CN118424101A - Ballasted track ballast square quantity real-time detection device and method - Google Patents
Ballasted track ballast square quantity real-time detection device and method Download PDFInfo
- Publication number
- CN118424101A CN118424101A CN202410456841.7A CN202410456841A CN118424101A CN 118424101 A CN118424101 A CN 118424101A CN 202410456841 A CN202410456841 A CN 202410456841A CN 118424101 A CN118424101 A CN 118424101A
- Authority
- CN
- China
- Prior art keywords
- ballast
- area
- point cloud
- track
- module
- 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
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000011897 real-time detection Methods 0.000 title description 5
- 238000001514 detection method Methods 0.000 claims abstract description 18
- 229910000831 Steel Inorganic materials 0.000 claims description 34
- 239000010959 steel Substances 0.000 claims description 34
- 241001669679 Eleotris Species 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 16
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 4
- 238000013135 deep learning Methods 0.000 claims description 3
- 230000000007 visual effect Effects 0.000 claims description 3
- 229910020929 Sn-Sn Inorganic materials 0.000 claims description 2
- 229910008827 Sn—Sn Inorganic materials 0.000 claims description 2
- 238000003708 edge detection Methods 0.000 claims description 2
- 238000001914 filtration Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 abstract description 3
- 238000012423 maintenance Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000011179 visual inspection Methods 0.000 description 1
Landscapes
- Machines For Laying And Maintaining Railways (AREA)
Abstract
The invention discloses a device and a method for detecting the ballast square quantity of a ballasted track in real time, which belong to the field of track detection of track traffic, and the device and the method for detecting the ballast square quantity of the ballasted track in real time are used for acquiring specific changes of time-space point cloud data of a full section along a railway by integrating a GNSS/DMI system, a mobile three-dimensional laser scanner module, an AI module and other devices through an advanced mobile laser image and AI technology, so that the ballast square quantity in different areas of the section of the track is calculated in real time, the technical effects of high-efficiency and high-precision railway detection are achieved, and the device and the method have the advantages of high detection efficiency, high three-dimensional scene restoration and the like, and have important significance for supporting railway operation decisions and disaster early warning.
Description
Technical Field
The invention belongs to the field of track detection of track traffic, and particularly relates to a device and a method for detecting the ballast square quantity of a ballasted track in real time.
Background
Counting the ballast square of a ballasted railway is very challenging in engineering, and is different from counting the number of regular components such as steel rails and sleepers, and the irregularity of the ballast square increases the complexity and uncertainty of calculation. The variation of track bed sections, such as straight and curved sections, different types of roadbeds, bridges, tunnels, and single and double line sections, makes obtaining accurate ballast data a difficult task. These data are critical to the basic decision, design and construction guidance for ballast maintenance.
The traditional railway ballast measuring method mainly relies on manual visual inspection or online approach single-point measuring operation by using a total station, a level gauge, a tape measure and the like, has the characteristics of low operation efficiency, long operation period and low measuring accuracy, and in addition, although the three-dimensional mobile scanning system is used for calculating the railway ballast square quantity by some methods, the method generally involves later data processing, so that the result timeliness is poor, the operation cannot be guided in a real-time alarm manner, no image assistance exists, and the accuracy of disease positioning is affected. Finding a quick and accurate measurement method is an urgent problem to be solved in the current stage.
The existing railway ballast measuring method is difficult to meet the requirements of quick and accurate measurement in field actual operation due to low efficiency, poor precision and lack of real-time data and image assistance.
Disclosure of Invention
Aiming at the problems, the invention provides a device and a method for detecting the ballast square quantity of a ballasted track in real time, which acquire specific changes of full-section space-time point cloud data along a railway by using advanced mobile laser images and AI technology, calculate the ballast square quantity in different areas of the section of the track in real time, achieve the technical effects of high-efficiency and high-precision railway detection, have the advantages of high detection efficiency, high three-dimensional scene restoration and the like, and have important significance for supporting railway operation decisions and disaster early warning.
