Research On Mileage Deviation Correction Algorithm Of Dynamic Detection Data Based On Track Static Detection Data

Track geometric state detection plays an important role in ensuring the safe and stable operation of trains.During the dynamic detection of track geometry state,due to the dynamic error of photoelectric encoder,wheel slide and slip,the error written into the detection system by the operator and other factors,the detection data will inevitably produce mileage deviation,resulting in waveform misalignment.When directly using the data with mileage deviation,it will affect the evaluation accuracy of track quality state and the reliability of line status deterioration trend analysis,and cannot ensure the effect of line on-site maintenance.The existing research results of dynamic detection mileage deviation correction have some problems,such as insufficient accuracy,high equipment maintenance cost and inconsistency with the line mileage after correction.In order to realize the accurate correction of dynamic detection mileage deviation and make the dynamic detection data mileage close to the line mileage.In this thesis,a mileage deviation correction method of dynamic detection data based on static detection data is proposed,and the principal point iteration correction(PPIC)algorithm suitable for mileage deviation correction of dynamic detection is established.The PPIC Algorithm consists of a rough correction model and a refined correction model.The rough correction model extracts the key point mileage of the plane curve by using the characteristics of the cant data at the principal point of the plane curve,and corrects the large mileage deviation by template matching.On the basis of rough correction,the refined correction model uses the optimized dynamic time warping(DTW)algorithm to eliminate the mileage residual deviation,and then adopts the cycle iterative correction method of the optimal waveform similarity section to realize the accurate correction of mileage deviation.Aiming at the problem that different mileage deviation correction methods have different results on the same data,a scientific and reasonable evaluation method is proposed to evaluate the effect of the correction algorithm,which comprehensively considers the mean difference,measurement uncertainty and the correlation coefficient of the track alignment and profile.In order to verify the effectiveness and practicability of the algorithm,experiment uses PPIC algorithm and the existing most stable Segment Similar Waveform Matching(SSWM)algorithm to process simulated data and measured data of high-speed railway.The results show that after the correction of PPIC algorithm,the waveform is more aligned,the mean difference and measurement uncertainty of dynamic and static detection irregularity become smaller,the correlation coefficient of dynamic and static detection data are significantly improved,and the correction accuracy of PPIC algorithm is better than that of SSWM.The mileage deviation correction accuracy of PPIC algorithm is better than a static detection interval(0.125m),which can effectively correct the mileage deviation of dynamic detection data.