中文摘要：

针对异步电机转子断条故障诊断中,原始信号包含的故障特征成分能量微弱,其提取过程较为繁琐,给断条故障的及时诊断带来不便,提出一种基于经验模态分解（Empirical Decomposition Mode,EMD）能量熵,主成分分析（Principal Component Analysis,PCA）和支持向量机（Support Vector Machine,SVM）相结合的新诊断方法,无须提取信号中的故障特征频率就能对电机断条故障做出准确的判断。该方法选取振动信号经过PCA处理后的EMD能量熵作为新的故障识别分类的特征量,然后支持向量机模型便可以根据断条故障前后振动信号EMD能量熵内在变化规律对转子正常和断条故障时的两种振动信号进行准确分类。通过实验分析表明,该方法操作便捷且简单,能够将转子正常和发生断条故障时的两种振动信号数据全部准确的识别分离,达到对转子断条故障进行有效识别诊断的目的,验证了方法的实用和有效性。

英文摘要：

In view to asynchronous motor rotor broken bar fault diagnosis,the energy of the fault characteristic components included in the original signal is weak,the extraction process is more tedious,and fault diagnosis is inconvenience to broken bar in time. Therefore a new method is put forward based on empirical mode decomposition（ EMD） energy entropy combination principal component analysis（ PCA） and support vector machine（ SVM）. The method does not need to extract the fault signal characteristic frequency and makes accurate judgment of motor broken bar fault. The method selection of vibration signal after dealing with the PCA of the EMD energy entropy as the characteristics of new fault identification classification,SVM model can then according to the article break fault vibration signals before and after the EMD energy entropy change rule of the rotor within normal and article break fault classification of two kinds of vibration signal accurately. Through the experimental analysis shows that the method is convenient and easy to operate,can the normal rotor and article break fault occurs in the two kinds of vibration signal data all accurate identification of separation,to make effective identification diagnosis for the rotor broken bar fault,the purpose of the validation of the method is practical and effective.