CN117842130A - Rapid continuous profile detection method based on line structured laser - Google Patents
Rapid continuous profile detection method based on line structured laser Download PDFInfo
- Publication number
- CN117842130A CN117842130A CN202311834563.6A CN202311834563A CN117842130A CN 117842130 A CN117842130 A CN 117842130A CN 202311834563 A CN202311834563 A CN 202311834563A CN 117842130 A CN117842130 A CN 117842130A
- Authority
- CN
- China
- Prior art keywords
- data
- rail
- point
- point cloud
- steel rail
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 38
- 229910000831 Steel Inorganic materials 0.000 claims abstract description 64
- 239000010959 steel Substances 0.000 claims abstract description 64
- 238000000034 method Methods 0.000 claims abstract description 33
- 238000004458 analytical method Methods 0.000 claims abstract description 5
- 230000000737 periodic effect Effects 0.000 claims abstract description 4
- 239000013598 vector Substances 0.000 claims description 26
- 230000008859 change Effects 0.000 claims description 15
- 238000001914 filtration Methods 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 10
- 238000013519 translation Methods 0.000 claims description 9
- 230000005540 biological transmission Effects 0.000 claims description 8
- 230000001360 synchronised effect Effects 0.000 claims description 8
- 238000005259 measurement Methods 0.000 claims description 7
- 230000001960 triggered effect Effects 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 description 9
- 238000004364 calculation method Methods 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 241001669679 Eleotris Species 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005265 energy consumption Methods 0.000 description 2
- 238000009434 installation Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000005299 abrasion Methods 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000000227 grinding Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005498 polishing Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61K—AUXILIARY EQUIPMENT SPECIALLY ADAPTED FOR RAILWAYS, NOT OTHERWISE PROVIDED FOR
- B61K9/00—Railway vehicle profile gauges; Detecting or indicating overheating of components; Apparatus on locomotives or cars to indicate bad track sections; General design of track recording vehicles
- B61K9/08—Measuring installations for surveying permanent way
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Mechanical Engineering (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a line structure laser-based rapid continuous profile detection method, belonging to the field of steel rail protection; the method comprises the following steps: firstly, setting initial line parameters for a track to be tested; newly creating a detection project according to the line ledger data of the track to be detected, and importing project data; then, a line structure laser acquisition device is built by taking the track trolley as a carrying platform, and real-time data acquisition is carried out on the track to be detected; transmitting the characteristic points to an embedded computer for analysis, and extracting the characteristic points; splicing the original point clouds of a plurality of sensors together by using a calibrated deflection angle, and combining the original point clouds into profile data containing the complete steel rail point clouds; and finally, writing the profile data collected each time into a database in real time in a coordinate format, so that the profile display and periodic steel rail profile management in the later period are facilitated. The invention can be carried on a railway car and a detection vehicle, can adapt to high-low speed environments, and uses an encoder to trigger signals, thereby improving the reliability and stability of data; the detection efficiency is high.
Description
Technical Field
The invention belongs to the field of steel rail protection, and relates to a line structure laser-based rapid continuous profile detection method.
Background
The safety problem of the railway is always the first problem to be solved in transportation, and in the age of rapid development of the railway, the problems existing in the railway need to be found out in time and solved.
In railway transportation systems, the accuracy of the rail profile is critical to ensure the safety, comfort and efficiency of train operation. Railway tracks are the infrastructure of train travel, whose geometry and surface conditions directly affect the stability, energy consumption and transport efficiency of the train. Therefore, the method has important significance for accurately and real-time detecting and monitoring the profile of the steel rail.
First, accurate profile detection can help prevent and identify potential problems: such as track deformation, wear or tear, etc. By timely finding out the problems, railway operators take effective maintenance and repair measures in advance, so that hidden dangers possibly causing accidents are prevented, and the safety of a railway system is ensured.
