Analysis And Diagnosis Research Of Open Circuit Fault Of Traction Inverter Of High-speed Railway EMUs

With the rapid development of China’s electrified railway,high speed EMUs account for more and more of the electrified railway.As the core component of the electric traction drive system of high-speed EMUs,power module devices mainly involve multi-field coupling in application occasions,and power module devices bear most of the electric thermal stress in this occasion,with high failure rate.By extracting fault features and accurately locating fault points,real-time online fault diagnosis is of great significance to the safe and reliable operation of traction drive system.Taking the traction drive system of CRH2 type EMU in China Railway high-speed(CRH)EMUs series as an example,the Neutral Point Clamped(NPC)three-level traction inverter has 12 power tubes,the occurrence of faults is frequent and the types are complex,so the single-tube and multi-tube faults of the power tube of three-level traction inverter can not be ignored.In order to further improve the accurate fault location of power module components of three-level traction inverter,this thesis focuses on the open circuit fault analysis,diagnosis and fault tolerance of one power device and multiple power devices.The main research contents are as follows:(1)Combined with the actual operation parameters of CRH2 EMU,the mathematical model of NPC three-level traction inverter of CRH2 EMU was established,and the electrical quantity of output side phase voltage,line voltage and three-phase current of the established model inverter was analyzed to verify the rationality and correctness of the established model.(2)Analysis of NPC three-level traction inverter under normal working mode and much power tube open-circuit fault switch state and circuit flow path of the switch tube.To pull the three level inverter output side current three-phase bridge arm as a diagnostic electric parameters,the contrast different work modes inverter output current distortion situation,classify fault diagnosis.(3)Fault diagnosis based on signal.The wavelet packet entropy of each fault electrical volume is extracted as fault feature information.The weights and thresholds of each transmission layer of BP neural network are further optimized by combining Particle Swarm Optimization(PSO)Algorithm with Genetic Algorithm(GA).It can classify and diagnose different faults,locate fault points accurately,verify and analyze the feasibility and real-time of the diagnosis method.(4)Construct the space voltage vector to analyze the normal operation of NPC three-level traction inverter and the voltage vector in fault mode.By connecting the faulty bridge arm with the DC bus bar,the three-level traction inverter can be derated and fault-tolerant under the condition of single-phase single-pipe failure.