Evaluation Of High Speed Railway Track Irregularity Based On Data-driven Mode

Based on the track geometry and vehicle response detection data,through the time-domain statistical analysis of track geometry,the time-frequency analysis of vehicle response,and the comprehensive evaluation analysis of track geometry and vehicle response,this thesis comprehensively evaluates the track state of the railway.The main research contents and conclusions are as follows:(1)Based on the preprocessing of the track inspection data,the time-domain characteristic parameters are selected,and the railway track is evaluated based on the sliding window.It is found that the sliding statistics can accurately locate the starting and ending position of the disease.The peak value of track irregularity is managed by using the score deduction function,and the operation acceptance management value is included in the local deviation evaluation,which can identify the potential bad sections.The mathematical statistics method is introduced to analyze the probability distribution characteristics of track irregularity.Generally speaking,the medium wave irregularity and vehicle body acceleration of the track obey the Logistic distribution,the long wave track irregularity obeys the Normal distribution,the track quality index(TQI)obeys the Weibull distribution,and the standard deviation of each single item irregularity obeys the three-parameter Burr distribution.(2)In view of the shortcomings of the current track quality index(TQI),considering the wavelength factor,the weighted track quality index(WTQI)is obtained based on coherence analysis,power spectrum calculation,and wavelet packet transform.Combined with the track spectrum inversion theory and probability distribution,its management value is determined.The index comprehensively considers the mileage,amplitude,and wavelength factors,which can effectively evaluate the track quality of the railway and identify the small value diseases that affect the vehicle response.(3)The existing comprehensive evaluation mainly adopts the analytic hierarchy process to determine the index weight,which is too subjective,and the evaluation system does not include the wavelength.Because of the above shortcomings,a comprehensive evaluation index system integrating track geometry and vehicle response,mean value management,and peak value management is established,the improved analytic hierarchy process entropy weight method is used to determine the index weight,and the track state is classified based on system clustering and fuzzy clustering to realize fine management.