利用吉布斯现象识别微弱振动脉冲

To Recognize Tiny Vibration pulse signal in Gearbox Based on Gibbs Phenomenon

周期信号与时域脉冲序列的频谱特点有明显差别,前者能量主要集中在一定的频段,后者能量分布较宽;二者傅立叶分解的余项收敛速度也有所不同。利用吉布斯现象所反映出的信号分析中的截断及误差原理,可以将经过调制的微弱脉冲的测量问题,给以比较好的解决。通过评价经过较强载波信号调制的微弱脉冲信号的高阶余项,可以对脉冲序列在高频范围内的能量优势给以获取、识别和评价。试验证明,该系统的技术参数满足机械系统特定脉冲故障分析的需求。文中所述的原理、测量和数据处理方法,对复杂机械中轴承的早期故障诊断具有非常重要的意义。

When analyzing the spectrums of a pulse series vs the periodic signals, there is an interesting phenomenon ,that the energy of pulse series is distributed in a more broad band than that of the periodic signals. So we can describe, recognize and estimate the pulse series energy by the higher frequency truncation error or remainder of the decomposing formula, at the same time ,the periodic signals should be filtered away,which could also be named as Gibbs Phenomenon. This smart technique is of significant value to the machine diagnosis , especially to evaluate the tiny faults in early stage of a bearing failure process.