The invention discloses a method for detecting the ballast square quantity of a ballast track in real time, which comprises the following specific steps:
step 1: drawing standard ballast bed section models with different working conditions according to different line types and structures, and storing.
Step 2: and acquiring three-dimensional point cloud data of the track section.
Step 3: and calling a corresponding standard ballast section model according to the standing book of which the current coordinate corresponds to the standard ballast model.
Step 4: and identifying and dividing the three-dimensional point cloud data into calculation areas, calculating the difference between the calculation areas and the standard ballast of the cross section of the ballast bed, and further counting the filling sum of the ballast weights of the areas.
Step 5: determining whether to synchronously acquire images of ballastless areas of the trackside cameras according to the filling sum of the ballast square quantities and transmitting the images to an upper computer unit; if the delta V is negative, the defect of the current area is needed to be repaired, and the shot high-definition camera picture is selected according to the image shooting interval and the current coordinate and transmitted to the upper computer unit.
Step 6: and the upper computer unit performs visual display and early warning on the detection result to form a final detection report.
The invention applies a real-time detection device for the ballast quantity of a ballast track to realize the method, and the device comprises a mobile three-dimensional laser scanner module, a memory unit, a trackside camera module, an AI module, an industrial personal computer synchronization unit and an upper computer unit.
The mobile three-dimensional laser scanner module is used for continuously collecting three-dimensional point cloud data of the track section in real time under high-speed movement.
The GNSS/DMI system is used for determining current real-time coordinates.
The memory unit is used for storing a standard ballast bed model, is matched with the GNSS/DMI system for use, acquires current real-time coordinates and calls a corresponding standard ballast bed section model.
And the AI module utilizes a deep learning algorithm to identify the three-dimensional point cloud data of the track section acquired by the received mobile three-dimensional laser scanning module in real time, and performs segmentation and calculation of the difference between the railway ballast square quantity of each part and the standard railway ballast, so that the identification precision and efficiency are ensured.
The trackside camera module is used for shooting a detail image of the ballastless area.
The upper computer unit is used for visually displaying ballastless pictures and giving early warning.
The industrial personal computer synchronization unit is connected with the AI module and the mobile three-dimensional laser scanner module; the industrial personal computer synchronization unit is used for synchronously moving the three-dimensional point cloud data of the track end surface acquired by the three-dimensional laser scanning module, so as to ensure accurate recording and real-time transmission of the data; and the industrial personal computer synchronization unit is used for recording the calculation result returned by the AI module, judging whether to acquire the image shot by the trackside camera or not by using the calculation result, outputting the image to the upper computer unit and early warning.
The invention has the advantages that:
(1) The detection efficiency is high. The electric rail car is utilized for operation, manual access is not needed, and the field operation scanning efficiency is high;
(2) The three-dimensional scene has high reduction degree. Non-contact measurement is carried out, and point cloud data and image data of all cross sections along the line are collected, so that good space reference is provided for detecting railway ballast diseases;
(3) The detection process is automated in real time. Classifying the section point cloud by using the efficient AI module 4, efficiently acquiring the section parameters of the railway ballast, and realizing automatic extraction of the railway ballast profile by combining with the roadbed section model, thereby realizing real-time detection and real-time maintenance of the railway ballast square quantity;
(4) The result precision is high. The high-precision high-resolution point cloud is utilized to extract and calculate the section parameters of the railway ballast, so that the accuracy of calculating the square quantity of the railway ballast is greatly improved compared with manual dotting, and the railway ballast is convenient to finely manage and maintain in the later stage.
Drawings
FIG. 1 is a block diagram of a real-time detection device for the ballast quantity of a ballasted track;
FIG. 2 is a flow chart of a method for detecting the ballast square quantity of the ballasted track in real time;
FIG. 3 is a cross-sectional model diagram of a standard ballast bed under different working conditions;
fig. 4 is a schematic diagram showing the phase difference between the point cloud data of different areas and the standard cross section.
