Railway existing line re-measurement method based on unmanned aerial vehicle non-contact measurement

Abstract

The application provides a railway existing line re-measurement method based on unmanned aerial vehicle non-contact measurement, aiming at solving the safety and efficiency problems in traditional re-measurement. The method adopts a multi-rotor unmanned aerial vehicle, carries high-precision LiDAR and high-definition camera, flies along the two sides of the railway line, does not need personnel to go on the line, reduces safety risks, and does not interfere with railway operation; the multi-rotor unmanned aerial vehicle quickly collects large-range, high-precision and high-density point cloud and image data, improves the efficiency and data quality of re-measurement work. The application is suitable for railway line reconstruction and maintenance projects, can comprehensively measure without affecting railway operation, uses the collected data for automatic or interactive processing, accurately extracts track center line, mileage, roadbed section and terrain elements; the method is not limited by sky window time, realizes all-weather operation, and effectively saves resources.

Description

Technical Field

[0001] This invention relates to the field of railway surveying technology, specifically to a method for re-surveying existing railway lines based on non-contact measurement using unmanned aerial vehicles (UAVs). Background Technology

[0002] As a vital national infrastructure, railways serve as the main artery of national economic development, playing a pivotal role in the transportation industry. With the rapid development of railway construction in my country, especially the large-scale construction of high-speed railways, railway operating speeds and frequencies are increasing, placing higher demands on train safety and smooth operation. To meet the needs of maintenance, repair, and track upgrading of existing railways, it is necessary to re-survey existing lines to obtain the geometric state of the tracks and the stability of track-related structures. Currently, traditional existing line re-surveying mainly relies on manual on-line measurement, using methods such as GNSS RTK + total station + electronic level measurement, and static (dynamic) track inspection trolley measurement. This work is limited by track maintenance windows, involves a large workload, and poses safety hazards for personnel working on the tracks. With the significant increase in railway speed, the high speed and density of train operations, and the shortening of track maintenance windows, traditional existing line measurement methods are insufficient to meet the needs of large-scale, high-quality railway development.

[0003] To address the above issues, the invention patent "Method for Railway Line Operation and Maintenance Measurement Based on Vehicle-Mounted LiDAR Technology" [Publication No.: CN105844995A] proposes a method for railway operation and maintenance measurement based on vehicle-mounted LiDAR technology. This method involves placing a 3D laser scanner on a train to quickly acquire high-density point cloud data of the railway track and its surroundings. The accuracy of the point cloud is improved by deploying target control points along the line, and finally, line measurement is performed based on the point cloud data. Compared to traditional measurement methods, this invention patent offers some efficiency improvements, but it suffers from three problems: First, both vehicle-mounted scanning data acquisition and target measurement must be performed online within the designated maintenance window, leading to difficulties such as the difficulty of obtaining maintenance windows, limited operating time, and significant personal safety risks. Second, vehicle-mounted LiDAR equipment is limited by overhead contact line clearance, typically installed relatively low, resulting in a limited field of view. This creates blind spots in areas such as high roadbeds, cuttings, bridges, and station platforms, leading to inconsistencies in the acquired point cloud data. Finally, vehicle-mounted LiDAR operates on railway tracks, and sensors and electrical equipment along the railway line cause electromagnetic interference to the laser scanner, inertial navigation system, and GNSS equipment, affecting the accuracy of data acquisition. The invention patent "Method for Extracting Railway Track and Centerline Based on Aerial Remote Sensing Imagery" [Publication No.: CN114187537A] uses imagery to extract existing line data, but image distortion and the horseshoe shape of the track result in insufficient accuracy in extracting non-angular information. Related research is still underway, and the technology is not yet mature. Summary of the Invention

[0004] This application provides a method and system for re-surveying existing railway lines based on non-contact measurement using unmanned aerial vehicles (UAVs), in order to solve the problems of time constraints, personal safety risks, and data accuracy in traditional railway re-surveying.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows:

[0006] A method for re-surveying existing railway lines based on unmanned aerial vehicle (UAV) non-contact measurement, the method comprising:

[0007] Preparation for the operation: Collect data on the existing railway centerline, terrain, and control network; select the UAV platform and airborne radar equipment according to the survey requirements.