Second, accurate profile information is critical to the operational stability and energy efficiency of the railroad train. Irregular track geometry can cause adverse effects of vibration, jolt, etc. during the train's travel and even cause damage to the train and equipment. By monitoring the profile, railway operators can adjust and optimize the track structure, improve the stability of the train, reduce the energy consumption and improve the transportation efficiency of the whole railway system.
The existing steel rail profile detection mainly uses contact type physical equipment, and the profile detection is carried out by using detection equipment after the position is manually selected.
How to realize continuous profile detection more accurately and more quickly and output results quickly becomes a problem to be solved urgently.
Disclosure of Invention
Aiming at the problems, the invention provides a rapid continuous profile detection method based on line structure laser, which is characterized in that the line structure laser is used for collecting the profile of a railway steel rail, a trolley is used as a vehicle-mounted platform, and an odometer is carried on the device, and is matched with a laser scanner, a high-definition camera, a data transmission module, a multi-sensor synchronous control unit, an embedded computer, various power supply equipment and the like; the sensor data of the laser of the left line structure and the right line structure are ensured to cover the whole steel rail through the installation of preset positions; high-speed continuous detection can be realized, and the accurate profile of each position is stored.
The method for detecting the rapid continuous profile based on the line structure laser comprises the following specific steps:
step one, setting initial line parameters for a track to be tested;
the initial parameters include starting mileage, ending mileage, railroad type, line/station, tie type, line type, etc.
Step two, newly creating a detection project according to the line ledger data of the track to be detected, and importing project data;
thirdly, constructing a line structure laser acquisition device by taking the track trolley as a carrying platform, and acquiring real-time data of the track to be detected;
the method comprises the following steps: various power supply equipment is arranged on a rail trolley, a lower beam of the rail trolley is a plane which is vertically arranged and perpendicular to the rail top surface of a steel rail, a line structure laser scanner is arranged at the bottom of the lower beam, a plurality of sensors of the laser scanner are distributed above the two sides of the left steel rail and the right steel rail, and the laser of the alignment steel rail guarantee line structure covers the whole rail head, the rail web and the rail bottom. The odometer, the multi-sensor synchronous control unit, the high-definition camera, the data transmission module and the embedded computer are all located above the laser scanner and are also installed on the plane of the underbeam.
The odometer is used for recording the mileage position of the measured data; the high-definition camera is used for recording the surface state of the steel rail on the operation site; the time-space synchronization of each sensor time is realized through the coordination of a multi-sensor synchronous control unit, and the one-to-one correspondence between the measured sensor data and the rail mileage is realized; the laser scanner is triggered by the carried encoder in the acquisition process, corresponding line structure sensor data and odometer data are stored in real time, and the line structure sensor data and odometer data enter the embedded computer for storage through the data transmission module and are displayed in real time.
Step four: transmitting the data acquired in real time to an embedded computer for analysis, and extracting characteristic points; and splicing the original point clouds of the sensors together by using the calibrated deflection angles, and combining the original point clouds into profile data containing the complete steel rail point clouds.
The specific process is as follows:
step 401, calculating a calibrated deflection angle; the method comprises the following steps:
aiming at a new steel rail to be detected, data of the steel rail to be detected is collected in real time and divided into areas which are the same as the standard steel rail, and each area is treated as follows:
firstly, calculating a normal vector of each point cloud through a point fitting plane aiming at each frame of two-dimensional point cloud in a current area, and finding out the fluctuation change of the normal vector of the area where the point cloud is located through calculating the curvature of each point cloud, wherein the point of the fluctuation change exceeding a threshold value is used as a characteristic point of the current area;
and then, manually rotating and translating the current area where the characteristic point is positioned according to the point cloud position in the standard steel rail corresponding to the characteristic point, so that the current area coincides with the point cloud area of the standard steel rail, and recording the deflection angle of the rotation and translation as a calibrated deflection angle.
