Contact network parameter detection algorithm and system based on three-dimensional laser point cloud
By acquiring 3D laser point cloud data and performing point cloud registration using the PCA-GICP algorithm, the problems of difficult component segmentation and inaccurate reference in catenary inspection were solved, enabling efficient and accurate detection of catenary parameters.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-19
- Publication Date
- 2026-03-27
AI Technical Summary
Existing point cloud-based contact wire inspection methods suffer from problems such as uneven point cloud density, complex structure leading to difficulties in component segmentation and inaccurate references, making it difficult to achieve reliable separation of key components such as the rail surface, contact wire, and catenary wire, thus affecting inspection accuracy and efficiency.
Three-dimensional laser point cloud data acquisition, spatial division, and PCA-GICP algorithm are used for point cloud registration. Point cloud data is acquired by a track-mounted self-propelled trolley, and coarse and fine registration are performed using the PCA-GICP algorithm. Key component feature points of the catenary are extracted, and their geometric parameters are calculated.
This improved the accuracy and automation level of overhead contact line inspection, ensuring the reliability and accuracy of measurement results and increasing inspection efficiency.
Smart Images
Figure CN121746437A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the technical field of three-dimensional laser measurement, and particularly relates to a catenary parameter detection algorithm and system for three-dimensional laser point clouds. BACKGROUND
[0002] The catenary is a key equipment for traction power supply of electrified railways, and the accuracy of geometric parameters such as the catenary height and the pull-out value directly affects the current collection quality and operation safety of the train pantograph. The traditional manual detection method is low in efficiency and high in risk, and the image-based detection method is susceptible to environmental light and weather interference and is insufficient in stability. In recent years, the three-dimensional laser scanning technology can obtain rich point cloud data, which provides a possibility for fine measurement of the catenary.
[0003] However, the existing catenary detection method based on point clouds still has obvious limitations. The point cloud density of the catenary is uneven, and the structure is complex, which leads to difficult component segmentation and inaccurate reference. The existing algorithm is insufficient in utilization of point cloud features, and the segmentation accuracy and generalization ability are limited, so it is difficult to realize reliable separation of key components such as the track surface, the contact line and the catenary cable, thereby restricting the precision and efficiency of the automatic detection of geometric parameters.
[0004] Therefore, the application provides a catenary parameter detection algorithm and system for three-dimensional laser point clouds to solve the above problems. SUMMARY
[0005] In view of the deficiencies of the prior art, the application provides a catenary parameter detection algorithm and system for three-dimensional laser point clouds, which solves the problems of difficult component segmentation and inaccurate reference caused by uneven point cloud density and complex structure of the catenary through the three-dimensional laser point cloud data acquisition, spatial division and PCA-GICP algorithm point cloud registration process.
[0006] The application is a catenary parameter detection algorithm for three-dimensional laser point clouds, which comprises the following steps: S1. Collecting three-dimensional laser point cloud data of a catenary system through a track self-propelled vehicle; S2. Dividing the three-dimensional laser point cloud data in space to form a point cloud data set, wherein the point cloud data set comprises track point clouds, contact line point clouds and catenary cable point clouds; S3. Registering the point cloud data set by using a PCA-GICP algorithm to form parameter measurement feature points, wherein the point cloud registration comprises coarse registration and fine registration, and the parameter measurement feature points comprise track feature points, contact line feature points and catenary cable feature points; S4. Calculating the spatial geometric parameters of the catenary parameter measurement feature points to form geometric parameter measurement results; S5. performing statistical analysis on the geometric parameter measurement results to output a measurement report, the measurement report including key geometric parameters of the catenary system.
[0007] The application is further provided that the track self-traveling trolley automatically travels on the railway track, the track self-traveling trolley is equipped with a laser scanner, the laser scanner acquires the three-dimensional laser point cloud data in real time by adjusting the scanning angle and the scanning speed, and the components of the three-dimensional laser point cloud data include the track, the contact wire and the catenary.
[0008] The application is further provided that the step of space division is: S21. performing preliminary screening on the three-dimensional laser point cloud data based on an elevation method to select track surface point cloud of the catenary system; S22. performing longitudinal segmentation on the track surface point cloud to form a track data set, and determining the spatial range of the contact wire and the catenary based on the track surface point cloud data; S23. taking the height of the track surface point cloud as a reference to extract the contact wire point cloud and the catenary point cloud in the spatial range.
