A novel data processing method for sectional rail measurements to detect track irregularities in high-speed railways

Static track inspection methods record the position and direction data through track surveys completed by trolleys that survey a track based on a track control network. The track inspection points must be measured in sections to decrease the likelihood of measurement errors. Certain track inspection points will be measured twice by adjacent stations, and different results will be obtained because of the measurement errors. To improve the regularity of the track, inspection point data from the sectional measurements must be processed, and differences in the results for the same inspection points must be eliminated. Therefore, this paper proposes a novel method referred to as the regularity for processing sectional measurement data (RPSMD) method, which overcomes the disadvantages of the currently available methods of processing inspection points via track-surveying trolleys and also improves the processing of nonoverlapping and overlapping inspection points. The adjustment criterion states that the difference value for the same adjusted overlapping points should be zero, and this criterion can be used to obtain the adjustment equation for each station. Using the adjustment equation, all station points can be corrected and the total regularity of the track points can be guaranteed. According to precisely measured ballastless tracks and their calculated three-dimensional coordinates, the RPSMD method and other available methods are verified by an experimentally designed precise mechanical device. The experimental results show that the accuracy of the nonoverlapping points adjusted by the RPSMD method is improved, and the accuracy is obviously higher than that obtained by the other available methods. Also, the accuracy of the RPSMD-adjusted overlapping points is much higher than that of the nonadjusted points.


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01662594
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Feb 8 2018 2:05PM