In the figure:
1-movement three-dimensional laser scanner module
2-Memory cell
3-Trackside camera module
4-AI module
5-Industrial personal computer synchronization unit
6-Upper computer unit
7-GNSS/DMI system
Detailed Description
The invention will be described in further detail with reference to the accompanying drawings.
The device for detecting the ballast square quantity of the ballast track in real time adopts the technical means of the three-dimensional laser scanner, the AI module and the trackside camera, and achieves the technical effects of high-efficiency and high-precision railway detection by collecting the specific change of the full-section three-dimensional point cloud data along the railway and carrying out real-time calculation and early warning, thereby having great significance for the safety and maintenance of railway transportation.
As shown in fig. 1, the method specifically includes: the system comprises a mobile three-dimensional laser scanner module 1, a memory unit 2, a trackside camera module 3, an AI module 4, an industrial personal computer synchronization unit 5, an upper computer unit 6 and a GNSS/DMI system 7.
The mobile three-dimensional laser scanner module 1 is used for continuously collecting three-dimensional point cloud data of the track section in real time under high-speed movement.
The GNSS/DMI system 7 is used to determine the current real-time coordinates.
The memory unit 2 is used for storing a standard ballast bed model, is used together with the GNSS/DMI system 7, acquires current real-time coordinates and calls a corresponding standard ballast bed section model.
And the AI module 4 utilizes a deep learning algorithm to identify the three-dimensional point cloud data of the track section acquired by the received mobile three-dimensional laser scanning module in real time, and performs segmentation and calculation of the difference between the railway ballast square quantity of each part and the standard railway ballast, so that the identification precision and efficiency are ensured.
The trackside camera module 3 is used for shooting a detail image of the ballastless area.
The upper computer unit 6 is used for visually displaying ballastless pictures and giving early warning.
The industrial personal computer synchronization unit 5 is connected with the AI module 4 and the mobile three-dimensional laser scanner module 1. The industrial personal computer synchronization unit 5 is used for synchronously moving the three-dimensional point cloud data of the track end surface acquired by the three-dimensional laser scanning module, so as to ensure accurate recording and real-time transmission of the data; meanwhile, the industrial personal computer synchronization unit 5 is used for recording the calculation result returned by the AI module 4, judging whether to acquire the image shot by the trackside camera or not by utilizing the calculation result, outputting the image to the upper computer unit 6 and early warning.
Based on the device, the method for detecting the ballast square quantity of the ballasted track in real time comprises the following specific steps as shown in fig. 2:
Step 1: drawing standard ballast bed section models with different working conditions according to different circuit types and structures, and storing the standard ballast bed section models into a memory unit 2;
According to the types of high speed/common speed- & gt single line/double line (multi-line) - & gt positive line/station line- & gt seamed/seamless- & gt sleeper type- & gt allowable speed- & gt curve widening and the like, standard track bed section models with different working conditions are drawn by combining conditions of roadbeds, tunnels, bridges and the like, and are stored in a memory unit 2. As shown in fig. 3, a standard ballast bed section model under certain working conditions is shown.
Step 2: the industrial personal computer synchronization unit 5 synchronously acquires the three-dimensional point cloud data of the track section acquired by the mobile three-dimensional laser scanner module 1 and transmits the three-dimensional point cloud data to the AI module 4 in real time.
Step 3: the GNSS/DMI system 7 determines the current real-time coordinates, and calls the corresponding standard ballast section model from the memory unit 2 to transmit to the AI module 4 according to the ledger corresponding to the standard ballast model with the coordinates of the current three-dimensional point cloud data.
Step 4: and identifying and dividing a calculation region by using the AI module 4, calculating the difference between the calculation region and the standard ballast of the section of the ballast bed, and further counting the filling sum of the ballast squares in each region.
