基于HHT变换和SVM的齿轮箱故障诊断

• 作 者: 姜旭刚 潘宏侠 郑朋
• 机 构: 中北大学信息与通信工程学院,中北大学机械工程与自动化学院,中国电信集团公司莱芜电信分公司,
• 出 处: 煤矿机械
• 发表日期: 2013年11期
• 关键词: 齿轮箱 Hilbert-Huang变换 支持向量机(SVM) 故障诊断

提出了基于HHT变换和SVM结合的齿轮箱故障诊断方法,介绍了固有模态函数、EMD分解和Hilbert谱以及支持向量机(SVM)理论。先对各种工况信号消噪,再利用EMD分解将信号分解为IMF分量,求出Hilbert谱和边际谱,再利用边际谱求出各信号的故障特征信息,最后利用支持向量机判别齿轮箱的故障类型。该实验证明了此方法诊断齿轮箱故障的有效性。

A gearbox fault diagnosis method based on HHT transform and SVM is proposed. The intrinsic mode function, EMD decomposition and Hilbert spectrum and support vector machine (SVM) theory are introduced. Firstly, the signals of various working conditions are de-noised, and then decomposed by EMD to decompose the signal into IMF components. The Hilbert spectrum and the marginal spectrum are obtained, and then the fault feature information of each signal is obtained by using the marginal spectrum. Finally, the support vector machine The type of fault. This experiment demonstrates the effectiveness of this method in diagnosing gearbox failures.