振动信号检测技术研究及其在故障诊断中的应用

Vibration Signal Detection with Applications to Fault diagnosis

针对复杂机械系统异常振动信号的检测,提出一种新的无混叠指数双线性时间-频率变换探测方法,新变换方法可以避免常用双线性变换中的频率混叠和信息丢失等问题,能够有效的抑制交叉项,具有较高的时频分辨率,给出新的双线性时-频变换的离散算法,通过仿真实验与齿轮箱早期故障检测,对新的时频变换的检测效果进行了验证。

A novel approach to detect abnormal transient vibration signal for complex mechanical system is presented,where alias-free exponential bilinear time-frequency transformation is adopted to avoid the frequency alias and information loss in the traditional bilinear distributions. Simultancously,the cross-terms are effectively deduced with higher time-frequency resolution,which is validated by a gearbox early fault detection.