Study On The Track Condition Evolution Of Wuhan-Guangzhou High-speed Railway

With the growth of railway operation time,the track condition changes and track defects gradually emerge,which will have an impact on the safety and comfort of traffic.From all over the world,track inspection vehicles or comprehensive inspection trains are commonly used to monitor changes in track condition,and China uses high-speed comprehensive inspection trains to conduct regular inspections of high-speed railway infrastructure.The Wuhan-Guangzhou high-speed railway has been in operation for 12 years since its operation,and accumulated a large amount of track status data.Based on the long-term test data,it is of great significance to carry out the analysis of WuhanGuangzhou high-speed railway track state,evolution and prediction method,and to guide the intelligent operation and maintenance of high-speed railway track.In this paper,an in-depth study has been carried out on the track state evaluation and long-term evolution law of Wuhan-Guangzhou high-speed railway,and the main contents are as follows.（1）Through the correlation coefficient analysis,coherence analysis and transfer analysis of track geometry and vehicle response,research on the correlation between track geometry and vehicle response.The peak value of the coherence function between the lateral acceleration and the alignment of the vehicle body in the subgrade section exceeds 0.2 at the wavelength of60m;the peak value of the coherence function between the lateral acceleration and the cross level of the vehicle body at the wavelength of 100 m is close to 0.7,and the peak value of the coherence function between the vertical acceleration and the cross level of the vehicle body at the wavelength of 100 m exceeds 100 m.0.2;there is almost no correlation between the acceleration of the car body and the gauge;the coherence function of the vertical acceleration of the car body and the profile of the bridge section has a peak value of more than 0.6 at the wavelength of 32 m and its frequency multiplication position;the correlation coefficient of the vertical acceleration of the car body and the height is high;The correlation coefficient between the lateral addition of the car body and the track direction is relatively high.（2）Research on track condition evaluation index.Through correlation analysis and trend analysis,the evaluation indexes of WuhanGuangzhou high-speed railway track state are selected,including: profile,alignment,long-wave profile,long-wave alignment,gauge,cross level and twist,gauge change rate,standard deviation of long-wave profile,standard deviation of long-wave alignment,TQI,peak value of vehicle lateral acceleration,peak value of vehicle vertical acceleration,maximum value of section derailment coefficient,maximum value of section wheel load reduction rate,maximum value of section axle transverse force,standard deviation of vehicle lateral acceleration and standard deviation of vehicle vertical acceleration.（3）Comprehensive evaluation method of track condition of Wuhan-Guangzhou high-speed railway.The evaluation system of track condition of Wuhan-Guangzhou high-speed railway is constructed,which includes track geometry local state,track geometry section state,vehicle response local state and vehicle response section state.A comprehensive evaluation method of Wuhan-Guangzhou high-speed railway based on analytic hierarchy process is proposed.The expert experience method and linear piecewise deduction function are used to realize quantitative classification by grading the whole kilometer,and the model is verified.The upward data of Wuhan-Guangzhou high-speed railway in July,2017 is selected for evaluation.It is found that the local state of track geometry,the local state of vehicle response and the state of sections are good,while the geometric state of track sections is poor,with more than 50 points deducted for many sections,which is mainly caused by TQI overrun caused by bridge creep and arching.It is necessary to arrange fine adjustment as soon as possible.（4）Through standard deviation analysis,cumulative distribution analysis and power spectrum density analysis,research on the development law of track geometry of Wuhan-Guangzhou high-speed railway.The development trend of gauge and alignment of subgrade and tunnel section is relatively stable from the operation to the fine adjustment of track.The profile and cross level first show an upward trend,and then remain stable.After the operation of the line,the standard deviation of profile in sections of the bridge begin to rise,and become stable after 5～6 years,and there is no further development trend after the fine adjustment.It is recommended to carry out the fine adjustment of the track after the line is opened for 5～6 years.After the operation of the line,the gauge and alignment show a steady trend,hardly developing,and the cross level rises after operation,and then remains stable.The standard deviation of each track geometric parameter after fine adjustment under each basic structure is obviously reduced.Linear regression,sliding linear regression and LSTM models are used to predict the state development law of track geometric sections under different foundations,and the prediction effect is good.（5）Through waveform comparison,standard deviation analysis,spectrum analysis,transfer analysis and coherence analysis,research on the factors affecting the vehicle response of the Wuhan-Guangzhou high-speed railway.The detection speed has little influence on the detection results of track geometry,but has great influence on the detection of vehicle response.With the increase of speed,the peaks of 1Hz lateral and vertical acceleration spectra shift to the left,and peaks gradually increase.The change of the transfer function of the test data of the same vehicle for different times is small,which indicates that the performance of the longterm service vehicle changes not so much;There are differences in peak positions,peak values and line heights of horizontal and vertical acceleration spectra of different types of inspection vehicles,such as CRH380 AJ,CRH380BJ,etc.The vehicle response inspection data of different vehicles of the same model,such as CRH380AJ-0201,0202 and 0203,have good consistency.The geometric changes of the track before and after the re-profiling of wheels are small,and the local abnormal situation of the lateral acceleration of the car body is obviously improved after the re-profiling of wheels.Grinding the rail has little influence on the track geometry,but great influence on the vehicle response,and it may cause local abnormal vehicle acceleration.The detection direction has little influence on the track geometry,but has great influence on the vehicle response,especially on the lateral acceleration of the car body.（6）Research on the development law of vehicle response of Wuhan-Guangzhou high-speed railway.The acceleration of the car body under different foundations and linear shapes has the same development trend.The geometric development trend of the horizontal and lateral track of the car body is the same;the vertical trend of the car body is the same as that of the vertical track.due to changes of the car.Using linear regression,sliding linear regression and LSTM model to predict the development law of the response of the same vehicle,the prediction effect is good.