摘要：针对感应电机转子断条与偏心复合故障特征受定子电流基波信号的影响而难以提取的难点，提出了一种基于定子电流Park矢量模EMD分解的诊断方法。首先从理论上分析了Park矢量模在凸显复合故障特征方面优于单相瞬时功率，然后通过仿真验证了这一结论的正确性。仿真结果还表明，基于定子电流Park矢量模EMD分解方法诊断灵敏度高、故障特征直观清晰，同时提取的故障特征IMF分量还可以从定量上对故障严重程度作出分析，是一种优越的可行方法。

关键词：感应电机；Park矢量；功率频谱；EMD分解；复合故障
Abstract: This paper presented a diagnosis approach based on stator currents Park vector module EMD, since the characteristics of induction motors rotor bar broken and eccentric fault are difficult to extract, which may influenced by stator current fundamental wave. Firstly, by theoretical analysis, Park vector module was proved to be better than singlephase instantaneous power in highlighting complex fault characteristics. Then, a simulation example was employed to show the efficiency of the proposed method. Finally, simulation result demonstrated that the diagnosis approach based on stator currents Park vector module EMD was of high sensitivity, and fault characteristic was clear to observe. Besides, fault characteristic IMF components could be used to determine fault severity by quantitative evaluation. This proposed approach was proved to be efficient.

Key words: induction motor；Park vector； power spectrum； EMD； complex fault