Research On Extraction Of Retest Elements Of Existing Railway Based On Vehicle LiDAR Technology

Railway has always been an indispensable transportation facility for national development.At the same time,it is also one of the main methods for resource-based and environment-friendly transportation.To build railways vigorously has become the consensus of all domestic parties.The railway is an important link for the development of the national economy.In recent years,the development of the railway in our country has been particularly evident.It has shown rapid development,especially in high-speed railways.The massive development of the railway has put forward higher requirements for the maintenance of railways and maintenance surveys.Therefore,it is particularly important to find a set of technical methods suitable for railway retesting.The on-board LiDAR measurement technology can acquire the spatial geometric information of the existing railway quickly,in real time and efficiently,and can replace the traditional technology in the railway retesting.At the same time,it is also necessary to find systematic,efficient,and accurate processing methods for obtaining a large amount of point cloud data to obtain real-time information on various elements of the railway.Therefore,in this paper,using vehicle-mounted LiDAR measurement technology,the following researches have been conducted for the purpose of fast and accurate acquisition of rail vertices and railway linear elements:(1)Describes the measurement principle of vehicle-mounted LiDAR system,the existing railway re-measurement method based on vehicle-mounted LiDAR technology,and establishes a base station control network and a target control network to improve the accuracy of obtaining initial point cloud data to meet the retesting accuracy requirements.(2)Summary of vehicle LiDAR point cloud data storage methods and reading methods,basic methods and theories of point cloud data processing.First,the filtering method of point cloud is studied,and then a filtering method based on improved least squares moving surface fitting is proposed.(3)On the basis of filtering,according to the three-dimensional coordinate information and reflection intensity of LiDAR point cloud,and extract the orbital vertices under the constraints of elevation,density and orbital geometry,and use the region growing method to extract the left and right orbit vertices.Separated.(4)Random sampling method is used to compress the vertices of the orbit,and then the chord slope graph is filtered using the fixed-length string slope algorithm combined with the Rlowess algorithm to accurately segment the vertices of the orbit.Orthogonal least-squares fitting is performed on the points in the corresponding segment to obtain the parameters of the straight line and the circular curve,and the parameters of the corresponding easing curve,the length and the mileage of the main point of the curve are calculated in combination with the parameters obtained by the fitting.