Step 402, extracting characteristic points of each region by using real-time data acquired by a laser acquisition device with a linear structure, rotating and translating according to a calibrated angle to obtain an original point cloud, and splicing the original point cloud;
step 403, filtering an influence point cloud generated by noise of the laser sensor from the spliced original point cloud;
the method comprises the following steps: dividing the original point cloud into dimensions, traversing the point cloud layer by layer according to the dimensions, and removing points which are not in the dimension range, wherein the points obtained after the traversing are the filtered point cloud;
and 404, removing Gaussian noise from each frame of filtered point cloud by adopting a Gaussian filtering method, and obtaining the final actually acquired profile data containing the complete steel rail point cloud.
Step five: and the profile data which is acquired each time and contains the complete steel rail point cloud is written into a database in real time in a coordinate format for storage, so that the later profile display and periodic steel rail profile management are facilitated.
The invention has the advantages that:
1. compared with the existing contact type physical equipment for collecting the profile of the steel rail, the rapid continuous profile detection method based on the line structure laser can only carry out single-point measurement by manual operation, has low operation efficiency and is easy to generate errors in the manual operation process; the invention can realize continuous detection at the speed of 80-120km/h, has high detection efficiency, is automatically detected by equipment in the measuring process, and has relatively small measuring error.
2. A rapid continuous profile detection method based on line structured laser can realize non-contact measurement of line structured light, acquire point cloud data and image data of all cross sections along the line, and can confirm whether the line structured light is caused by a real environment or not through the image data if the point cloud scanning is wrong.
3. A rapid continuous profile detection method based on line structure laser has the advantages of higher accuracy of a line structure laser scanner, lower loss accuracy of scanning data and very high accuracy of the output detected profile due to the cooperation of other sensors.
4. A rapid continuous profile detection method based on line structure laser has strong detection environment adaptability; the device can be mounted on a rail car and a detection vehicle, can collect rail profile data when the vehicle checks a rail, can adapt to high-speed and low-speed environments, and uses the encoder to trigger signals, so that the reliability and stability of the data are improved.
5. A line structure laser-based rapid continuous profile detection method reduces labor cost and potential safety hazards. At present, most of the devices also adopt manual point taking measurement, the manual measurement period is overlong, and a large number of potential safety hazards exist in long-time up-road operation. The line structure laser continuous measurement mode detects at the speed of 80-120km/h, has high detection efficiency, and can effectively reduce the number of detection personnel and the time for going up the line, thereby reducing the labor cost and the hidden danger of safe production.
Drawings
FIG. 1 is a flow chart of a method for line structured laser-based rapid continuous profile detection in accordance with the present invention;
FIG. 2 is a diagram of a line structure laser acquisition device constructed by taking a track trolley as a carrying platform;
FIG. 3 is raw data collected by line structured laser sensors on both sides of a track of the present invention;
FIG. 4 is a graph showing the data of the line structured laser sensor of the present invention after correcting the deflection angle;
FIG. 5 shows the cloud data of the steel rail points after outliers are filtered by a filtering algorithm;
FIG. 6 is a comparison of the standard profile of the present invention with the scanned profile after transformation;
fig. 7 is a view showing the point cloud profile after deflection angle correction and various filtering according to the present invention.
Detailed Description
The following provides a complete and detailed description of embodiments of the invention, taken in conjunction with the accompanying examples and figures.
It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary only for explaining the present invention and are not to be construed as limiting the present invention.
The invention provides a line structure laser-based rapid continuous profile detection method, which comprises the following specific steps as shown in fig. 1:
step one, setting initial line parameters for a track to be tested;
according to the account data provided by the owner, initial line parameters required by the test route are written, including a starting mileage, a terminating mileage, a railway type, a line/station line, a sleeper type, a line type and the like.