[0009] The application is further provided that the elevation method separates the track surface point cloud from other area point clouds by calculating the height value of each point in the three-dimensional laser point cloud data, and the spatial range is determined according to the structural characteristics of the catenary system by setting the height range and the position parameters.
[0010] The application is further provided that the PCA-GICP algorithm includes a principal component analysis algorithm and a generalized iterative closest point algorithm, the principal component analysis algorithm performs the coarse registration, and the generalized iterative closest point algorithm performs the fine registration.
[0011] The application is further provided that the coarse registration roughly aligns the point cloud data set by the principal component analysis algorithm, the coarse registration calculates the covariance matrix of the point cloud data set, extracts the main eigenvalue and the eigenvector based on the covariance matrix, and rotates the point cloud data set according to the principal component direction determined by the eigenvector to obtain the coarse registration result.
[0012] The application is further provided that the fine registration is based on the coarse registration result, and the covariance matrix is iteratively rotated and the covariance matrix is translated by the generalized iterative closest point algorithm, and the catenary parameter measurement feature points are acquired in the process of the fine registration.
[0013] The application is further provided that the track feature points include left track feature points and right track feature points, the left track feature points include left track upper surface feature points and left track inner side surface feature points, the right track feature points include right track upper surface feature points and right track inner side surface feature points, the left track feature points and the right track feature points determine a track surface reference, and the calculation of the spatial geometric parameters is to calculate the track superelevation according to the track left track upper surface feature points and the right track upper surface feature points, to calculate the track gauge according to the track left track inner side feature points and the right track inner side feature points, to calculate the contact wire gauge and the pull-out value according to the contact wire feature points and the track surface reference, and to calculate the catenary gauge and the pull-out value according to the catenary feature points and the track surface reference.
[0014] The application is further provided that the statistical analysis is to calculate the standard deviation, the maximum value, the minimum value and the average value of the geometric parameters in the geometric parameter measurement results to form an analysis result, and to generate the measurement report based on the analysis result.
[0015] A catenary parameter detection system of three-dimensional laser point cloud, comprising: A track self-propelled trolley, which automatically travels along a railway track and collects three-dimensional laser point cloud data of a catenary system during the traveling; A point cloud data processing module, which pre-processes the three-dimensional laser point cloud data to form pre-processed point cloud data; A point cloud segmentation module, which segments the pre-processed point cloud data to form catenary component point clouds; A point cloud registration module, which accurately aligns the catenary component point clouds to form catenary point cloud data; A geometric parameter calculation module, which calculates key geometric parameters of the catenary according to the catenary point cloud data to form catenary parameters; A measurement result output module, which generates a measurement report according to the catenary parameters.
[0016] The application has the following beneficial effects: The application accurately extracts key component point clouds of the catenary system by spatial division and registration of three-dimensional laser point cloud data, calculates geometric parameters of the catenary system, improves the processing accuracy of point cloud data, and makes the measurement results more reliable and accurate.
[0017] The PCA-GICP algorithm is used for point cloud registration, the double steps of coarse registration and fine registration are adopted, the catenary parameter measurement feature points are efficiently obtained, and accurate geometric parameters such as track superelevation, track gauge, contact wire gauge and pull-out value are calculated, thereby improving the efficiency and automation level of catenary detection. BRIEF DESCRIPTION OF DRAWINGS
[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the description of the embodiments will be briefly introduced.
[0019] Figure 1 A point cloud data processing flowchart of the present application.
[0020] Figure 2 A schematic diagram of point cloud coordinate calibration and segmentation of the present application.
[0021] Figure 3 A result schematic diagram of rail point cloud registration based on the PCA-GIC algorithm of the present application.
[0022] Figure 4 A feature point extraction result schematic diagram based on point cloud registration of the present application.
[0023] Figure 5 A main step flowchart of parameter calculation based on feature points of the present application. DETAILED DESCRIPTION
[0024] The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. The described embodiments are only some of the embodiments of the present application, not all the embodiments.
[0025] Embodiment 1 Please refer to Figures 1-5 The present application is a catenary parameter detection algorithm for three-dimensional laser point cloud, comprising: S1. Collecting three-dimensional laser point cloud data of the catenary system by a track self-propelled trolley. The track self-propelled trolley automatically travels on the railway track, and the track self-propelled trolley is equipped with a laser scanner. The laser scanner acquires three-dimensional laser point cloud data in real time by adjusting the scanning angle and scanning speed. The components of the three-dimensional laser point cloud data include the track, the contact wire and the messenger wire.