1) Analyzing the three-dimensional point cloud data of each section, setting a certain threshold value according to the fixed positions LD and RD of the three-dimensional laser scanner module 1 to the left and right steel rails, respectively obtaining point cloud data of areas near the left and right steel rails, filtering and denoising by utilizing algorithms such as SOR, CSF and the like, respectively performing ICP matching by using a standard track model and the point cloud data of the areas near the left and right steel rails, and obtaining center coordinates, rotation angles of the rail surfaces of the left and right steel rails and point cloud data forming the profile of the steel rails;
2) And constructing a left steel rail surface coordinate system and a right steel rail surface coordinate system by utilizing the rail surface centers and the rotation angles of the left steel rail and the right steel rail, and converting the point cloud data coordinates of the areas near the left steel rail and the right steel rail into the rail surface coordinate system.
3) Extracting the boundaries of the left and right steel rails, the sleeper and the sleeper box in the point cloud data by using a canny edge detection algorithm, respectively obtaining starting and ending coordinates of the left and right side slopes, the left and right steel rails and the sleeper box, setting the area from the outer side of the left steel rail to the left side edge of the sleeper as L0, the area from the inner side of the left steel rail to the inner side of the right steel rail as L1, the area from the outer side of the right steel rail to the right side of the sleeper as L2, the area from the left side of the sleeper to the starting point of the left side slope as L3, the area from the right side of the sleeper to the starting point of the right side slope as L4, and the area between the starting and ending positions of the left side slope and the right side slope as L5 and L6.
4) Calculating the point cloud area V 0＝Sn, L and n E (0, 1,2,3,4,5 and 6) of each area divided in the step 3; v 0 is the current area ballast square volume V0, L is the point cloud sampling interval, and S n is the point cloud area of each area.
5) As shown in FIG. 4, the difference between the area of the orange point cloud partial area and the area of the standard ballast cross section model in the area is the difference between the square quantity area of the railway ballast with the current cross section, and the sampling interval is multiplied, namely the difference DeltaV 0＝(Sn-Sn')*L,n∈(0,1,2,3,4,5,6),Sn' of the square quantity of the railway ballast in the area is the point cloud area of the standard ballast cross section in each area。
6) Counting the total filling quantity delta V=delta V 0+ΔV1+…ΔV6 of the railway ballast of each area, wherein delta V is the total quantity to be filled of the railway ballast of the current component; and 0 to 6 are region numbers indicating the division in the step 4.
Step 5: the industrial personal computer synchronization unit 5 determines whether to synchronously acquire images of ballastless areas of the trackside cameras according to the filling sum DeltaV of the ballast square quantity and transmits the images to the upper computer unit 6。
If the delta V is negative, the defect of the current area is needed to be repaired, and the shot high-definition camera picture is selected according to the image shooting interval and the current coordinate and transmitted to the upper computer unit 6.
Step 6: the upper computer unit 6 performs visual display and early warning on the detection result to form a final detection report as shown in the following table.
The invention combines various high-precision devices such as a GNSS/DMI system 7, a mobile three-dimensional laser scanner module 1, a trackside camera module 3 and the like, and realizes the rapid real-time acquisition of the full-section space-time point cloud data along the railway and the high-definition extraction of ballastless images; the application of the AI technology and the synchronous control of the industrial personal computer obviously improve the transmission and analysis capacity of real-time data; the defects of the traditional detection method in real-time, accuracy and comprehensiveness are effectively overcome, and an efficient and reliable solution is provided for railway track maintenance.
