希尔伯特变换在故障诊断中的应用

许多机械故障信息以调制形式存在于振动噪声之中,希尔伯特变换是提取调制信息的一种有力工具。本文首先介绍希尔伯特变换在数字信号处理中提取幅值和相位调制方法,它利用调制信号和它的希尔伯特变换形式构成的解析信号,进行取模和反正切变换,即可获得包络和相位信号。这方法简单,适合于微机实时处理。由于现场实测信号存在各种噪声和干扰,对于被处理的信号需作预处理,因此本文引入时间同步平均,自适应消噪技术用于信号预处理,这有利于提高信号处理结果的信噪比以及诊断的可靠性。最后本文对三个典型机械零件故障的实际诊断作扼要介绍,实测说明希尔伯特变换和它相关的预处理技术结合应用于故障诊断是非常成功的,它为机械故障的早期诊断,和提高诊断灵敏度开辟了新的途径。

HILBERT TRANSFORM APPLICATION IN FAULT DIAGNOSIS

Many mechanical fault information occur in the form of modulation of vibration and noise signal, however Hilbert transform is a powerful tool for extracting modulation informa tion. In this paper, the demodulating method is introduced using Hilbert transform, the enve lope and phase modulation information are extracted by taking module and inverse tangen transform for an analytic signal which is constructed with the modulation singal and its Hilbert transformation. This method is simple in algorithm, and it is adapt for signal pro cessing in real time. Due to the random noise influence on measurement in practice, the mod ulation signal is needed to be preprocessed, then the technique of Synchronize Average and Adaptive Noise Cancelling are used in signal processing, it is favorable to increase the rate of signal to noise and reliability of fault diagnosis. At last three engineering examples of faul diagnosis in typical mechanical parts are presenten, it shows that Hilbert transform and signa preprocessing technique related with it applied in fault diagnosis is very successful, and it is considered as a new route to early fault diagnosis and increment of sensitivity in faul diagnosis.