Step two, newly creating a detection project according to the line ledger data of the track to be detected, importing project data, and initializing to synchronize the time of a plurality of sensors;
thirdly, constructing a line structure laser acquisition device by taking the track trolley as a carrying platform, and triggering an encoder to acquire real-time data of the track to be detected;
the method comprises the following steps: the line structure laser acquisition device uses the track trolley as a carrying platform, and various power supply equipment is arranged on the track trolley, so that the line structure laser acquisition device can detect the track trolley at the speed of 80-120km/h, and the working efficiency is greatly improved. The lower beam of the rail trolley is a plane which is vertically placed and perpendicular to the top surface of the steel rail, the bottom of the lower beam is provided with a line structure laser scanner, a plurality of sensors of the laser scanner are distributed above the two sides of the left steel rail and the right steel rail, and the laser of the line structure is aligned with the steel rail at a certain angle to cover the whole rail head, rail web and rail bottom. The odometer, the multi-sensor synchronous control unit, the high-definition camera, the data transmission module and the embedded computer are all located above the laser scanner and are also installed on the plane of the underbeam.
The odometer is used for recording the mileage position of the measured data; the laser scanner is installed at a preset position, so that line structure laser sensors at the left side and the right side of the steel rail cover the whole steel rail, and complete steel rail profile information can be collected conveniently; the high-definition camera is used for recording the surface state of the steel rail on the operation site; the time-space synchronization of each sensor time is realized through the coordination of the multi-sensor synchronous control unit, the stability of data during profile acquisition is ensured, and the one-to-one correspondence between measured sensor data and rail mileage is realized; the laser scanner is triggered through the carried encoder in the acquisition process, corresponding line structure sensor data and odometer data are stored in real time, the reliability of profile data is guaranteed, acquired data are transmitted to the embedded computer through the data transmission module, the acquired laser point clouds are stored after being combined, meanwhile, the acquired laser point clouds are displayed to the acquisition interface in real time, and the normal operation of the sensor and the effectiveness of the acquired data are guaranteed.
Step four: transmitting the data acquired in real time to an embedded computer for analysis, and extracting characteristic points; and splicing the original point clouds of the sensors together by using the calibrated deflection angles, and combining the original point clouds into profile data containing the complete steel rail point clouds.
The existing original Iterative Closest Point (ICP) algorithm is to construct a rotation translation matrix based on corresponding point pairs by solving the corresponding point pairs between a source point cloud and a target point cloud, transform the source point cloud to a coordinate system of the target point cloud by utilizing the solved matrix, estimate an error function of the transformed source point cloud and the target point cloud, and iterate until a given error requirement is met if the error function value is larger than a threshold value. The ICP algorithm adopts least square estimation to calculate the transformation matrix, has simple principle and better precision, but has slower algorithm calculation speed due to the adoption of iterative calculation, and has easy sinking into a local optimal solution.
According to the method, points with larger feature degree are extracted as key points according to normal vector angle change, initial matching point pairs are searched by calculating the curvature of the key points, and the initial matching point pairs are used as marks to improve the efficiency of an ICP algorithm.
The specific process is as follows:
step 401, aiming at a new steel rail to be tested, collecting data of the steel rail in real time, dividing the steel rail into areas which are the same as the standard steel rail, and extracting characteristic points of each area based on normal vectors;
firstly, aiming at a single-frame two-dimensional point cloud of a line structure in a current area, fitting a plane through points, such as a least square method; calculating normal vectors of the point cloud, calculating included angles between normal vectors of adjacent points by using methods such as point multiplication or inverse cosine of the vectors, and finding out points with the normal vector included angle change exceeding a threshold value or meeting specific conditions as characteristic points of the current area;
the normal vector at any point in the three-dimensional point cloud is called the normal, and the line is perpendicular to the tangential plane of the point; if the normal vector change of the flat area is more gentle, the normal vector change of the area with larger fluctuation change is larger; calculating the curvature of each characteristic point to obtain the fluctuation of the neighborhood of the characteristic point;
therefore, when the normal vector in the neighborhood where a certain point is located is greatly changed according to the change trend, the point can be set as a characteristic point, namely, the included angle between the normal line of the certain point and the normal line of the neighborhood point around the certain point is larger than a given threshold value, and the point is the characteristic point.