Claims (4)
1. A method for detecting the ballast square quantity of a ballast track in real time is characterized by comprising the following steps: the method comprises the following specific steps:
step 1: drawing standard ballast bed section models with different working conditions according to different circuit types and structures, and storing;
Step 2: collecting three-dimensional point cloud data of a track section;
Step 3: calling a corresponding standard ballast bed section model according to a standing book corresponding to the standard ballast bed model in the current coordinate;
Step 4: identifying and dividing the three-dimensional point cloud data into calculation areas, calculating the difference between the calculation areas and the standard ballast of the section of the ballast bed, and further counting the filling sum of the ballast weights of the areas;
Step 5: determining whether to synchronously acquire images of ballastless areas of the trackside cameras according to the filling sum of the ballast square quantities and transmitting the images to an upper computer unit; if the delta V is a negative value, indicating that the ballastless of the current area needs to be repaired, selecting a shot high-definition camera picture according to the image shooting interval and the current coordinate, and transmitting the shot high-definition camera picture to an upper computer unit;
step 6: and the upper computer unit performs visual display and early warning on the detection result to form a final detection report.
2. The method for detecting the ballast square quantity of the ballasted track in real time according to claim 1, wherein the method comprises the following steps: in the step 1, standard track bed section models with different working conditions are drawn according to the conditions of high speed/common speed, single line/double line or multiple lines, positive line/station line, seam/seamless, sleeper type, allowable speed, curve widening, roadbed, tunnel, bridge and the like.
3. The method for detecting the ballast square quantity of the ballasted track in real time according to claim 1, wherein the method comprises the following steps: the specific method of the step 4 is as follows:
1) Analyzing the three-dimensional point cloud data of each section, respectively acquiring point cloud data of areas near the left and right steel rails according to the fixed positions LD and RD of the three-dimensional laser scanner module to the left and right steel rails, filtering and denoising, and respectively performing ICP matching by using a standard track model and the point cloud data of the areas near the left and right steel rails to acquire the center coordinates, the rotation angles of the rail surfaces of the left and right steel rails and the point cloud data forming the outline of the steel rails;
2) Constructing a left steel rail surface coordinate system and a right steel rail surface coordinate system by utilizing rail surface centers and rotation angles of the left steel rail and the right steel rail, and converting the point cloud data coordinates of the areas near the left steel rail and the right steel rail into the rail surface coordinate system;
3) Extracting the boundaries of a left steel rail, a right steel rail, a sleeper and a sleeper box in point cloud data by using a canny edge detection algorithm, respectively obtaining starting and ending coordinates of a left side slope, a right side slope, a left side rail, a right side rail and a sleeper box, setting an area from the outer side of the left steel rail to the left side edge of the sleeper as L0, an area from the inner side of the left steel rail to the inner side of the right steel rail as L1, an area from the outer side of the right side rail to the right side of the sleeper as L2, an area from the left side of the sleeper to the starting point of the left side slope as L3, an area from the outer side of the sleeper to the starting point of the right side slope as L4, and an area between the starting and ending positions of the left side slope as L5 and L6;
4) Calculating the point cloud area V 0＝Sn x L of each area, and n E (0, 1,2,3,4,5, 6); v 0 is the current area ballast square volume V0, L is the point cloud sampling interval, a fixed value and S n is the point cloud area of each area;
5) The area difference value of the point cloud partial area in the standard ballast section model is the difference of the current section ballast square quantity area, and the sampling interval is multiplied, namely the difference value DeltaV 0＝(Sn-Sn')*L,n∈(0,1,2,3,4,5,6),Sn' of the ballast square quantity of the point cloud area is the point cloud area of the standard ballast section in each area;
6) Counting the total filling quantity delta V=delta V 0+ΔV1+…ΔV6 of the railway ballast of each area, wherein delta V is the total quantity to be filled of the railway ballast of the current component; and 0 to 6 are point cloud area numbers representing the division.
