用自适应陷波器检测异步电动机转子断条的仿真

Simulation of Adaptive Notch Filter do Detect Rotor Bar Breaking in Induction Motors

转子断条是笼型异步电动机常见的一种故障,采用快速傅立叶变换(FFT)对定子电流进行分析是目前应用最广泛的转子断条在线检测方法,但是存在灵敏度低的缺点。当电机断条时,定子电流将产生(1-2s)f1频率的附加分量但幅值非常小,若直接将定子电流信号去作频谱分析,将受幅度很大的电网频率分量f1的影响,因而提出采用自适应陷波器的方法对异步电动机定子电流信号进行处理,通过LMS算法来调节自适应陷波器的两个权值,以达到对工频信号极大的衰减,虽不能完全消除但幅值已明显减小。其结果有利于转子故障特征量的提取,从而提高检测的灵敏度。仿真结果证明了该方法的有效性。

Rotor bar breaking is a common fault in the squirrel-cage motors. At present the most widely used detecting method for rotor bar breaking is frequency spectrum analysis of stator current by Fast Fourier Transform （FFT）,but low sensitivity is the disadvantage of this method. There is a very small additional current component of frequency in the stator current signal when motor has broken rotor bars, in case that the frequency spectrum analysis is carried out directly, because the line frequency component with large amplitude will influence on this result. In this paper, adaptive notch filter is applied to process the signals of the stator current in induction motors. Two parameters of the adaptive notch filter will be adjusted by LMS algorithm in order to attenuate the line frequency component greatly. Though this method can not eliminate the line frequency component completely, the amplitude will be reduced evidently. The result is a great help to extract the feature component of rotor fault and improve the sensitivity of fault diagnosis. The simulation show that the method is valid and effective.