The detection steps are as follows:
(1) Given a point P i (X i, Y i, Z i ) T Calculating a domain point P in a sphere with radius r i1 ，P i2 ，…，P ik 。
(2) The unit normal vector n of this point is calculated using the following formulas 2.1,2.2 i Wherein P is i Is the three-dimensional centroid of the neighborhood point set, C is P i Is calculated by eigenvalue decomposition.
(3) Given a threshold T, if n i And all n i1 ，n i2 ，…，n ik The absolute values of the inner products of (a) are all greater than T, then P is considered to be i And the feature points are not feature points, and the feature points are the opposite feature points. The degree of variation f of the normal vector of a given point can also be calculated by 2.4 i Wherein θ is ij For point p i Included angle of normal vector of (c) and normal vector of its neighboring point, f i The larger the relief of the region, f i The smaller the area, the flatter the area, and an appropriate threshold T is selected to determine whether any point on the point cloud is a feature point.
Three eigenvectors of matrix CCorresponding three eigenvalues, and the eigenvector corresponding to the smallest eigenvalue is the point p i Normal vector of (2); calculated normal vector n i Only a numerical value, and no positive or negative division, the direction of the normal vector is generally defined to be consistent toward the viewpoint direction (from the viewpoint v p To p i Direction of (2) so n can be determined using equation 2.3 i Is a direction of (2). The normal vector n of the neighborhood point of the point can be calculated by the same method i1 ，n i2 ，…，n ik ：
The key point curvature calculation uses a matching algorithm, such as nearest neighbor matching, to match similar key points in two frames, and feature descriptors can be used to improve the accuracy of the matching.
The process comprises the following steps: performing eigenvalue decomposition according to covariance matrix C in 2.2 formula to obtain each eigenvalue of C, if the eigenvalue satisfies lambda 0 ≤λ 1 ≤λ 2 The surface curvature at point P is:
a smaller delta indicates a flatter neighborhood and a larger delta indicates a larger fluctuation of the neighborhood.
And step 402, manually rotating and translating the current area where the characteristic point is positioned according to the point cloud position in the standard steel rail corresponding to the characteristic point, so that the current area coincides with the point cloud area of the standard steel rail, and recording the deflection angle of the rotation and translation as a calibrated deflection angle.
During calibration, firstly scanning a standard steel rail to obtain point cloud; then, the steel rail of the track to be tested is scanned to obtain characteristic points, the positions of different characteristic points of the track to be tested are manually adjusted to enable the characteristic points to be matched with the point cloud of the profile of the standard steel rail through translation or rotation, the characteristic points are overlapped within a certain threshold value, and then specific adjustment parameters of each characteristic point are recorded.
Step 403, extracting characteristic points of each region by using real-time data acquired by the line structure laser acquisition device, rotating and translating according to a calibrated angle to obtain an original point cloud, and splicing the original point cloud;
step 404, filtering an influence point cloud generated by noise of the laser sensor from the spliced original point cloud;
the method comprises the following steps: dividing the original point cloud into dimensions, traversing each point in the point cloud in a specified dimension and a value range of the dimension, judging whether the point is in the range, then eliminating points which are not in the range, and obtaining the points in the range after traversing is finished as filtered point clouds;
and 405, removing Gaussian noise from each frame of filtered point cloud by adopting a Gaussian filtering method, and obtaining the final actually acquired profile data containing the complete steel rail point cloud.
As shown in fig. 4, a suitable filtering range is selected, and outliers outside the steel rail are filtered to obtain point clouds in the steel rail range.
Step five: and the profile data which is acquired each time and contains the complete steel rail point cloud is written into a database in real time in a coordinate format for storage, so that the later profile display and periodic steel rail profile management are facilitated.
