基于3s/2r变换的改进互功率谱MUSIC算法的电动机故障诊断

Motor Fault Diagnosis for Improved Cross Power Spectrum MUSIC Algorithm Based on 3s-2r Transformation

根据信息集成定理,从定子电流的故障特征出发,提出一种基于同步旋转坐标变换,将ia、ib、ic三相电流转化为id和iq的利用空间谱的分析方法。这种改进的互功率谱的方法能有效消除有色噪声,其锐化谱峰使得信噪比不同时的信号得到有效估计,且修正的MUSIC算法还解决了基波分量过大导致有效故障信号被湮没的问题,充分挖掘了定子三相电流中的故障信息。该方法在基于一对极的异步电动机转子断条和偏心启动情况进行了实验验证,并通过分析图谱得出结论,该改进的方法能有效发现电动机存在的故障。

According to the information integration theorem, and analyzes the fault characteristics of state current, based on synchronous rotating coordinator transformation, proposes one analytic method that used space spectrum to transform the three phase currence ia, ib , ic into id and iq. This improved cross power spectrum method can effectively eliminate the colored noise. The sharpen peaks effectively estimated the signals of different noise ratio. And also the improved MUSIC Algorithm is proposed to solve the problem when the fundamental harmonic is too strong to cover the characteristics of faults diagnosis. The optimized method can finds the fault information in three- phase stator current. To be more specific, an experiment for the problem of rotor-fault and eccentricity start-up is raised to verify the validity of this new method. And the improved MUSIC algorithm also solves the problem that fault signals.