Research On Fault Diagnosis Of Key Components Of High Speed EMU Traction System

The reliability of the traction drive system plays a decisive role in driving safety of High-Speed EMUs.As the operating environment changes and the operating time increases,the key components in the traction control system will experience varying degrees of performance degradation,causing various failures.Therefore,this paper takes the key components of the traction drive system of the High-Speed EMU as the research object,and focuses on the fault mechanism of the traction inverter IGBT,current sensor and intermediate DC side capacitor.On this basis,an appropriate method is selected to design the fault diagnosis scheme,which is verified by simulation experiments.Specific research contents include the following four aspects:(1)Introduce the working principle and control method of the traction drive system,and analyze the failure causes and action mechanisms of the traction inverter IGBT,current sensor and intermediate DC side capacitor.The simulation model of traction drive system was built on MATLAB/Simulink platform,including pulse rectifier circuit,middle dc circuit and traction inverter circuit.The failure simulation of three key components is realized through the simulation model,which lays a foundation for the subsequent fault diagnosis.(2)Diagnose the open circuit fault of the inverter IGBT.The diagnosis content includes two parts: feature extraction and pattern recognition.When IGBT of traction inverter has an open circuit fault,part of waveforms of stator three-phase current will be missing.Based on this feature,this paper uses Wavelet Transform to separate frequency band and reduce noise of inverter IGBT under 22 fault situations,and extracts fault feature vector for the training of diagnosis model.In the selection of diagnostic model,this paper introduces BP neural network and support vector machine classifiers respectively,and verifies them by simulation.The simulation results show that both of them can diagnose the IGBT fault of the converter,but the latter is less dependent on samples and has stronger anti-noise capability.(3)Diagnose open circuit fault of current sensor and develop fault-tolerant control scheme.Based on the coordinate transformation,the diagnosis method obtains the current measurement value and the reference value in the ?? phase coordinate system through Clark transformation and Park inverse transformation respectively,and compares the difference between the two with the given threshold to judge the fault type.When the current measurement value is abnormal,the reference value is selected to replace it,so as to realize fault-tolerant control.In the selection of threshold value,this paper improves the existing literature research,and proposes a method to replace the named value with the unit value of each variable,which simplifies the related algorithm.In addition,considering the problem of IGBT misdiagnosis caused by sensor fault,this paper carries out the simulation of sensor fault alone and composite sensor and IGBT fault respectively in Simulink platform,and the simulation results verify the effectiveness of the above method.(4)Identify the degraded parameters of the DC side capacitance to achieve the purpose of fault diagnosis.The equivalent series resistance ESR and the equivalent series capacitance C were selected as the key degradation parameters,and the parameters were identified by the recursive least square method under the time-domain model and the improved three-phase current reconstruction method,respectively.Simulation experiments show that both methods can complete the identification accurately,but the improved three-phase current reconstruction method has higher identification accuracy and lower hardware cost,so it has better practical application value.