Three-phase Asynchronous Motor Stator Failure Diagnostic Method Research

The asynchronous motor is a type of power device widely used in industry,agriculture and automated production processes.The motor stator winding easily fail due to prolonged high-load operation in a harsh working environment.This situation leads to abnormal operation of motor and other equipments of power system.Moreover,serious production accidents occur.Huge economic losses can be avoided if the asynchronous motor is diagnosed in time at the initial stage of the fault.This thesis takes three-phase induction motor as the research object.The diagnosis methods of motor stator open-circuit fault and stator inter-turn short circuit fault have been studied.The following work will be mainly completed:Firstly,this thesis establishes the voltage,current,electromagnetic torque and flux linkage equations of a three-phase asynchronous motor under three conditions.The conditions include normal,single-phase open-circuit fault and turn-to-turn short-circuit fault.The stator open-circuit and inter-rotor short-circuit faults are simulated using MATLAB/Simulink software,Besides,the current,voltage,speed and electromagnetic torque are analyzed under normal and fault conditions.To provide a database for subsequent fault diagnosis,the motor’s three-phase stator current signal is collected in normal and fault conditions.Secondly,a fault diagnosis method is proposed based on wavelet transform and knowledge classification,due to low accuracy of the traditional stator open-circuit fault diagnosis method of asynchronous motor.To begin with,the wavelet transform is employed to decompose the stator current signal of the three-phase induction motor.Wavelet transform decomposition coefficients are obtained for motor signals in normal and fault states.Next,the RMS value of the signal wavelet packet decomposition sub-band node coefficient is calculated and used as a fault characteristic value.Finally,the defect feature values are classified using the Bayesian classifier.Compared with BP neural network and classified decision tree method,the simulation results show that this algorithm accurately identify the open-circuit fault of the stator winding of threephase induction motor.Therefore,shows that the method has high diagnostic accuracy and high diagnostic speed.Thirdly,this thesis proposes a fault diagnosis method based on Welch power spectrum support vector machines.In order to solve the problem of spectrum leakage in the diagnosis of short circuit faults between stator windings of three-phase asynchronous motors using Fourier transform.To begin with,the current signal of the original three-phase induction motor is denoised using the wavelet transform.Next,the Fourier transform of the three-phase current is improved by adding the Hanning window function.The Welch power spectrum and power spectral density distribution of the resulting three-phase current signal are used as fault characteristics.Finally,the short-circuit fault between the stator windings is classified using the support vector machine.The simulation results show that the fault diagnosis method can accurately identify the short-circuit fault between turns of the stator winding of asynchronous motor.The effectiveness and feasibility of this method are verified.Finally,a physical experimental platform for online diagnosis of three-phase asynchronous motor systems was established.Completed the overall,hardware,and software design of three parts.The experimental platform collects the stator current signals of the motor stator winding during normal and open circuit faults.Then,the proposed algorithm is verified and the experimental results show the feasibility and effectiveness of the proposed diagnostic method.