The Research Of Fault Analysis And Diagnostic Methods For Asynchronous Motors

Asynchronous motors have the advantage of simple structure, low cost and high reliability. So they made a wide range of applications in the modern industry, agriculture, transportation, and people’s daily life. Asynchronous motors’faults will not only damage the motor itself, it will also result in shut down of the motors, causing huge economic losses and catastrophic consequences. Therefore, the research on the characteristics of asynchronous motor fault and diagnosis method has very important value.Focused on the fault diagnosis for asynchronous motors, this paper completed the following tasks mainly:Various fault diagnosis methods of asynchronous motors are analyzed, advantages and disadvantages of these methods are compared. Based on the working principle and the electromagnetic relation of asynchronous motor, the characteristic mechanism of inter-turn short circuit and rotor bar breaking fault were analyzed.On the basis of accurate calculation of the stator and rotor windings inductance parameters, the relationship between inductance and harmonics is analyzed. Then, a multi-loop model of asynchronous motor is established. On the basis of multi-loop model and the mechanism of fault, simulation of asynchronous motors faults is researched. The obvious features of stator current of asynchronous motor are compared. The relationship between the number of broken rotor bars and a new fault feature is analyzed. The simulation also proved that stator winding faults affect three-phase current phase.Based on wavelet analysis theory, various wavelet analysis methods of signal noise reduction are compared. Through the camparation, a denoising method for asynchronous motor fault signal based on wavelet packet theory and energy threshold segmentation is put forward. It is difficult to extract rotor faults feature from the stator current signal of motors with the fluctuated load. So the demodulation characteristics of Hilbert transform for stator current signal is analyzed, and the feasibility of applying wavelet analysis to determine the load fluctuations and rotor bars breaking fault is studied. Through these studies, a rotor fault diagnosis method based on wavelet transform coefficients and Hilbert demodulation is proposed. The results of simulation and the analysis of measured signal proved that above methods could denoise stator current signal and extract rotor faults feature effectively, thus diagnose the rotor bar breaking fault accurately and effectively. Figure43, Table4, references61.