Research On Stator Inter-turn Short-circuit Fault Diagnosis Of AC Motor Based On Inverter Harmonics

With the popularization of energy-efficient equipment,variable-speed AC motors have been widely used in the fields of industrial production,railway transportation and aerospace.However,due to environmental erosion or mechanical vibration,the reliability of the AC motors is reduced.Consequently,intelligent industrial equipment with features of condition monitoring and fault diagnosis are increasingly drawing attentions.Stator winding inter-turn short circuit is one of the most common faults in the electrical motors.Detecting and diagnosing this kind of fault rapidly and effectively can be of importance to ensure efficient operation,reduce maintenance cost,guarantee personnel safety and lessen disastrous accidents.In this thesis,the stator winding inter-turn short-circuit fault of the AC motors is researched.Analyzing the modeling theory and the diagnosing technology,a fault diagnosis method based on the inverter switching harmonics are proposed,followed with its expanded strategies of interference suppression.Then,experiments are carried out for further verification.The structure of this thesis is organized as follows:First,reliable mathematical model provides a theoretical basis for the diagnosing research.In consideration of this,following the fault model of induction machine(IM)in stationary frame and interior permanent-magnet synchronous motor(IPMSM)in synchronous frame,a saturation model of IM with stator faults is deduced,and the frequency-response fault models of IM and IPMSM are proposed.Based on the models above,four different kinds of diagnosis methods are carried out,in the order of fundamental wave,low-order harmonic,high-frequency injected harmonic and the switching-band injected harmonic.In terms of the reliability and the robustness,the comparison among the methods indicates the short-comings of the traditional methods.Second,due to the multivariable and strongly coupled electromagnetic characteristics,traditional fault indicators can be easily affected by control mode,operating condition and external disturbance.In this thesis,the switching excitation is analyzed theoretically,and the expression of the fault characteristics is demonstrated.Then,a stator fault diagnosis method based on the inverter switching harmonics is proposed,and the motor status evaluation with high reliability and high robustness are achieved.To extract the fault information among diverse harmonics.a rapid extraction of weak signals in a strongly noisy environment is implemented by Teager-Kaiser operators,and the fault characteristics are obtained using the windowed DFT frequency-domain interpolation algorithm.According to the experimental results in fault IM platform,the proposed method is validated to be effective.Third,as stator asymmetry due to resistance variation generates the characteristic components similar to those a stator inter-turn fault does,diagnosing the inter-turn short circuit under the stator asymmetry condition can be difficult.Therefore,the effects of stator asymmetry on the controller internal signals are analyzed,and the internal relation between stator asymmetry and inverter harmonic interference is illustrated.Research shows that the interference components introduced by stator asymmetry and closed-loop regulator,can also be raised at the characteristic frequency,thus reducing the diagnosing reliability.Based on frequency-response fault IM models,a expanded strategies of interference suppression is proposed,with which the inter-turn fault can be diagnosed in case of stator asymmetry.According to the experimental results in fault IM platform,this strategy is proved to be effective.Finally,with a salient-pole motor,even-order space harmonics can be introduced in the air-gap function,and with the modulating effect,the amplitude of fault components decreases and the frequency thereof disperses.Consequently,the diagnostic accuracy can be reduced because of the mutual interference between adjacent spectral harmonics.In the thesis,the theoretical analysis is performed on the high-frequency impedance variation of fault IPMSM,following which the expression of fault characteristics is derived.Furthermore,based on the comparison among switching harmonics of different carrier-bands,the switching excitations that can be applied to the IPMSM are selected.Then,a decoupling strategy is proposed to achieve the diagnostic result under the effect of asymmetrical stator and salient-pole rotor,simultaneously.The experimental results of the IPMSM platform validate the effectiveness of the proposed method.