电机与负载轴承耦合故障的谱峭度诊断方法

Spectral Kurtosis Diagnosis Method for Coupling Failure of Motor and Load Bearing

当机电传动系统中电机轴承与负载轴承同时发生故障时,呈现出振动信号复杂、轴承故障信号信噪比差、故障特征不明显的特点。针对上述问题,提出了基于小波和谱峭度相结合的轴承耦合故障诊断方法。首先,通过小波分解可以将耦合故障的多频带故障特征分解到各个子频带中,减少不同故障之间的相互影响;其次,根据谱峭度最大原则自动选择最佳带通滤波进行滤波,对滤波后的信号进行包络分析,从包络谱中对故障特征频率进行快速有效的识别,进而确定故障类型和故障位置;最后,利用支持向量机实现电机轴承耦合故障的模式分类。实验结果表明,利用该方法能够有效地滤除噪声干扰,提取强谐波信号下的弱故障特征,效果优于传统谱峭度分析方法。

When the motor bearing and the load bearing fail simultaneously in the electromechanical drive system,the vibration signal is complex.The signal-to-noise ratio of bearing fault signal is poor,and the fault features are not obvious.To solve the above problems,a new method for bearing diagnosis based on the combination of wavelet and spectral kurtosis is proposed.Firstly,the multi-band fault features of coupling faults can be decomposed into different sub-band signal by wavelet decomposition.It can reduce the mutual influence between different faults.Secondly,according to the principle of maximum spectral kurtosis,the optimal bandpass filtering is automatically selected for filtering.The enveloped spectrum of filtering signal is obtained by Hilbert transform method.The fault characteristic frequency is quickly and effectively identified from the envelope spectrum,and then the fault type and fault location are determined.Finally,support vector machine(SVM)is used to classify coupling faults.Experimental results show that this method can effectively remove noise and extract weak fault features under strong harmonic signals.The effect is superior to traditional spectral kurtosis analysis method.