Research On Detection Method Of High Powerinverter For Metro Vehicles

Urban traffic problems become more and more highlighted with the rapid development of urbanization process and the dramatic increase of city’s population. As the most rapid development transportation in recent years the Metro vehicle which have the characteristics of large capacity, on time, low pollution and high speed. At present, a lot of large cities as well as medium-sized cities have established rail transit to relieve traffic pressure. Auxiliary power supply system is one of the most important part of urban rail transit power supply system, which is used to provide power for interior lighting, air conditioning and ventilator. The comfort of passengers is also affected by auxiliary power supply system. The auxiliary inverter is the core of the auxiliary power supply system and is the components which has the highest failure rate. The current fault diagnosis and the protection of the auxiliary inverter power supply and protection is not perfect enough and there is no useful fault diagnosis methods for the open-circuit fault of IGBT which is used as the power device.The auxiliary inverter system of Shanghai pearl line is set as the research object in this paper. A simulation model of the auxiliary inverter system has been built in the software of MATLAB/SIMULINK. The influences of the aging of power device and the capacitance change of filter capacitor on output circuit are analyzed and the open circuit faults of auxiliary inverter power supply power device are analyzed emphatically. On the basis of the simulations of open circuit fault output waveform of the power device in the main circuit inverter system, the extraction methods of fault feature data are put forward and the characteristic data of various kinds of fault cases are built. the classification of the fault data failure data is achieved by the use of support vector machine(SVM).The following works are mainly done:First of all, the composition structures and the working principle of key components of auxiliary inverter system of Shanghai pearl line have been studied. The key componentsof auxiliary inverse system also have been carried on the simulation and the correctness of the simulation model of inverter system is verified. The influences of the change of internal resistance of the power device in the inverter and the capacitance change of filter capacitor on both output circuit and harmonic content are analyzed and the cause of the power device open circuit fault are analyzed emphatically. The different kinds of open circuit fault condition are classified and the changes of the output voltage of inverter after the open circuit fault are simulated. This article is mainly concern with the problem that there is no more than two power device with open circuit fault at the same time because the probability of the three power device failure at the same time is small.Then, The transform principle of clark is studied. The three-phase output voltage vector of the inverter is transformed into two phase space vector. According to the characteristics of different track of space vector performance under different fault types,the centroid method is adopted to extract fault feature data and the deficiency of the method is analyzed. According to the conduction order of power device, a method which utilize the space vector modulus values to extract the fault feature is proposed and the fault feature vector is built.Finally, the classification principle based on support vector machine(SVM) is studied.By the use of LIBSVM toolbox, the fault feature data classification is extracted. The recognition of inverter circuit fault has been implemented and the effectiveness of the method has been verified.