结合相邻系数TQWT去噪在旋转机械故障诊断中的应用

The application of adjacent coefficient TQWT denoising to the fault diagnosis of rotating machinery

加强对旋转机械的故障诊断对提高产品质量和长期安全运行具有重要意义。然而在故障诊断的过程中,有用的弱特征通常是具有强大的噪声背景,从而增加了特征提取的难度。为此,提出一种使用相邻系数可调品质因子小波变换(TQWT)的诊断方法,这种新兴可调品质因子小波变换与传统恒定的品质因子小波变换相比,具有可以随着信号的振荡使品质因子一致的优良性质。与此同时,采用相邻系数降噪是为了避免对约定逐项阈值技术过度矫正。结合这两种方法的优点,提出的去噪方法比其它方法更具有实用性和有效性。这种方法被应用到一个带用外圈缺陷滚动轴承和齿轮箱故障的检测,处理结果表明,该方法能够成功地识别故障特征,表明该方法比传统的小波阈值去噪方法更有效。

To strengthen the fault diagnosis of rotating machinery is of great significance to improve the quality of products and long-term safe operation. However, in the process of fault diagnosis, the weak feature is usually a strong noise background, which increases the difficulty of feature extraction. So is proposed, and a use of neighboring coefficient the tunable Q-factor wavelet transform（TQWT） of the diagnostic method, the emerging new the tunable Q-factor wavelet transform and the traditional constant Q factor of wavelet transform new compared with along with the oscillating signal and Q factor of excellent properties. At the same time, the adjacent coefficient is used in order to avoid excessive noise correction to the agreed itemized threshold technology. Combining the advantages of these two methods, the proposed method is more practical and effective than other methods. This method is applied to a belt with outer race fault rolling bearing and gear box fault detection, results show that, the method can successfully identify fault features and show that the method than the traditional wavelet threshold to noise method is more effective.