Research On Open-Switch Fault Diagnosis For Three-Phase Voltage-Source Inverters Based On Currents And Voltages

Power electronics technology,as one of important branches in the field of electric power,has been extensively applied in the fields of aerospace,nuclear power,smart grid,new energy power generation and rail traffic,among which three-phase voltage source inverter(VSI)plays an important role in electrical energy exchange.Any advanced electrical system must be able to ensure the reliability and safety of three-phase VSI,and it is of great significance to study the fault diagnosis technology for the improvement of the reliability and safety of three-phase VSI.In this paper,the new fault diagnosis methods are proposed based on current and voltage analysis.In addition,the proposed methods are verified in the simulation and experimental platform.(1)On the basis of summarizing the status quo of existing three-phase VSI fault diagnosis technology,and deeply understanding of the circuit topology and the working principle,this paper discusses the different of three-phase VSI open-switch fault types,analyzes the characteristics of output current and voltage before and after the fault occurs.(2)This paper presents a fault diagnosis method based on current observer.State space model has a good application in three-phase VSI fault diagnosis technology,but the traditional state space model is usually linear or based on switch function model.This paper introduces discrete signals to the state space model,and constructs the mixed logic dynamic model of the three-phase VSI.The residual values of observed and measured currents are used to compute the value of vector angle,which can realize detection and localization within a time equivalent to one sixth of the fundamental period of current.In addition,the average value of current residual vector in a fundamental period is adopted to deal with the interference caused by measurement errors,model parameter errors and noises,which is to the benefit of the improvement of reliability and robustness.(3)This paper presents a fault diagnosis method based on current signal processing.The fault detection method is improved to enhance the reliability and robustness,which can realize the fault detection in the eighth current period.After then,Clark transformation is used to get the normalized currents,and zero crossing detection method is used to convert the sine wave into square wave,which can obtain the current polarity.The current polarity and fault detection algorithm are used to realize fault diagnosis.In addition,the diagnosis method has the ability to ensure the robustness to the transient condition,such as load variations and current frequency variations,and prevent misinterpretation under the transient condition,although all of the switches operate in a normal state.(4)This paper presents a fault diagnosis method based on line voltage.The voltage-based method can be used to diagnose the inverter fault more quickly,but it usually needs to add additional hardware.In addition,it usually needs to set a threshold value for auxiliary judgment,which is often a fixed threshold and has no adaptive performance.This paper puts forward to improve the traditional method based on voltage model,and need to add no additional hardware.The proposed method only uses the existing information,such as current and duty cycle,to build a discrete mean line voltage model,and the deviation between expected value and estimated value of line voltage is used for fault diagnosis.Considering the parameter error,switch delay and other factors,an adaptive threshold value is designed.(5)The proposed fault diagnosis methods are simulated using MATLAB/simulink,and verified by experiment based on dSPACE platform.