[0026] S2. Spatially dividing the three-dimensional laser point cloud data to form a point cloud data set, which includes track point cloud, contact wire point cloud and messenger wire point cloud. The steps of spatial division are: S21. Preliminary screening of the three-dimensional laser point cloud data based on the elevation method, and selecting the track surface point cloud of the catenary system.
[0027] S22. Longitudinally cutting the track surface point cloud to form a track data set, and determining the spatial range of the contact wire and the messenger wire according to the track surface point cloud data.
[0028] S23. Taking the height of the track surface point cloud as a reference, extracting the contact wire point cloud and the messenger wire point cloud in the spatial range.
[0029] The elevation method separates the track surface point cloud from other area point clouds by calculating the height value of each point in the three-dimensional laser point cloud data. The spatial range is defined by setting the height range and position parameters according to the structural characteristics of the catenary system.
[0030] S3. The PCA-GICP algorithm is used to register the point cloud dataset to form parametric measurement feature points. Point cloud registration includes coarse registration and fine registration. Parametric measurement feature points include track feature points, contact wire feature points, and catenary feature points. The PCA-GICP algorithm includes principal component analysis (PCA) and generalized iterative nearest-point (GNP) algorithm. PCA performs coarse registration, and GNP performs fine registration. Coarse registration uses PCA to roughly align the point cloud dataset. Coarse registration involves calculating the covariance matrix of the point cloud dataset, extracting the principal eigenvalues and eigenvectors based on the covariance matrix, and rotating the point cloud dataset according to the principal component directions determined by the eigenvectors to obtain the coarse registration result. The covariance matrix satisfies: in, This represents each point cloud data point. It is the center of mass, and .
[0031] The principal component direction is the normalized result of the eigenvector corresponding to the largest eigenvalue, expressed as: in, It is the point cloud covariance matrix The eigenvector corresponding to the largest eigenvalue.
[0032] Fine registration, based on the coarse registration results, employs a generalized iterative nearest-point algorithm to iterate the rotation and translation covariance matrices, acquiring contact network parameter measurement feature points during the fine registration process. The difference between the rotation matrix and the identity matrix is measured using the Frobenius norm, calculated as follows: in, Describe the Frobenius norm. For rotation registration matrix, It is the identity matrix. Rotation matrix The elements in the i-th row and j-th column, identity matrix The i row and number j Column elements.
[0033] The local curvature of the point cloud satisfies: in, Let be the eigenvalues of the covariance matrix of the set of points in the neighborhood of a given point. It is the smallest eigenvalue.
[0034] During the fine registration process, the translation relationship between the source point cloud and the target point cloud is determined by the difference in the coordinates of their centers, and the translation vector satisfies: Where (x0,y0,z0) is the center of the source point cloud and (x1,y1,z1) is the center of the target point cloud.
[0035] In the generalized iterative nearest-point algorithm, the error vector and the joint covariance matrix satisfy the following relationship: in, , For the origin point cloud and the target point cloud, This is the error vector after point transformation. For the joint covariance matrix, For point The neighborhood point set, The neighborhood mean Let be the covariance at that point.
[0036] S4. Calculate the spatial geometric parameters of the contact wire parameter measurement feature points to form the geometric parameter measurement results. Track feature points include left rail feature points and right rail feature points. Left rail feature points include feature points on the upper surface of the left rail and feature points on the inner surface of the left rail. Right rail feature points include feature points on the upper surface of the right rail and feature points on the inner surface of the right rail. The left and right rail feature points determine the rail surface reference. The spatial geometric parameters are calculated based on the left and right rail upper surface feature points to calculate the track superelevation: in, For extremely high, and These represent the feature point heights of the left and right rails, respectively.
[0037] Calculate the track gauge based on the feature points inside the left and right rails: in, For track gauge, and These represent the x-coordinates of the feature points inside the left and right rails, respectively.
[0038] Based on the contact wire feature points and rail surface reference, calculate the contact wire guide height and pull-out value: Project the contact wire feature points vertically onto the rail surface elevation, and calculate the vertical distance as the contact wire guide height: wherein, is the contact wire guide height, is the coordinate of the contact wire feature point, and is the coordinate of the left rail feature point, respectively.
[0039] wherein, is the contact wire pull-out value, is the coordinate of the contact wire feature point, and is the coordinate of the left rail feature point, respectively.
[0040] wherein, is the super-elevation, and is the height of the feature point of the left rail, respectively.
