检测感应电机转子断条故障及故障程度的关键在于查找定子电流谱中频率为（1±2s）f的间谐波含有率。传统的谱分析由于加窗和截断带来的频谱泄漏,使特征谱难以辨识,采用现代谱估计中的扩充Prony算法,基于定子电流信号的 AR 参数估计,计算信号的谱线构成,依据故障电机稳态运行时的定子电流特征,选取合适的阶数,能够精确地检测到（1±2s）f故障特征分量的幅值和频率,从根本上回避了频谱泄漏问题,为感应电动机转子断条故障的MCSA检测法提供了一条新的途径。通过实验室实测算例验证了该方法的有效性,在电机空载和轻载运行时也能够准确地检测出故障特征信号,诊断出故障程度。
英文摘要:
Search of the Inter-harmonics component which frequency is （1±2s）f in stator current spectrum can complete the fault diagnosis for rotor broken bar of induction motor. Introduces a window during the traditional analysis method will cause the spectrum leakage, and unable to distinguish characteristic spectrum. In this paper, using the extended Prony which belong to modern spectrum estimation based on AR parameters of stator current signal, can detect the amplitude and frequency of fault feature component accurately, and avoided the problem of spectrum leakage in the fundamental. The method provides a new way for MCSA to detect rotor broken bar of induction motor. It also analyzes calculation examples by the test in laboratory, achieve the detection of fault characteristic when the motor is operating at no load or light load.
