从振动信号中提取故障特征是滚动轴承故障诊断的常用方法。提出了利用零空间追踪算法和包络谱分析进行滚动轴承故障诊断的方法。首先对轴承故障振动信号进行零空间追踪(NullSpacePursuit)ff解,降低噪声的干扰,提取高能量成分。然后对高能量分量进行包络谱分析得到振动信号特征频率,与理论故障特征频率对比后确定故障类型。仿真和试验结果显示该方法能够成功提取特征频率,辨别轴承故障类型,具有一定的实际意义。
Exacting fault feature from vibration signal is a common methodology for rolling bearing fault diagnosis. Using Null Space Pursuit algorithm and envelope spectrum for rolling bearing fault diagnosis is proposed in this oaoer. Firstly separating the bearing fault signal in null space pursuit can reduce the affect of noise and extract subcomponents with a high energy. Then, calculate the envelope spectrum from the subcomponents and the signal's characteristic frequencies. Finally, determined the fault position by comparison with theoretical fault characteristic frequencies. The approach on simulation and real vibration signals verifies the effectiveness of the proposed method.