[0041] According to the feature points of the cable and the track surface reference, the guide height and the pull-out value of the cable are calculated: wherein, is the cable guide height, is the coordinate of the cable feature point.
[0042] wherein, is the cable pull-out value, is the coordinate of the cable feature point.
[0043] S5. Perform statistical analysis on the geometric parameter measurement results to output a measurement report, the measurement report including key geometric parameters of the overhead contact system. The statistical analysis forms an analysis result by calculating the standard deviation, maximum value, minimum value and average value of the geometric parameters in the geometric parameter measurement results, and generates the measurement report based on the analysis result.
[0044] A three-dimensional laser point cloud overhead contact system parameter detection system, comprising: a track self-propelled trolley, the track self-propelled trolley automatically travels along the railway track and collects three-dimensional laser point cloud data of the overhead contact system during the travel.
[0045] a point cloud data processing module, the point cloud data processing module pre-processes the three-dimensional laser point cloud data to form pre-processed point cloud data.
[0046] a point cloud segmentation module, the point cloud segmentation module segments the pre-processed point cloud data to form a catenary component point cloud.
[0047] a point cloud registration module, the point cloud registration module accurately aligns the catenary component point cloud to form a catenary point cloud data.
[0048] a geometric parameter calculation module, the geometric parameter calculation module calculates key geometric parameters of the catenary according to the catenary point cloud data to form catenary parameters.
[0049] a measurement result output module, the measurement result output module generates a measurement report according to the catenary parameters.
[0050] Embodiment 2 Please refer to Figure 2 On the basis of Embodiment 1, three-dimensional laser point cloud data of the catenary system is collected by the track self-propelled trolley to accurately obtain geometric parameters of the track, contact wire and messenger wire, thereby providing reliable data support for subsequent point cloud processing and catenary system analysis. The specific implementation is as follows: 1. Device configuration and parameter description Device model and specifications: The track self-propelled trolley is equipped with a Leica RTC360 laser scanner, with a scanning accuracy of 1 mm and a resolution of 0.5 mm. The maximum scanning range of the laser scanner is 80 meters, and the scanning frequency is 2000 points per second.
[0051] The working frequency of the laser scanner is 2000 data points per second, and the accuracy of the point cloud data is 0.5 mm.
[0052] The working voltage of the device is 12V, and the power consumption during data collection is about 20W.
[0053] 2. Data collection process Travel speed and data collection frequency: The travel speed of the track self-propelled trolley is set to 12 kilometers per hour, i.e. 3.33 meters per second. At this speed, the laser scanner collects data at a frequency of 2000 data points per second.
[0054] In a 500-meter track section, the travel time of the track self-propelled trolley is 9 minutes. In 9 minutes, the amount of point cloud data collected by the laser scanner is: Data distribution: Track data: Among the 1080000 points, the track part including the upper surface and the inner surface of the track accounts for 60% of the data, i.e. about 648000 points.
[0055] Contact wire data: Contact wire accounts for 25% of the data, i.e. about 270000 points.
[0056] Cable data: Cable accounts for 15% of the data, i.e. about 162000 points.
[0057] The distribution ratio of the data is based on the characteristics of the actual railway track and the catenary structure. The track part usually accounts for a larger proportion of the total data, while the data of the contact wire and the cable is less.
[0058] 3. Data transmission and processing Data transmission: The point cloud data collected by the laser scanner is transmitted in real time to the data processing module through the wireless network. The amount of data transmitted per second is about 15KB, so the total amount of data transmitted in 9 minutes is: The wireless transmission technology adopts Wi-Fi5 standard to ensure the real-time and stability of the data.
[0059] 4. Data acquisition results In a 500-meter-long track section, the track self-propelled trolley collected about 1080000 point cloud data points. The data covers the three-dimensional spatial information of the track, contact wire and cable, with an accuracy of 1mm and a resolution of 0.5mm, which can accurately represent the geometric characteristics of the track, contact wire and cable.
[0060] Data details: The track data includes the upper surface and inner side surface point cloud data of the left and right tracks, the contact wire data covers the height and position of the contact wire, and the cable data contains the geometric shape and position data of the cable.
[0061] Through the above steps, the data acquisition process ensures the accurate recording of each component, and the data is transmitted in real time to the background processing module through the wireless transmission technology, providing a reliable basis for subsequent point cloud registration, geometric parameter calculation and catenary system performance analysis, ensuring the high precision and reliability of the data.