[0008] Control network deployment and measurement: Deploy base station control network and target control network, and measure the three-dimensional coordinates of target control points in the target control network;

[0009] Drone route planning: Using railway centerline and terrain data to plan the flight path of drone platforms;

[0010] Flight data acquisition: Several GNSS base stations are set up at the base station control points of the base station control network, and point cloud and image data are acquired using airborne radar equipment;

[0011] Scanning data preprocessing: The GNSS data of the airborne radar, inertial navigation data and GNSS data of the GNSS base station are fused and solved to generate a high-precision trajectory file, which in turn generates initial point cloud data and image data, and the initial point cloud data and image data are quality checked.

[0012] Point cloud refinement and correction based on target control points: The initial point cloud data is corrected using target control points to improve the absolute accuracy of the point cloud data;

[0013] Line measurement element information extraction: Based on the refined and corrected point cloud data, the measurement information is extracted, including the track centerline, mileage, roadbed cross section and topography. The survey of the line and ancillary facilities is completed with reference to the image data.

[0014] Verification and organization of remeasurement results: Verify the extracted measurement information and compile a remeasurement table.

[0015] Furthermore, the drone platform is a multi-rotor drone platform with PPK function, equipped with high-precision LiDAR and high-definition camera, and has a flight time of no less than 60 minutes.

[0016] The laser spot frequency of the airborne radar equipment is not less than 1000KHz, and the scanning field of view is not less than 50 degrees.

[0017] Furthermore, the base station control network uses the existing CPI control network and CPI I control network of the railway; if the density and accuracy of the existing control network are insufficient, a new base station control network is built on both sides of the railway to supplement it.

[0018] Furthermore, the deployment and measurement of the target control network specifically includes:

[0019] A rectangular target plate is manufactured, wherein the target plate is printed with alternating black and white colors, and the printed surface is covered with a frosted film.

[0020] Target plates are placed in pairs on both sides of the railway with the same target spacing. The diagonal direction of the target plate is at a 90° angle to the railway to improve the laser scanning effect and the recognition rate of the target control point at the center of the target plate.

[0021] GNSS rapid static measurement and leveling round-trip measurement were performed on the target plate to determine its horizontal position and elevation.

[0022] The planar position and elevation data of the target plate are calculated and adjusted to obtain the three-dimensional coordinates of the target control points.

[0023] Furthermore, the flight path of the drone platform is planned using railway centerline and terrain data, specifically including the following steps:

[0024] Flight route design parameter planning: Based on the scanning field of view (fov) of the airborne radar equipment, calculate the flight route spacing to ensure that the flight route spacing covers the part outside the railway land boundary. At the same time, set the lateral overlap P to be greater than 30% to ensure the point cloud density.

[0025] The spacing between flight paths, D, is shown in the following formula:

[0026]

[0027] The flight altitude H is calculated using the flight path spacing D, as shown in the following formula:

[0028]

[0029] Once the flight altitude H is determined, the point frequency and linear velocity parameters of the airborne radar equipment, as well as the flight speed of the UAV platform, are determined based on the density of the point cloud.

[0030] Flight strip partitioning: Based on the length of the railway line and the flight time of the UAV, the entire measurement area is divided into several flight strip partitions to ensure that each partition can be completed in a single flight, and that there are overlapping areas between adjacent partitions for data stitching.

[0031] Flight path design and terrain-following flight: Design terrain-following flight paths based on terrain data to ensure uniform density of ground point clouds;

[0032] Conduct safety assessments and export data for the planned flight routes.

[0033] Furthermore, several GNSS base stations are set up at the base station control point, with a distance of 5 to 8 km between adjacent GNSS base stations.

