基于独立分量分析与希尔伯特-黄变换的轴承故障特征提取

滚动轴承早期故障信号具有能量小、频带分布宽等特征，易受到其它能量较大振源信号的干扰。传统的希尔伯特-黄变换（HHT）对信噪比大、多频率调制信号常因不能对其所包含的固有模式函数（IMF）实现准确分离和去除调制干扰分量而失效。本文提出了基于HHT和独立分量分析（ICA）的滚动轴承诊断新方法。该方法首先利用经验模式分解（EMD）将滚动轴承振动信号分解成若干平稳的本征模式函数IMF分量，通过提取若干包含主要信息的IMF分量，应用带通滤波器和Hilbert变换获取IMF分量的高频包络波形，再应用ICA分离包络波形并进行频谱分析，进而判断滚动轴承的运行状况。仿真和试验分析结果验证了本方法的可行性。

Rolling element bearing fault feature extraction based on HHT and independent compoment analysis

Vibration signal generated by incipient faults of rolling element bearing is usually with low energy and dispersed frequency distribution,it is easily merged in strong disturbances of other vibration sources.Using Hilbert-Huang transformation(HHT) directly may fail to remove noise from observed modulated signals and separate their intrinsic mode functions(IMFs) accurately since they are with lower signal-to-noise ratio and multi-frequency modulation.Here,a new fault-diagnosis approach was proposed based on ...