[0062] Example 3 Please refer to Figure 5 On the basis of Example 1 and Example 2, the geometric parameters of the catenary system are calculated based on three-dimensional laser point cloud data to realize accurate measurement and state evaluation of the catenary system. The specific implementation is as follows: 1. Extraction and measurement of track feature points The feature points of the left track and the right track are extracted from the three-dimensional laser point cloud data, and the specific data is as follows: Left rail upper surface feature point: The left rail upper surface feature point coordinates are x1 = 10.120 m, y1 = 50.650 m, z1 = 0.120 m, collected by a laser scanner.
[0063] Left rail inner surface feature point: The left rail inner surface feature point coordinates are x2 = 10.100 m, y2 = 50.640 m, z2 = 0.118 m.
[0064] Right rail upper surface feature point: The right rail upper surface feature point coordinates are x3 = 10.200 m, y3 = 50.650 m, z3 = 0.121 m.
[0065] Right rail inner surface feature point: The right rail inner surface feature point coordinates are x4 = 10.220 m, y4 = 50.630 m, z4 = 0.119 m.
[0066] 2. Track superelevation calculation The track superelevation is determined by the height difference between the left rail upper surface feature point and the right rail upper surface feature point. According to the laser scanning data, the z coordinate of the left rail upper surface feature point is 0.120 m, and the z coordinate of the right rail upper surface feature point is 0.121 m.
[0067] The track superelevation calculation formula is: 3. Gauge calculation The gauge is determined by the lateral distance between the left rail inner feature point and the right rail inner feature point. According to the laser scanning data, the x coordinate of the left rail inner feature point is 10.100 m, and the x coordinate of the right rail inner feature point is 10.220 m.
[0068] The gauge calculation formula is: 4. Contact wire height and pull-out value calculation Contact wire feature point: The contact wire feature point coordinates are x5 = 10.150 m, y5 = 50.650 m, z5 = 0.115 m, according to the three-dimensional laser point cloud data.
[0069] Contact wire height: The contact wire height is calculated according to the track surface reference, i.e. the average value of the z coordinates of the left rail and the right rail, and the height difference of the contact wire feature point. The calculation formula is as follows: The average value of the z coordinates of the left rail and the right rail is: The contact wire pull-out value calculation formula is: 5. Conductor cable height and pull-out value calculation The characteristic point of the messenger wire is at x6 = 10.180 m, y6 = 50.630 m, and z6 = 0.122 m.
[0070] The messenger wire gauge is calculated based on the height difference between the track reference and the messenger wire. The calculation formula is as follows: The z-coordinate of the messenger wire characteristic point is 0.122 m, and the track reference is 0.1205 m. The messenger wire gauge calculation formula is: The messenger wire pull-out value is calculated based on the horizontal coordinate of the messenger wire characteristic point and the x-coordinate of the track reference. The calculation formula is as follows: The x-coordinate of the messenger wire characteristic point is 10.180 m, and the x-coordinate of the track reference is 10.160 m. The messenger wire pull-out value calculation formula is: 6. Results summary Through the above calculations, the key geometric parameters of the overhead contact line system are obtained, and the specific data are as follows: Track superelevation = 0.001 m, gauge = 0.120 m, contact wire gauge = -0.0055 m, contact wire pull-out value = -0.010 m, messenger wire gauge = 0.0015 m, and messenger wire pull-out value = 0.020 m.
[0071] The above data ensure the accuracy of the overhead contact line parameters, providing a reliable basis for the state analysis and detection of the overhead contact line system.
[0072] In summary, the measurement results provide a scientific basis for the state analysis and evaluation of the overhead contact line, helping to improve the detection accuracy and reliability of the overhead contact line system, and providing support for subsequent maintenance and optimization decisions. Accurate calculation of geometric parameters ensures the safety and stability of the overhead contact line system during operation, reducing the risks that may be caused by parameter deviations.
[0073] The above only describes certain exemplary embodiments of the present application in a descriptive manner, and it is needless to say that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present application. Therefore, the above figures and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present application.
Claims
1. A contact wire parameter detection algorithm for three-dimensional laser point clouds, characterized in that: S1. Collect three-dimensional laser point cloud data of the overhead contact system using a track-mounted self-propelled trolley; S2. Spatially divide the three-dimensional laser point cloud data to form a point cloud dataset, which includes track point cloud, contact line point cloud and catenary point cloud; S3. The point cloud dataset is registered using the PCA-GICP algorithm to form parameter measurement feature points. The point cloud registration includes coarse registration and fine registration. The parameter measurement feature points include track feature points, contact wire feature points and catenary feature points. S4. Calculate the spatial geometric parameters of the contact wire parameter measurement feature points to form geometric parameter measurement results; S5. Perform statistical analysis on the geometric parameter measurement results and output a measurement report, which includes the key geometric parameters of the overhead contact system.