[0034] Furthermore, point cloud refinement and correction based on target control points: The initial point cloud data is corrected using target control points to improve the absolute accuracy of the point cloud data. Specifically, this includes:

[0035] Target identification and preliminary correction: The target position is identified by fusing initial POS data and point cloud data; if the target center is not directly scanned in the point cloud data, the coordinates of the center point are calculated based on the target outline, and this position is associated with the most recent point cloud data and assigned a corresponding GPS timestamp.

[0036] POS trajectory line correction: Using the target's position in the point cloud and the known coordinates of the control network, the POS trajectory line is corrected in reverse to ensure that the trajectory line is consistent with the target control network coordinates;

[0037] High-precision point cloud data generation: The corrected trajectory line is re-fused and solved with the scanned data to generate point cloud data in the WGS84 coordinate system;

[0038] Point cloud data coordinate transformation: The point cloud data is transformed from the WGS84 coordinate system to the engineering independent coordinate system to obtain the final high-precision point cloud data.

[0039] Furthermore, in the extraction of line measurement element information, the extracted centerline measurement information includes:

[0040] Based on the shape and elevation information of the track point cloud, a starting point is specified on any rail, and the rail surface is automatically tracked to obtain a rough outline of the track line.

[0041] Using the rail lines as a reference, the distance between the left and right rails and the reference line is set, the center line extraction spacing is set, the rail section is cut, the section point cloud is registered with the standard rail section size, and when the matching error is minimal, the center of the standard rail tread is taken as the rail vertex, and the elevation is taken as the elevation of the nearest point cloud of the rail vertex, thus obtaining the three-dimensional coordinates of the center of the left and right rails.

[0042] The centerline plane is the average of the centers of the left and right rails, and the centerline elevation is the minimum of the left and right rail elevations, thus obtaining the centerline coordinates of the line.

[0043] Furthermore, in the extraction of route measurement element information, the mileage measurement is based on the centerline measurement results. Starting from the permanent structures in the point cloud, the mileage is calculated along the centerline to obtain any mileage of the route.

[0044] Furthermore, in the verification and organization of the re-measurement results, gross errors or mistakes are eliminated by calculating the track gauge and checking the elevation difference.

[0045] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0046] 1. The method of this invention, employing unmanned aerial vehicle (UAV) non-contact measurement, solves the problem of traditional railway line resurvey. This method avoids direct personnel work on the track, significantly reducing safety hazards. Because it is not limited by train running time (maintenance windows), it can be used in all weather conditions, greatly improving surveying efficiency. Especially given the current context of continuously increasing train speeds, traditional on-track measurement methods are becoming increasingly difficult to implement, while this invention provides a superior technical solution.

[0047] 2. The method of this invention offers a wide data acquisition range and high cost-effectiveness. Compared with traditional re-survey methods, this invention can cover a wide area of ​​approximately 200 meters on both sides of the railway line, obtaining more comprehensive information. This method can acquire detailed data over a large area in a single measurement, which can be used multiple times by different professions and needs, effectively reducing the necessity of repeated measurements. Therefore, this invention not only improves the efficiency of data acquisition but also significantly saves manpower and material costs, making it particularly suitable for surveying needs that change frequently. Attached Figure Description

[0048] Figure 1 A flowchart illustrating a method for re-surveying existing railway lines based on unmanned aerial vehicle (UAV) non-contact measurement, as provided in one embodiment of the present invention;

[0049] Figure 2 A design drawing and a physical image of a target plate provided for one embodiment of the present invention;

[0050] Figure 3 This is a schematic diagram of the target plate layout along the line according to an embodiment of the present invention;

[0051] Figure 4 This is a schematic diagram of unmanned aerial vehicle (UAV) route planning parameters provided in one embodiment of the present invention;

[0052] Figure 5 This is a schematic diagram of unmanned aerial vehicle (UAV) route planning provided in one embodiment of the present invention;

[0053] Figure 6 This is a schematic diagram of matching a standard rail cross section with a rail point cloud, provided as an embodiment of the present invention.