Examples:
the detection method for acquiring the railway steel rail profile based on the high-precision line structure laser in the embodiment, as shown in fig. 1, comprises the following steps:
step 1: newly-built engineering and obtaining account data
The device comprises a trolley serving as a vehicle-mounted platform, an equipment-mounted odometer system, a laser scanner, a high-definition camera, a data transmission module, a multi-sensor synchronous control unit, an embedded computer, various power supply equipment and the like. Through the installation of preset positions, the data of the left line structure laser sensor and the right line structure laser sensor are ensured to cover the whole steel rail;
and then writing data required by the test route according to the ledger data provided by the owner, wherein the data comprises parameters such as starting mileage, ending mileage, railway type, line/station line, sleeper type, line type and the like.
And finally, newly creating a project, and combining the newly created project into a detection project through the provided line ledger data.
Step 2: acquiring data after engineering configuration parameters are imported;
the detection engineering acquires data, and the equipment realizes the time-space synchronization of each sensor time through the coordination of the synchronization module, so that the stability of the data during the profile acquisition is ensured. In the acquisition process, the data is triggered through an encoder carried by the equipment, the laser data and the odometer data are stored in real time, the reliability of profile data is guaranteed, and the effectiveness of the acquired data is guaranteed by displaying the acquired data to an acquisition interface in real time.
Step 3: displaying profile data in real time
(1) Analyzing data, as shown in fig. 2 and 3, extracting points with larger normal vector change feature degree as key points by calculating the curvature of each point cloud, splicing the original point clouds of a plurality of lasers of the steel rail together by using a calibrated deflection angle, and combining the original point clouds into data containing the complete steel rail point clouds.
(2) The embodiment proposes to use a line structure laser sensor to detect the track profile, and because of the fact that certain noise exists due to the hardware of the line structure laser, the measurement accuracy is affected, and the embodiment uses a self-grinding filtering method: and traversing each point in the point cloud in a specified dimension and a value range under the dimension, judging whether the point is in the range, then eliminating points which are not in the range, wherein the points in the range after the traversing is finished are the filtered point clouds, so that the influence point clouds generated by line structure optical noise can be well eliminated, and then the point with the best effect is reserved.
And then, the rail profile point cloud acquired by the single-frame laser sensor adopts a common Gaussian filtering method, and the Gaussian noise is removed and then the finally actually acquired profile is output. As shown in fig. 4, a suitable filtering range is selected, and outliers outside the steel rail are filtered to obtain point clouds in the steel rail range.
(3) Then dividing the rail point cloud into a rail head point cloud and a rail web rail bottom point cloud through threshold segmentation, taking the rail web rail bottom point cloud as an initial point set P, matching a standard point set Q of a designed rail, extracting points with larger change as key points according to the change of normal vector included angles in the kth iteration, and finding a point set S corresponding to a target point set T in the initial point set P from the Q through curvature search of the key points. And calculating a transformation matrix from T to S, updating the point sets until the average distance between the two point sets is smaller than a set threshold L, and recording the rotation translation matrix at the moment.
And according to the translation matrix and the rotation matrix obtained by calculation, applying the same rotation translation transformation to the rail head part point cloud, and comparing the transformed point cloud with the rail head point cloud of the standard contour, wherein the compared result is shown in fig. 5.
Step 4: storing profile data
And outputting the final scanning result after all the correction results are acted, as shown in fig. 6. The laser profile data of the line structure light scanning can be stored in real time, and if a database exists, the laser profile data can be directly written into the database. The profile collected each time can be stored in a fixed manner, so that later display and subsequent management operation are facilitated.
The method can store the basic profile of each acquired frame track, the stored profile is used for profile deformation analysis at a later stage, and the acquired profile can be periodically managed. If standard profile data exist, the stored profile can be compared with the standard profile to calculate the abrasion value, so that the later steel rail polishing scheme can be conveniently output.
