Research On Fault Prediction And PHM Application Of Traction System Of Shanghai-Nanjing High-Speed Railway CRH2 EMU

In recent years,with the continuous progress of the middle and long term railway network planning in China,more and more high-speed rail lines are designed,built and operated.Meanwhile,the number of CRH series EMUs invested in the high-speed rail line is also expanding.How to carry out scientific management of such a large number and more complex rail transit passenger transport equipment has become a huge problem facing the railway operation management department,so the technology of equipment fault prediction and health management based on "CBM" has become a hot spot of research.With the help of modern advanced sensor technology,collect and store a large amount of real-time monitoring data produced by the EMU equipment system every day,using the system information reflected by these data to predict the current running state of the equipment and on this basis to develop advanced maintenance strategies,it is an important research content of the maintenance work of the high speed rail CRH EMU at present.The traction system(power supply part)of CRH2 type EMU running on Shanghai Nanjing high speed rail line is selected as the research object.Aiming at the limitation of the traditional equipment maintenance mode of EMU "plan",through the construction of equipment degradation state of the hidden semi Markoff model based on(HSMM)to estimate the performance of the current state and remaining life of the PHM evaluation system.The purpose of this paper is to transform the current maintenance mode of EMU equipment from the traditional "plan repair" into "state repair",which provides scientific decision basis for the railway operation management department to improve the maintenance efficiency,reduce the maintenance cost and improve the reliability of the equipment operation.The main work of this paper is as follows:(1)The traction system of CRH2 type EMU running on Shanghai Nanjing high speed rail line is taken as the research object.By analyzing the structure and working mechanism of the system,the characteristic parameters of the performance state of the system are extracted,and the whole life cycle of the system is divided into four degenerate states based on the time series according to the system state degradation mechanism.(2)Based on the analysis of the characteristic parameters of the system and the mechanism of state degradation,the HSMM model based on the degenerate state of the system is constructed and used to evaluate the current degradation state of the system.Based on the current degradation state,the PHM estimation of the remaining service life of the system is made.(3)Because of the defects of the HSMM model itself,the precision of the HSMM model trained by the traditional BW algorithm combined with the field experiment data is very low,so the improved PSO algorithm(TPSO)is used to optimize the parameter estimation of the state classifier and the whole life cycle model,and the HSMM model is expected to improve the system state recognition and the residual life estimation.The accuracy of the calculat ion.(4)The application of the pre processed system monitoring data to the improved HSMM model,the fault prediction and residual life estimation are carried out for the traction system of CRH2 type EMU.The result shows that,the improved HSMM model has greatly improved the accuracy of prediction and the accuracy of PHM estimation of residual life compared with the traditional HSMM model.