中文摘要：

传统的通过FFT变换对电机定子电流信号直接进行频谱分析,难以有效识别出转子断条故障。本文提出了一种在频域采用RELAX算法进行故障诊断的方法,它根据信号最小能量准则,结合粗估计和细估计,得到各频率分量的幅值和相位信息,从而直接构造各频率成分的复频域表达式,最后在频域内消除基波分量。相对于在时域内剔除基波分量的方法,它具有直接,快速的特点,缺点是对各频率分量的参数估计不是特别准确,但实验证明,该方法有效地突出了电机故障特征,在电机早期故障诊断中具有一定的应用价值。

英文摘要：

The traditional method of spectrum analysis on the current signal via FFT is hard to diagnose the fault of the broken rotor bars.This paper presents a fault diagnosis method using the RELAX algorithm in frequency domain.It can estimate the amplitude and phase values of various frequency components using coarse and fine estimation according to the criterion of minimum energy.Finally the fundamental component can be eliminated in the frequency domain when the domain expressions of the above frequency components are constructed.Compared with the method of eliminating the fundamental component in time domain,this method has the advantage of computing faster but less accurate.However,it has been proved that the algorithm can highlight the fault characteristic and value a lot in early motor fault diagnosis.