The present invention is not limited to the above-mentioned embodiments, and any changes or substitutions that can be easily understood by those skilled in the art within the technical scope of the present invention are intended to be included in the scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.
Claims (4)
1. A line structure laser-based rapid continuous profile detection method is characterized by comprising the following specific steps:
step one, setting initial line parameters for a track to be tested; newly building a detection project according to the line ledger data of the track to be detected, and importing project data;
step two, constructing a line structure laser acquisition device by taking a track trolley as a carrying platform, and acquiring real-time data of a track to be detected;
the method comprises the following steps: various power supply devices are arranged on a rail trolley, a lower beam of the rail trolley is a plane which is vertically arranged and perpendicular to the top surface of a rail, a line structure laser scanner is arranged at the bottom of the lower beam, a plurality of sensors of the laser scanner are distributed above two sides of a left rail and a right rail, and the laser of a line structure is aligned to the rail to cover the whole rail head, the rail web and the rail bottom; the odometer, the multi-sensor synchronous control unit, the high-definition camera, the data transmission module and the embedded computer are all positioned above the laser scanner and are also arranged on the plane of the underbeam;
transmitting the data acquired in real time to an embedded computer for analysis, and iterating a nearest point algorithm to extract characteristic points according to the change of the normal vector included angle; splicing the original point clouds of a plurality of sensors together by using a calibrated deflection angle, and combining the original point clouds into profile data containing the complete steel rail point clouds;
the specific process is as follows:
step 401, calculating a calibrated deflection angle; the method comprises the following steps:
aiming at a new steel rail to be tested, data of the steel rail to be tested is collected in real time and divided into areas identical to those of a standard steel rail, and feature points are extracted from each area according to a normal vector included angle change iterative nearest point algorithm:
firstly, calculating a normal vector of each point cloud by a point fitting plane aiming at each frame of two-dimensional point cloud in a current area;
then, calculating the curvature of each point cloud, finding out the fluctuation change of the normal vector of the area where the point cloud is located, and selecting the point of which the fluctuation change exceeds a threshold value as a characteristic point of the current area;
finally, according to the point cloud position in the standard steel rail corresponding to the characteristic point, manually rotating and translating the current area where the characteristic point is located to enable the current area to coincide with the point cloud area of the standard steel rail, and recording the deflection angle of rotation and translation as a calibrated deflection angle;
step 402, real-time data acquired by the line structure laser acquisition device are similarly divided into areas, feature points of each area are extracted, rotation and translation are carried out according to the calibrated angles of the areas, and an original point cloud of each area is obtained;
step 403, splicing original point clouds of all areas, and filtering influence point clouds generated by noise of the laser sensor;
step 404, removing Gaussian noise from each frame of filtered point cloud by adopting a Gaussian filtering method, and obtaining profile data which is finally and actually acquired and contains the complete steel rail point cloud;
step five: and the profile data which is acquired each time and contains the complete steel rail point cloud is written into a database in real time in a coordinate format for storage, so that the later profile display and periodic steel rail profile management are facilitated.
2. A line structured laser based rapid succession of profile detection methods as in claim 1, wherein said initial parameters include start mileage, end mileage, railroad type, line/stop line, tie type, and line type.
3. The method for rapid continuous profile detection based on line structured laser as claimed in claim 1, wherein in the second step, the odometer is used for recording the mileage position of the measurement data; the high-definition camera is used for recording the surface state of the steel rail on the operation site; the time-space synchronization of each sensor time is realized through the coordination of a multi-sensor synchronous control unit, and the one-to-one correspondence between the measured sensor data and the rail mileage is realized; the laser scanner is triggered by the carried encoder in the acquisition process, corresponding line structure sensor data and odometer data are stored in real time, and the line structure sensor data and odometer data enter the embedded computer for storage through the data transmission module and are displayed in real time.