[0054] Figure 7 A schematic diagram of existing railway airborne point cloud results provided in one embodiment of the present invention:

[0055] Figure 8 A schematic diagram of centerline extraction for line resurvey elements provided in an embodiment of the present invention:

[0056] Figure 9This is a schematic diagram of the extraction of cross-sectional elements for line re-measurement provided in one embodiment of the present invention. Detailed Implementation

[0057] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings. Similar elements in different embodiments are referred to by associated similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of this application. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, certain operations related to this application are not shown or described in the specification. This is to avoid obscuring the core parts of this application with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0058] Furthermore, the features, operations, or characteristics described in the specification can be combined in any suitable manner to form various embodiments. At the same time, the steps or actions in the method description can be rearranged or adjusted in a manner obvious to those skilled in the art. Therefore, the various orders in the specification and drawings are only for the clear description of a particular embodiment and do not imply a necessary order, unless otherwise stated that a particular order must be followed.

[0059] Example 1:

[0060] See Figure 1 A method for re-surveying existing railway lines based on non-contact measurement using unmanned aerial vehicles (UAVs) specifically includes the following steps:

[0061] S1. Work Preparation. Collect data on the existing railway centerline, topographic maps, and control network information. Based on the specific needs of the engineering survey, select a suitable multi-rotor UAV platform and airborne radar equipment.

[0062] In S1, the laser spot frequency of the radar equipment is not less than 1000KHz, the scanning field of view is not less than 50 degrees, the rotary-wing UAV platform supports PPK post-differential function, can perform terrain-following flight, and has an endurance of not less than 60 minutes.

[0063] S2. Control Network Deployment and Measurement. The control network includes a base station control network and a target control network. The base station control network is used for the establishment of UAV flight base stations to meet the requirements of subsequent PPK calculation. The target control network is used to correct the point cloud and improve the absolute accuracy of the results. If the density and accuracy of the existing control network points meet the requirements for base station setup, a separate base station control network is not required. The target control network is deployed manually, located along the railway land boundary on both sides, i.e., outside the protective net. The point markers are prefabricated KT boards. The three-dimensional coordinates of the target control points are obtained using GNSS rapid static and leveling survey methods.

[0064] The existing CPI and CPI I control networks of the railway can be used as a priority for the base station control network. When the density and accuracy of the control network points do not meet the survey requirements, a new base station control network can be built along both sides of the railway line.

[0065] In S2, the specific deployment methods for the target control network include:

[0066] S201. Design the target board style. The target board should be rectangular, exceeding 50 cm in size, and made of KT board no thinner than 5 mm. The target board should be printed with alternating black and white colors to form target control points. The printed surface of the target board should be covered with a frosted film to enhance the laser reflection effect. See [link to documentation]. Figure 2 , Figure 2 (a) shows the design style of the target plate. Figure 2 (b) is a picture of the actual target plate.

[0067] S202. Utilize network satellite maps to pre-select target control points, and deploy them in pairs along both sides of the railway line outside the protective netting at a target spacing of 300 meters; see [link / reference]. Figure 3 During on-site deployment, ensure that the angle between the target plate and the railway line is close to 90 degrees to optimize the vertical relationship between the laser scanning direction and the target's reflective stripes, thereby increasing the number of point clouds on the target, reducing scanning omissions, and facilitating the identification of the target center later.

[0068] S203. Perform rapid static GNSS measurements on the target's planar position, with an observation time of approximately 25 to 30 minutes, and connect with at least two base station control points. Use a leveling reciprocating measurement method to determine the target's elevation.

[0069] S204. Calculate and adjust the target plane and elevation measurement data to obtain the final target control point data.

[0070] S3. Drone Flight Path Planning. Utilize existing railway lines and terrain data to plan drone flight paths using either built-in or third-party flight path planning software. Use the centerline of the railway line as a reference, with the land boundaries on both sides as buffer zones. Lay out round-trip flight paths along both sides of the railway line outside the buffer zones, and set flight parameters reasonably while ensuring that the flight path overlap is better than 30%. Consider the terrain undulations and design terrain-following flight paths.

