Font Size: a A A

Research On High-precision Full-section Profile Measurement And Stitching Reconstruction Algorithm For Rails

Posted on:2021-02-09Degree:MasterType:Thesis
Country:ChinaCandidate:Y J BaoFull Text:PDF
GTID:2392330647967267Subject:Intelligent perception and control
Abstract/Summary:PDF Full Text Request
The rapid development of high-speed railway has greatly promoted China's economic growth,facilitating people's travel and cargo transportation.At the same time,the quality requirements for rails are also more stringent.At present,China's welding rail bases mainly rely on manual inspection for rail profile detection,which not only requires more labor,but also has low measurement accuracy,long time-consuming and inconvenient data storage,which seriously hinders the rapid development of high-speed railway.With the development of laser detection technology,non-contact profile detection has become an important means of pre-weld rail profile detection technology.Therefore,this paper uses a high-precision laser profiler to develop a high-precision,high-efficiency,high-automatic rail profile detection system to meet the needs of pre-weld inspection in China.In this paper,based on the direct laser triangulation method,four laser profilometers are used to simultaneously acquire the contour data of the top,bottom,left and right sides of the rail at the same time,first extract and segment the rail waist feature points,and then pass the Newton iterative algorithm.The center of the arc at the waist of the rail is obtained and the feature points are matched with the reference coordinates to obtain the transformation matrix.The coordinates of the four laser profilometers are unified,and the precise stitching reconstruction of the rail profile is completed.In order to achieve high-precision measurement,the data is preprocessed by the upper computer,and the anomaly point is removed by the chord height difference method of adaptive threshold,which avoids the influence of iron filings on the rail surface;the double orthogonal wavelet method is used to eliminate the abnormal point.The data is denoised,so that the denoised data more accurately reflects the rail profile;finally,the data is linearly fitted by the improved minimum absolute residual method.In order to improve the function of the system,the distortion measurement model was built by the coplanar method according to the manual measurement method.The flatness measurement model was built by the stringmeasurement method,and the verticality measurement model was built by the dihedral method to measure the rail comprehensively.The system is tested on the basis of repeatability accuracy and accuracy.The limit error of repeatability test is less than 0.05 mm,the standard deviation is less than 0.015mm;the accuracy error of comparison test is between 0.000?0.015 mm,which meets the requirements of the Ministry of Railways.The system can accurately and reliably detect the rail profile and characteristic parameters.
Keywords/Search Tags:rail detection, laser profiler, feature point matching, iterative algorithms, data filtering
PDF Full Text Request
Related items