在行星变速箱特征提取中的应用

针对行星变速箱在运行时产生的非线性非平稳振动,且故障特征信号微弱等问题,提出一种新的特征提取方法——改进多尺度符号动力学信息熵。在传统的符号动力学信息熵原理的基础上,通过改进传统方法的符号化过程,在考虑条件概率情况下计算信息熵,并引入多尺度概念,使得所提特征具有更大优势。最后求解行星变速箱故障模拟试验台采集到的三种状态下的振动信号改进多尺度符号动力学信息熵,并基于提出的特征评价指标对改进多尺度符号动力学信息熵、时频熵、排列熵、样本熵等特征的计算结果进行了对比。结果表明,该方法能够有效的提取行星变速箱运行状态特征,具有更高的敏感度。

MMSDE and its application in feature extraction of a planetary gearbox

Aiming at nonlinear non-stationary vibration and weak fault feature signals caused in planetary gearbox operation, a new feature extraction method called the modified multi-scale symbolic dynamic entropy (MMSDE) was proposed here. Based on traditional symbolic dynamic entropy (SDE) principles, traditional method’s symbolization process was improved, information entropy was calculated considering conditional probability, and the multi-scale concept was introduced to make extracted features have greater advantages. Finally, MMSDEs for vibration signals under 3 states collected on a planetary gearbox fault simulation test platform were solved. The results were compared with corresponding time-frequency entropy, permutation one and sample one based on feature evaluation indexes proposed here. The results showed that the proposed method can effectively extract a planetary gearbox’s operational state features with a higher sensitivity.