一种旋转机械振动信号小波去噪方法

摘要

In view of the traditional method cannot be effective applied in non-stationary rotating machinery signal de-noising, a new wavelet analysis method is introduced. Based on the analysis of noise characteristics of the rotating machinery signal, the combination of hard threshold function and Module Maximum Value method is experimented. A new effective evaluation system is proposed from multi-perspectives and it can conclude a comprehensive and objective evaluation of the de-noising signal. Experimental results show that the method can effectively remove the noise and identify fault diagnosis with rotating machinery.%针对旋转机械的非平稳振动信号难以用传统方法处理的问题,本文讨论了小波方法在旋转机械振动信号去噪中的应用。针对旋转机械故障信号的特点,提出阈值去噪和模极大值相结合的小波去噪方法对振动信号进行处理,并利用多种指标对去噪后信号做出全而客观的评价。实验结果表明该方法可以有效的提高故障信号信噪比,有利于提高后期故障诊断的准确率。