[0071] S3 specifically includes the following steps:

[0072] S301. First, plan the flight path design parameters. After selecting the radar equipment, the scanning field of view (fov) of the equipment can be determined. The railway boundary distance S is the distance between the two protective nets. Considering flight safety, the flight path is laid outside the railway boundary; therefore, the flight path spacing D > S. To ensure the density of the rail point cloud, the lateral overlap P is greater than 30%. See [reference needed]. Figure 4 .

[0073] Use the formula:

[0074]

[0075] The relative flight altitude H can be calculated:

[0076]

[0077] After determining the flight altitude H, based on the density of the point cloud obtained (in this embodiment, the density of the railway resurvey point cloud is not less than 2000 pt / ㎡), the point frequency and linear velocity of the radar parameters, as well as the flight speed of the UAV platform, are determined, with the flight speed ≤8m / s.

[0078] S302. Based on the total length of the railway line, the measurement area is divided into multiple flight zones to ensure that each zone can be completed in a single flight. The overlapping area between adjacent zones is designed. In this embodiment, the distance between at least two pairs of target control points is about 600m to ensure the integrity and continuity of the data.

[0079] S303. Utilize the flight path design software built into the UAV platform or third-party flight path design software, such as FPS Smart software, to design flight paths. The flight path should be designed with reference to the DEM of the survey area, employing terrain-following elevation variation to ensure uniform ground point cloud density. For publicly available DEM data sources, such as SRTM, which have low resolution, a low-resolution orthophoto flight of the survey area can be performed first to obtain higher-resolution real terrain data before flight path design. Set zigzag flight paths at the start and end of the flight path segment for inertial navigation calibration to improve radar data accuracy and stability. See [link to relevant documentation]. Figure 5 .

[0080] Step 304: After the route design is completed, export the KML file format, load the route into the map, and evaluate the route safety.

[0081] S4. Flight Data Acquisition. Data acquisition is achieved through POS-assisted aerial photography with ground base stations. Multiple GNSS base stations are set up at offline base station control points. Based on the zoned range of the flight strip design, point cloud and image data are acquired using radar equipment mounted on UAVs.

[0082] In S4, ground GNSS base stations are set up at base station control points, with multiple GNSS base stations deployed in each survey area, and the distance between adjacent GNSS base stations is 5 to 8 km.

[0083] Before acquiring flight data in S4, the target control points deployed in S2 must be checked to ensure that the target boards are not lost or covered, and that the angle and orientation are correct. During the formal flight operation, the base station should be turned on for observation half an hour in advance, and the base station sampling frequency should be set to 1Hz. After the flight is completed, the base station should be turned off for half an hour. After the flight is completed, the POS data and scan data should be copied in a timely manner, and the integrity and size of the data should be checked.

[0084] S5. Scanning Data Preprocessing. Using Inertial Explorer (IE) software, airborne GNSS data, inertial navigation data, and ground base station GNSS data are fused and processed to obtain high-precision trajectory files. Initial point cloud data and imagery are generated based on these trajectory files, and the quality of the acquired data is checked.

[0085] S6. Point cloud refinement and correction based on target control points. The accuracy of the preprocessed point cloud data is based on the PPK differential calculation, which is insufficient to meet the accuracy requirement of better than 2cm for existing line re-measurement. The POS trajectory line is corrected by using target control points deployed along both sides of the line, thereby improving the absolute accuracy of the final point cloud.

[0086] In S6, point cloud refinement and correction based on the target control network refers to correcting the solved POS trajectory line using the target control points to obtain a high-precision trajectory line, which is then fused to obtain high-precision point cloud data. Specifically, this includes the following steps:

[0087] S601. Using the POS fusion point cloud data from the initial solution, the target position is identified from the point cloud. The center position of the target may not necessarily have a scanning point. The coordinates of the center point can be obtained by fitting the target contour. At the same time, the GPS timestamp information of the point cloud closest to the center point is assigned to the fitted point.

[0088] S602. At this time, the target has two sets of coordinates: point cloud coordinates and control network coordinates. The POS trajectory line is corrected in reverse by using time information to obtain a high-precision POS trajectory line that is consistent with the target control network coordinates. The POS trajectory line corrected in segments needs to share two pairs of target points.

