基于时频融合的转速估计及轴承故障特征提取研究

Speed estimation based on time-frequency fusion and its application in feature extraction of bearing fault

针对通过齿轮啮合频率与转速间存在的比例关系估计齿轮振动转速,提出基于齿轮啮合振动的时频融合分布估计转速新方法,将Wigner-Ville分布与小波尺度谱进行融合,对融合后的时频分布用峰值搜索法估计转速。与基于时频分布的转速估计方法相比,该方法的估计精度与对噪声的不敏感度更高。用该方法估计齿轮传动装置转速,并对轴承故障进行阶比分析以评估其效果表明,该方法能有效获得轴承故障阶比。

The rotating speed of a transmission system provides critical information of the system dynamic features, it is helpful for fault diagnosis with varying rotating speed. For a gear transmission with a fixed transmission ratio, the rotating speed can be estimated with the gear meshing vibration because there is a proportional relationship between the gear meshing frequency and the rotating speed. A speed estimation method based on time-frequency fusion of gear meshing vibration was proposed here. The rotating speed was estimated using the fusion of Wigner-Ville distribution and wavelet scalogram, and the peak search of the fusion image. It was shown that compared with the traditional speed estimation method based on time-frequency distribution, the proposed method has a higher precision and a less sensitivity to noise; this speed estimation method is applied to the speed estimation of a gear transmission and then the estimated speed is used for order analysis of fault bearing vibration ; at last, bearing fault orders can be effectively obtained.