2. The contact wire parameter detection algorithm for three-dimensional laser point clouds according to claim 1, characterized in that: The self-propelled trolley moves automatically on the railway track. The self-propelled trolley is equipped with a laser scanner. The laser scanner acquires the three-dimensional laser point cloud data in real time by adjusting the scanning angle and scanning speed. The components of the three-dimensional laser point cloud data include the track, contact wire, and catenary.
3. The contact wire parameter detection algorithm for three-dimensional laser point clouds according to claim 1, characterized in that: The steps for spatial partitioning are as follows: S21. Based on the elevation method, the three-dimensional laser point cloud data is initially screened to select the track surface point cloud of the catenary system; S22. The track surface point cloud is longitudinally segmented to form a track dataset, and the spatial range of the contact line and catenary is determined based on the track surface point cloud data; S23. Using the height of the track surface point cloud as a reference, extract the contact line point cloud and the load-bearing cable point cloud within the spatial range.
4. The contact wire parameter detection algorithm for three-dimensional laser point clouds according to claim 3, characterized in that: The elevation method separates the track surface point cloud from other area point clouds by calculating the height value of each point in the three-dimensional laser point cloud data. The spatial range is defined by setting height range and position parameters according to the structural characteristics of the catenary system.
5. The contact wire parameter detection algorithm for a three-dimensional laser point cloud according to claim 1, characterized in that: The PCA-GICP algorithm includes principal component analysis and generalized iterative nearest point algorithm. The principal component analysis algorithm performs the coarse registration, and the generalized iterative nearest point algorithm performs the fine registration.
6. The contact wire parameter detection algorithm for three-dimensional laser point clouds according to claim 5, characterized in that: The coarse registration is performed by the principal component analysis algorithm to roughly align the point cloud dataset. The coarse registration involves calculating the covariance matrix of the point cloud dataset, extracting the main eigenvalues and eigenvectors based on the covariance matrix, and rotating the point cloud dataset according to the principal component direction determined by the eigenvectors to obtain the coarse registration result.
7. The contact wire parameter detection algorithm for a three-dimensional laser point cloud according to claim 6, characterized in that: The fine registration is based on the coarse registration result, and the covariance matrix is iteratively rotated and translated using the generalized iterative nearest point algorithm. During the fine registration process, the contact wire parameter measurement feature points are obtained.
8. The contact wire parameter detection algorithm for three-dimensional laser point clouds according to claim 1, characterized in that: The track feature points include left track feature points and right track feature points. The left track feature points include feature points on the upper surface of the left track and feature points on the inner surface of the left track. The right track feature points include feature points on the upper surface of the right track and feature points on the inner surface of the right track. The left track feature points and the right track feature points determine the track surface reference. The spatial geometric parameters are calculated as follows: the track superelevation is calculated based on the left and right track upper surface feature points; the track gauge is calculated based on the left and right track inner surface feature points; the contact wire guide height and pull-out value are calculated based on the contact wire feature points and the track surface reference; and the catenary guide height and pull-out value are calculated based on the catenary feature points and the track surface reference.
9. The contact wire parameter detection algorithm for three-dimensional laser point clouds according to claim 1, characterized in that: The statistical analysis involves calculating the standard deviation, maximum value, minimum value, and average value of the geometric parameters in the geometric parameter measurement results to form an analysis result, and generating the measurement report based on the analysis result.
10. A contact wire parameter detection system for three-dimensional laser point clouds, characterized in that: A self-propelled track trolley automatically travels along the railway track and collects three-dimensional laser point cloud data of the catenary system during the travel process; A point cloud data processing module, which preprocesses the three-dimensional laser point cloud data to form preprocessed point cloud data; A point cloud segmentation module, which segments the preprocessed point cloud data to form point clouds of contact wire components; A point cloud registration module, which precisely aligns the point clouds of the contact wire components to form contact wire point cloud data; A geometric parameter calculation module calculates key geometric parameters of the catenary based on the catenary point cloud data to form catenary parameters; The measurement result output module generates a measurement report based on the contact wire parameters.