基于双树复小波包及SVM的齿轮故障诊断

Gear Fault Diagnosis Based on Dual-tree Complex Wavelet Packet and SVM

齿轮是卫星传动与控制系统的重要组成部件,其故障诊断的效果在一定程度上反映了卫星系统在役考核中的可靠性情况.针对齿轮故障诊断特征提取难、故障识别率较低的问题,提出了一种集合双树复小波包分解、主成分分析及支持向量机的齿轮故障诊断方法.利用双树复小波包变换将原始信号分解为3层,得到8个子信号.再分别以信号时域指标、频域指标、时频域结合指标为特征对子信号进行特征提取.利用主成分分析对特征向量进行降维,将降维后的数据输入到支持向量机中进行故障诊断.研究结果表明,利用频域指标对齿轮振动信号进行特征提取的效果最佳,并且方法的故障总识别率达到100%,高于以往采用的结合经验模态分解及支持向量机的方法,为卫星齿轮故障进行快速准确识别提供了有效的技术手段.

Gear is one of the important components in satellite transmission and control system,and its fault diagnosis effect reflects the reliability of the satellite system in service assessment to a certain extent.Aiming at the problem of difficult feature extraction of gear fault diagnosis and low fault recognition rate,a gear fault diagnosis method based on ensemble dual-tree complex wavelet packet decomposition,principal component analysis and support vector machine SVM is proposed.The dual-tree complex wavelet packet transform is used to decompose the original signal into 3 layers to obtain 8 sub-signals.Then take the time domain index,frequency domain index and time-frequency combination index as features to extract the features of the sub-signals.Principal component analysis is used to reduce the dimension of the feature vector,and finally the reduced data is input into the support vector machine for fault diagnosis.The results show that the feature extraction of gear vibration signal using frequency domain index is the best,and the total fault recognition rate reaches 100%,which is higher than the previous method combining empirical mode decomposition and support vector machine.It provides an effective technique for rapidly and accurately identifying gear faults.