摘  要：在电动机运行过程中，转子断条故障将导致电动机出力降低，性能恶化。因此研究更高效的电动机故障诊断方法来对其进行检测迫在眉睫。针对电动机转子出现断条故障时定子电流信号的非平稳特征，提出了一种基于经验模态分解（EMD）和神经网络相结合的转子故障诊断方法。该方法首先将原始信号分解为突出了原信号不同时间尺度的局部特征信息的内在模函数(IMF)分量，然后将各IMF分量输入到BP网络中进行训练学习和故障诊断。将此方法应用于电动机转子断条故障的识别，实验结果表明，该方法能快速准确地识别转子断条故障。
关键词：转子故障；经验模态分解；内在模函数；神经网络；故障诊断

Abstract: In the process of the motor running, the broken-bar fault of rotor will reduce motor-output and worsen its performance. So research more efficient motor fault diagnosis method to detect is extremely urgent. For non-stationary characteristics of the stator current signal when broken rotor bar fault occurred in motor, a fault diagnosis approach based on EMD and NN combined was proposed. First, the original signals were decomposed into a number of IMFs including highlighted the local characteristics information of original signals’different time scales; then, the IMFs were put into BP neural network training learning and fault diagnosis. The method is applied to the diagnosis of the broken rotor bars, the experimental results indicate that this method can identify rotor in failure rapidly and accurately.
Key words: rotor faults; EMD; IMF; NN; fault diagnosis