4. The method for detecting the ballast square quantity of the ballasted track in real time according to claim 1, wherein the method comprises the following steps: the detection device is realized by the following steps:
the detection device comprises a mobile three-dimensional laser scanner module, a memory unit, a trackside camera module, an AI module, an industrial personal computer synchronization unit and an upper computer unit;
The mobile three-dimensional laser scanner module is used for continuously collecting three-dimensional point cloud data of the track section in real time under high-speed movement;
The GNSS/DMI system is used for determining current real-time coordinates;
The memory unit is used for storing a standard ballast bed model, is matched with the GNSS/DMI system for use, acquires current real-time coordinates and calls a corresponding standard ballast bed section model;
The AI module utilizes a deep learning algorithm to identify the three-dimensional point cloud data of the track section acquired by the received mobile three-dimensional laser scanning module in real time, and segments and calculates the difference between the railway ballast square quantity of each part and the standard railway ballast, so that the identification precision and efficiency are ensured;
The trackside camera module is used for shooting a detail image of the ballastless area;
The upper computer unit is used for visually displaying ballastless pictures and giving early warning;
The industrial personal computer synchronization unit is connected with the AI module and the mobile three-dimensional laser scanner module; the industrial personal computer synchronization unit is used for synchronously moving the three-dimensional point cloud data of the track end surface acquired by the three-dimensional laser scanning module, so as to ensure accurate recording and real-time transmission of the data; and the industrial personal computer synchronization unit is used for recording the calculation result returned by the AI module, judging whether to acquire the image shot by the trackside camera or not by using the calculation result, outputting the image to the upper computer unit and early warning.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311818592 | 2023-12-27 | ||
| CN2023118185923 | 2023-12-27 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118424101A true CN118424101A (en) | 2024-08-02 |
Family
ID=92322308
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410456841.7A Pending CN118424101A (en) | 2023-12-27 | 2024-04-16 | Ballasted track ballast square quantity real-time detection device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118424101A (en) |
-
2024
- 2024-04-16 CN CN202410456841.7A patent/CN118424101A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110647798B (en) | Automatic track center line detection method based on vehicle-mounted mobile laser point cloud | |
| CN108248634B (en) | Track switch and track full-sized abrasion vision measurer and its measurement method | |
| CN105905132B (en) | A kind of track condition intelligent patrol detection apparatus and method | |
| CN110986820B (en) | Track intrusion detection method and device and electronic equipment | |
| CN104567708B (en) | Full section of tunnel high speed dynamical health detection means and method based on active panoramic vision | |
| KR101111569B1 (en) | Monitering System of Railroad Facilities using Railway Vehicle | |
| CN103630088B (en) | High accuracy tunnel cross-section detection method based on bidifly light belt and device | |
| CN109492071B (en) | Railway high-precision map data processing method and system | |
| CN115503788B (en) | Ballast track bed section scanning detection system | |
| CN109470207A (en) | A kind of complete detection method for tunnel | |
| CN114312877B (en) | Railway comprehensive inspection system | |
| CN108981618A (en) | A kind of tunnel just branch flatness analysis method and system | |
| CN111579560A (en) | Defect positioning device and method used in tunnel | |
| CN111561967A (en) | Real-time online detection method and system for pantograph-catenary operation state | |
| CN115435700A (en) | Device for quickly and automatically measuring limit of trackside equipment | |
| CN114740510A (en) | Operation and maintenance railway digital tamping method and system based on linear measurement | |
| CN114379598B (en) | Railway comprehensive inspection system | |
| CN108596968B (en) | Sleeper counting method based on track 3D depth image | |
| CN115562284A (en) | Method for realizing automatic inspection by high-speed rail box girder inspection robot | |
| CN114379607A (en) | Comprehensive railway inspection method | |
| CN117848238A (en) | Rail transit station actual limit measurement method based on laser point cloud | |
| CN118424101A (en) | Ballasted track ballast square quantity real-time detection device and method | |
| CN105346568B (en) | Method for positioning main lines and branch lines through GPS longitude and latitude | |
| CN114913131A (en) | Vertical absolute and relative vertical deformation detection method for ballastless track structure | |
| CN115035087A (en) | Novel railway line image detection method and system |
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 |