Zoom-Prony方法及其在转子断条故障严重程度定量估算中的应用

Zoom-Prony’s Method and Its Application in Quantitative Evaluation of Induction Machine Rotor Broken-Bar Fault

在基于定子电流信号分析的感应电动机故障诊断中,通过求基波两侧由转子断条故障引起的边频分量的幅值、相位和频率,可以进行故障严重程度的定量估算。由于Prony方法对噪声非常敏感,而实际电动机电流信号包含丰富的谐波和噪声,且故障特征分量的幅值很小,直接应用Prony方法很难估算出故障特征分量的参数。本文提出Zoom-Prony方法,将Prony方法与选带细化技术(Zoom技术)相结合,对转子断条故障特征频率附近的小频段进行局部放大分析,估算这两个边频分量的参数,提高了计算的准确性,缩短了计算时间。此外,由于故障特征频率与基频过于靠近且幅值相差太大,边频分量常常被基频分量所掩盖而无法得到正确的结果,因此本文通过两次应用该方法来消除基频分量的影响。仿真与实验验证了这一方法的有效性。

In motor current signal analysis （MCSA） based fault diagnosis, quantitative evaluation of induction machine rotor broken-bar fault is allowed if the frequencies, amplitudes and phases of the fundamental side-band components aroused by rotor broken-bar fault are known. There are a lot of harmonics and noise in the stator current, and the amplitudes of the fault feature components are small. So it is difficult to evaluate the parameters of these fault feature components correctly by Prony＇s method which is sensitive to noise. In this paper, the Zoom-Prony＇s method is proposed to derive the fundamental side-band components to evaluate the broken-bar fault quantitatively, in which the Prony＇s method and zooming technique are conjugated to analyze the spectrum in a given narrow bandwidth. The zooming technique improves the Prony＇s method in both accuracy and speed. In addition, the side-band fault feature components are always whelmed by the fundamental component. In this paper, the proposed method is used twice to diminish the influence of the fundamental component. The validity of the proposed method is verified by simulation and experiment.