基于改进小波算法的舰炮齿轮箱故障特征提取

Feature Extraction of Naval Gun Gearbox Based on Improved Wavelet Arithmetic

小波变换具有优良的时频分析特性,适合于提取齿轮箱等机械设备的故障特征。但是由于小波滤波器的频响应特性不理想,因此直接对信号进行小波变换会产生一定的虚假频率,从而影响故障特征提取的准确性。针对小波分析的混频现象,在深入分析该现象产生原因的基础之上提出一种改进算法,并给出了利用该算法提取故障特征的具体步骤,为舰炮齿轮箱故障特征提取提供了一个新途径。仿真结果表明,该方法能有效地消除混频现象,提高了故障特征提取的准确性。

Wavelet transform has well time frequency analysis characteristic and is fit for extracting the fault feature of machine like gearbox.But it would produce false frequency if the signals were transformed directly by using the wavelet method because the filter′s filtering property is not ideal which consequently influence the veracity of fault feature extraction.Improved wavelet arithmetic was brought forward by in-depth analysis aimed at the frequency mixing phenomenon,and the concrete steps how to use it also were brought forward.This provides a new way for extracting the fault feature of naval gun gearbox.The results of simulation indicate this method could eliminate the frequency mixing phenomenon effectively and increase the precision of feature extraction.