[0089] S603. The corrected trajectory line is fused with the scanned data again to obtain the point cloud results in the WGS84 coordinate system.

[0090] S604. Using the two sets of coordinates from the base station control network covering the survey area, the seven parameters are calculated from the WGS84 geodetic height results and the engineering independent coordinate system leveling height results. The point cloud results are then transformed to the engineering independent coordinate system to obtain the final point cloud results. (See [link to documentation]). Figure 7 .

[0091] S7. Extraction of Line Measurement Elements. Based on point cloud data, track centerline extraction, mileage measurement, roadbed cross-section and longitudinal section measurement, and topographic surveying are performed. Referring to the acquired image data, the survey of the line and its ancillary facilities is completed.

[0092] S7 is used for extracting elements for line resurvey. Based on the final result point cloud, it can perform centerline measurement, mileage measurement, cross-sectional and longitudinal section measurement, etc., completely moving the "field site" to the "computer", greatly improving work efficiency.

[0093] Among them, see Figure 8 The specific steps for midline extraction are as follows:

[0094] First, based on the shape and elevation information of the track point cloud, a starting point is manually designated on any rail, and the software will automatically track the rail surface to obtain a rough outline of the track line.

[0095] Using the rail lines as a reference, set the distances between the left and right rails and the reference lines, for example, 0 for the left rail and 1.5 for the right rail. Set the centerline extraction spacing. The software algorithm will first perform rail cross-section cutting, and then register the cross-section point cloud with the standard rail cross-section dimensions. See [link / reference]. Figure 6 When the matching error is minimized, the center of the standard rail tread is taken as the rail vertex, and the elevation is taken as the elevation of the nearest point cloud of the rail vertex. This will give us the three-dimensional coordinates of the center of the left and right rails.

[0096] The centerline plane is the average of the centers of the left and right rails. The centerline elevation is the minimum of the left and right rail elevations. In curved sections, the superelevation of the outer rail is the inner rail elevation, thus obtaining the centerline coordinates of the line.

[0097] See Figure 9 Mileage measurement is based on the centerline measurement results. Starting from permanent structures such as bridges, culverts, and stations in the point cloud, the mileage of any route can be automatically calculated along the centerline.

[0098] S8. Verify and organize the extracted retest results, and compile the retest tables.

[0099] In S8, the main focus is on centerline inspection, which can eliminate gross errors or mistakes by calculating track gauge and checking elevation differences.

[0100] Example 2:

[0101] In this embodiment, a railway speed-up and renovation project located in Northwest China successfully adopted the patented technology to complete the resurvey of the existing railway line. A CPI control network already exists along the railway line, with control points spaced 4 kilometers apart, directly serving as the base station control network required by this patent. Target control points are deployed in pairs every 300 meters on both sides of the railway line, and the target plates are designed to meet… Figure 2 The pattern shown measures 50 cm by 50 cm. In key areas such as stations, the deployment of target control points has been increased.

[0102] For aerial mapping, a hexacopter UAV equipped with a Riegl VUX240 radar was selected. The radar's point frequency was set to 1800 kHz, and its scanning linear velocity was 350 lines per second. The UAV's flight altitude was set to 90 meters, and its flight speed to 6 meters per second. This configuration resulted in an average point cloud density of 3000 points per square meter. After precise target correction and coordinate transformation, the absolute accuracy of the point cloud data was verified, with a planar mean square error of 0.012 meters and an elevation mean square error of 0.008 meters, meeting the accuracy requirements for resurveying existing lines.

[0103] Using this high-precision point cloud data, the extraction of key elements for the track resurvey was completed. Specifically, the track centerline data, through horizontal and vertical profile fitting by the design team, was largely consistent with the engineering log data, meeting the design accuracy requirements. Other resurvey results also passed field sampling inspections and all met relevant specifications.