4. The method for rapid continuous profile detection based on line structured laser as claimed in claim 1, wherein said step 403 comprises: dividing the original point cloud into dimensions, traversing the point cloud layer by layer according to the dimensions, and removing points which are not in the dimension range, wherein the points obtained after the traversing is finished are the filtered point cloud.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311834563.6A CN117842130A (en) | 2023-12-28 | 2023-12-28 | Rapid continuous profile detection method based on line structured laser |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311834563.6A CN117842130A (en) | 2023-12-28 | 2023-12-28 | Rapid continuous profile detection method based on line structured laser |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117842130A true CN117842130A (en) | 2024-04-09 |
Family
ID=90533969
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311834563.6A Pending CN117842130A (en) | 2023-12-28 | 2023-12-28 | Rapid continuous profile detection method based on line structured laser |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117842130A (en) |
-
2023
- 2023-12-28 CN CN202311834563.6A patent/CN117842130A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2023019709A1 (en) | Automatic detection method of conductor height and pull-out value of overhead line system based on vehicle-mounted mobile laser point cloud | |
| US6909514B2 (en) | Wheel profile inspection apparatus and method | |
| CN109708615B (en) | Subway tunnel clearance dynamic detection method based on laser scanning | |
| CN100480627C (en) | Steel rail wearing integrative parameter vehicle-mounted dynamic measuring device and method | |
| CN114049294B (en) | Road surface technical condition detection method and equipment based on three-dimensional contour | |
| CN109060821B (en) | Tunnel disease detection method and tunnel disease detection device based on laser detection | |
| CN110293993B (en) | Switch detection device and system | |
| CN109855562B (en) | Method and device for measuring rail abrasion | |
| CN105761305B (en) | A kind of vehicle traveling detection method and system | |
| CN100419373C (en) | Apparatus and method for detecting obstacles around current collector of railway vehicle | |
| CN109060820B (en) | Tunnel disease detection method and tunnel disease detection device based on laser detection | |
| CN102914290A (en) | Metro gauge detecting system and detecting method thereof | |
| CN115503788B (en) | Ballast track bed section scanning detection system | |
| KR102255978B1 (en) | Apparatus and method for generating tunnel internal precise map based on tunnel internal object detection using 3D sensor | |
| CN110132990A (en) | A kind of profile of steel rail detection method based on image recognition | |
| Yao et al. | The online monitoring system of pantograph slider based on 2D laser displacement sensors | |
| CN116128834A (en) | Rail transit platform limit rapid detection method based on three-dimensional laser scanning | |
| CN114897777A (en) | Full-automatic extraction method of laser point cloud of overhead line system supporting facility considering spatial relationship | |
| CN112581521B (en) | Method for extracting central line of magnetic suspension track | |
| CN114119957A (en) | Method and device for detecting high-speed railway steel rail profile | |
| CN112964195A (en) | Power supply rail geometric parameter comprehensive detection method and system based on laser triangulation | |
| CN117842130A (en) | Rapid continuous profile detection method based on line structured laser | |
| CN115682956B (en) | Dynamic compensation method and system for contact rail detection data during vehicle body vibration | |
| CN110525463A (en) | Underground mine track transportation rail checks vehicle and inspection method | |
| Larionov et al. | Multiphysical system of operational monitoring of the condition of the railway track |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information |
Inventor after: Jiang Jun Inventor after: Chang Siming Inventor after: Zhao Guanchuang Inventor after: Hou Yinqing Inventor after: Liu Yongqian Inventor after: Zhang Yucheng Inventor after: Wang Junping Inventor after: Xu Hailong Inventor after: Liu Xin Inventor after: Shao Shichun Inventor before: Jiang Jun Inventor before: Zhao Guanchuang Inventor before: Hou Yinqing Inventor before: Liu Yongqian Inventor before: Wang Junping Inventor before: Xu Hailong |
|
| CB03 | Change of inventor or designer information |