弧齿锥齿轮早期缓变微小故障诊断与研究

Slowly Early Varying Small Fault Diagnosis and Research of the Spiral Bevel Gears

针对矿山传动机械中硬齿面弧齿锥齿轮的早期缓变微小故障诊断和识别的问题,提出了一种结合小波包特征提取和支持向量机多分类算法结合的故障识别方法。采集矿山现场的弧齿锥齿轮振动信号,经小波包分解算出信号能量谱,以能量谱作为SVM的信号特征输入。结果表明,该方法用在硬齿面弧齿锥齿轮诊断时,故障的识别率为100%,是一种可行的工业齿轮故障诊断方法。

Against the problems of early slowly slight fault diagnosis and identification of hardened spiral bevel gears in mine transmission machinery, a combination of wavelet packet feature extraction and fault identification method combined with sup- port vector machine multi - classification algorithm was adopted in this article. Acquisition vibration signal were decomposition by wavelet packet and then the spectrum of the signal energy ware calcu- lated. The energy spectrums were regarded as the signal characteristics of the input of the SVM. The results show that the method used in hardened spi- ral bevel gears diagnostic fault identification can reach the rate of 100~ and it is a viable industrial gear fault diagnosis method.