[0104] Those skilled in the art will understand that all or part of the functions of the various methods in the above embodiments can be implemented by hardware or by computer programs. When all or part of the functions in the above embodiments are implemented by computer programs, the program can be stored in a computer-readable storage medium, which may include: read-only memory, random access memory, disk, optical disk, hard disk, etc., and the program is executed by a computer to achieve the above functions. For example, the program can be stored in the memory of a device, and when the program in the memory is executed by the processor, all or part of the above functions can be achieved. In addition, when all or part of the functions in the above embodiments are implemented by computer programs, the program can also be stored in a server, another computer, disk, optical disk, flash drive, or external hard drive, etc., and can be downloaded or copied to the memory of a local device, or the system of the local device can be updated. When the program in the memory is executed by the processor, all or part of the functions in the above embodiments can be achieved.

[0105] This invention proposes a non-contact railway line resurvey method based on unmanned aerial vehicle (UAV) measurement, aiming to overcome the difficulties and limitations encountered in traditional manual track surveying. By employing a multi-rotor UAV equipped with LiDAR and a camera, this method can rapidly acquire extensive, high-precision, and high-density point cloud data and high-resolution imagery along both sides of the railway line without requiring personnel to be on the track. This technical solution significantly improves the efficiency and safety of railway resurveys, while expanding the scope of data acquisition and ensuring the comprehensiveness of information.

[0106] Based on these point cloud and image data, field personnel can extract line elements automatically or interactively, meeting the professional needs of various stages of railway resurveying and line reconstruction. Since this method requires no on-track work and is not limited by railway operating hours, it significantly improves operational efficiency and safety, providing a new and efficient technical means for resurveying existing railway lines.

[0107] The above examples illustrate the present invention only to aid in understanding it and are not intended to limit the scope of the invention. Those skilled in the art can make various simple deductions, modifications, or substitutions based on the ideas of this invention.

Claims

1. A method for re-surveying existing railway lines based on unmanned aerial vehicle (UAV) non-contact measurement, characterized in that, The method includes: Preparation for the operation: Collect data on the existing railway centerline, terrain, and control network; select the UAV platform and airborne radar equipment according to the survey requirements. Control network deployment and measurement: Deploy base station control network and target control network, and measure the three-dimensional coordinates of target control points in the target control network; Drone route planning: Using railway centerline and terrain data to plan the flight path of drone platforms; Flight data acquisition: Several GNSS base stations are set up at the base station control points of the base station control network, and point cloud and image data are acquired using airborne radar equipment; Scanning data preprocessing: The GNSS data of the airborne radar, inertial navigation data and GNSS data of the GNSS base station are fused and solved to generate a high-precision trajectory file, which in turn generates initial point cloud data and image data, and the initial point cloud data and image data are quality checked. Point cloud refinement and correction based on target control points: The initial point cloud data is corrected using target control points to improve the absolute accuracy of the point cloud data; Line measurement element information extraction: Based on the refined and corrected point cloud data, the measurement information is extracted, including the track centerline, mileage, roadbed cross section and topography. The survey of the line and ancillary facilities is completed with reference to the image data. Verification and organization of remeasurement results: Verify the extracted measurement information, organize it, and compile a remeasurement table; Planning flight paths for drone platforms using railway centerlines and terrain data includes the following steps: Flight route design parameter planning: Based on the scanning field of view (fov) of the airborne radar equipment, calculate the flight route spacing to ensure that the flight route spacing covers the part outside the railway land boundary. At the same time, set the lateral overlap P to be greater than 30% to ensure the point cloud density. The spacing between flight paths, D, is shown in the following formula: Calculate flight altitude H using the route spacing D: Once the flight altitude H is determined, the point frequency and linear velocity parameters of the airborne radar equipment, as well as the flight speed of the UAV platform, are determined based on the density of the point cloud. Flight strip partitioning: Based on the length of the railway line and the flight time of the UAV, the entire measurement area is divided into several flight strip partitions to ensure that each partition can be completed in a single flight, and that there are overlapping areas between adjacent partitions for data stitching. Flight path design and terrain-following flight: Design terrain-following flight paths based on terrain data to ensure uniform density of ground point clouds; Conduct safety assessments and export data for the planned flight routes; Point cloud refinement and correction based on target control points: This involves correcting the initial point cloud data using target control points to improve the absolute accuracy of the point cloud data. Specifically, this includes: Target identification and preliminary correction: The target position is identified by fusing initial POS data and point cloud data; if the target center is not directly scanned in the point cloud data, the coordinates of the center point are calculated based on the target outline, and this position is associated with the most recent point cloud data and assigned a corresponding GPS timestamp. POS trajectory line correction: Using the target's position in the point cloud and the known coordinates of the control network, the POS trajectory line is corrected in reverse to ensure that the trajectory line is consistent with the target control network coordinates; High-precision point cloud data generation: The corrected trajectory line is re-fused and solved with the scanned data to generate point cloud data in the WGS84 coordinate system; Point cloud data coordinate transformation: Transform the point cloud data from the WGS84 coordinate system to the engineering independent coordinate system to obtain the final high-precision point cloud data; In the extraction of line measurement element information, the extracted centerline measurement information includes: Based on the shape and elevation information of the track point cloud, a starting point is specified on any rail, and the rail surface is automatically tracked to obtain a rough outline of the track line. Using the rail lines as a reference, the distance between the left and right rails and the reference line is set, the center line extraction spacing is set, the rail section is cut, the section point cloud is registered with the standard rail section size, and when the matching error is minimal, the center of the standard rail tread is taken as the rail vertex, and the elevation is taken as the elevation of the nearest point cloud of the rail vertex, thus obtaining the three-dimensional coordinates of the center of the left and right rails. The centerline plane is the average of the centers of the left and right rails, and the centerline elevation is the minimum of the left and right rail elevations, thus obtaining the centerline coordinates of the line.

