Rotor Fault Diagnosis Of Induction Motor Based On NME-EMD

Motor is the most widely used electrical driving equipment at present.In industrial applications,the rotor of motor is prone to failure due to long-term high load operation or frequent start and stop,which will not only cause interference to normal industrial production activities,but also may harm personal safety.In view of rotor failure of induction motor,the traditional method is to make fast Fourier transform of the collected stator current signal and analyze it according to the simulation spectrum diagram.The fault characteristic component amplitude of the rotor is much smaller than the fundamental wave component,but the fault characteristic frequency is close to the fundamental wave frequency,so it is difficult to effectively identify the fault characteristics of the rotor.Rotor faults mainly include rotor broken bar and rotor eccentricity.Aiming at the problem that rotor broken bar fault of asynchronous motor is difficult to diagnose,this paper adopts the fault diagnosis method based on NME-EMD(EMD decomposition extraction maximum sample entropy principle),builds the SIMULINK model of rotor broken bar fault,and obtains stator current signal by adjusting parameters.Then,the EMD decomposition is performed to obtain different IMF components,which may contain fundamental frequency components,high-order harmonics and residual components.The IMF components containing fault characteristic information are selected for demodulation to obtain instantaneous frequency spectrum diagram,which can directly reflect the characteristic frequency of the rotor broken strip fault.The greater the severity of the fault,the more obvious the fault characteristic signal is.When the severity of the fault is small,the characteristic signal of the fault is easily submerged by the fundamental frequency signal,which increases the difficulty of the fault diagnosis process.Therefore,this paper improved the method based on EMD decomposition and adopted the method based on NME-EMD maximum sample entropy extraction.The IMF component of maximum sample entropy was selected for separate analysis,and notch filter was used for filtering to eliminate the influence of fundamental frequency signals.Fault diagnosis was carried out according to the obtained spectrum diagram.The improved method for rotor eccentricity fault diagnosis is roughly the same as that for broken bar fault diagnosis.The difference is that the rotor eccentricity fault module is adjusted to simulate rotor eccentricity fault under different working conditions in the diagnosis process.The simulation results show that the rotor fault diagnosis method based on NMEEMD eliminates the influence of fundamental wave component,and makes the fault features in different fault cases more obvious and easier to extract.With the deepening of the fault severity,the fault characteristic signal increases constantly,which makes the method more intuitive and accurate to detect the fault characteristics of the rotor.Figure[56]Table[2]Reference[80]...