2. The method for re-surveying existing railway lines based on UAV non-contact measurement according to claim 1, characterized in that, The drone platform is a multi-rotor drone platform, equipped with a high-precision LiDAR and a high-definition camera, with a flight time of no less than 60 minutes. The laser spot frequency of the airborne radar equipment is not less than 1000kHz, and the scanning field of view is not less than 50 degrees.

3. The method for re-surveying existing railway lines based on UAV non-contact measurement according to claim 1, characterized in that, The base station control network uses the railway CPI control network; if the density and accuracy of the CPI control network are insufficient, new base station control networks will be built on both sides of the railway to supplement it.

4. The method for re-surveying existing railway lines based on UAV non-contact measurement according to claim 1, characterized in that, The base station control network uses the railway CPII control network; if the density and accuracy of the CPII control network are insufficient, new base station control networks will be built on both sides of the railway to supplement it.

5. The method for re-surveying existing railway lines based on UAV non-contact measurement according to claim 1, characterized in that, The deployment and measurement of the target control network specifically include: A rectangular target plate is manufactured, wherein the target plate is printed with alternating black and white colors, and the printed surface is covered with a frosted film. Target plates are placed in pairs on both sides of the railway with the same target spacing. The angle between the diagonal direction of the target plate and the direction of the railway line is 90°, which improves the laser scanning effect and the recognition rate of the target control point at the center of the target plate. GNSS rapid static measurement and leveling round-trip measurement were performed on the target plate to determine its horizontal position and elevation. The planar position and elevation data of the target plate are calculated and adjusted to obtain the three-dimensional coordinates of the target control points.

6. The method for re-surveying existing railway lines based on UAV non-contact measurement according to claim 1, characterized in that, Several GNSS base stations are set up at the base station control point, with a distance of 5 to 8 km between adjacent GNSS base stations.

7. The method for re-surveying existing railway lines based on UAV non-contact measurement according to claim 1, characterized in that, In the extraction of route measurement element information, mileage measurement is based on the centerline measurement results. Starting from the permanent structures in the point cloud, the mileage of the route is calculated along the centerline to obtain any mileage of the route.

8. The method for re-surveying existing railway lines based on UAV non-contact measurement according to claim 1, characterized in that, In the verification and organization of the results of the re-measurement elements, gross errors or mistakes are eliminated by calculating the track gauge and checking the elevation difference.

Images