Automatic leak repairing robot and leak repairing method thereof

Abstract

The invention discloses an automatic leak repairing robot and a leak repairing method thereof, and the method comprises the following steps: S1, enabling the leak repairing robot to move to a target carriage along a track, positioning the position of the carriage through a scanning device, and recognizing the identification number of the carriage; s2, pre-stored compartment door parameters are called according to the compartment identification number; s3, moving to a vehicle door position according to the compartment door parameters, and identifying door slot information and adhesive residue information; s4, the moving speed is adjusted according to the door crack information and the residual glue information, and spraying and leak repairing are conducted on the door crack; and S5, after leakage repairing of the current vehicle door is completed, the vehicle door is moved to the next vehicle door, and S3 and S4 are executed circularly till leakage repairing operation of the whole compartment is completed. According to the invention, a machine vision perception technology is adopted, a 2D camera image and 3D radar point cloud data program algorithm is innovated, and the length, width and position information of the compartment door slot are accurately identified; interference factors such as dirt, residual glue and ambient light change of the compartment door slot are solved through processing modes such as noise filtering, data acquisition and intelligent compensation, and the recognition rate of the compartment door slot is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of carriage leak repair, and in particular to an automatic leak repair robot and a leak repair method thereof. Background Art

[0002] In the existing technology, coal-carrying train carriages are prone to abnormal conditions of the middle and small doors due to long-term use, and even problems such as deformation, leakage, and damage of the doors, resulting in coal spillage. The scattered coal will cause serious harm to people and the environment along the railway, and affect the safety of train operation.

[0003] In order to solve the problem of coal spillage caused by leaks in the middle and small doors of coal-carrying train carriages, a reinforcement and leak repair station was established, and manual operations were used to reinforce and repair the doors. Manual inspections were conducted to determine whether there were gaps in the small and middle doors on both sides of each carriage, and the gaps were repaired with foam strips or foam glue. The manual operation method required a large amount of manpower, and there were many problems such as high labor intensity and long construction time.

[0004] In view of the above situation, an automatic leak repair robot and its leak repair method are needed to replace manual reinforcement and leak repair operations, make the work simple and convenient, fundamentally eliminate potential safety hazards, improve the industry's inherent safety level, and achieve the goal of reducing manpower and increasing efficiency. Summary of the Invention

[0005] According to an embodiment of the present invention, an automatic leak repairing robot and a leak repairing method thereof are provided, comprising the following steps: S1, the leak repair robot moves along the track to the target carriage, locates the carriage position and recognizes the carriage identification number through the scanning device; S2. Calling pre-stored carriage door parameters according to the carriage identification number; S3. Move to the door position according to the door parameters and identify the door gap information and residual glue information; S4. Adjust the moving speed according to the door gap information and residual glue information to spray and seal the door gap; S5. After completing the current door leak repair, move to the next door and execute S3 and S4 in a loop until the entire carriage leak repair operation is completed.

[0006] Furthermore, the door gap information includes the door gap length, door gap width, door gap depth and the exact position of the door gap.

[0007] Furthermore, identifying the door gap information and residual glue information includes the following steps: Collect image data through 2D camera and perform preprocessing; Use the YOLO algorithm to perform preliminary target detection on the preprocessed image and identify the approximate location of the door gap; Based on the gradient changes of the RGB channels in the residual glue area, the YOLO detection results are optimized through the Transformer encoding and decoding mechanism to identify the residual glue location.

[0008] Furthermore, identifying the door gap information and residual glue information also includes the following steps: Use a 3D camera to collect point cloud data of the door gap area and perform preprocessing; Align 2D image data and 3D point cloud data through a hybrid calibration method; Eliminate, expand, statistically filter, and grid the point cloud data to extract depth and normal vector features; The residual glue position is analyzed based on the changes in depth and normal vector, and the coordinate information of the door gap and residual glue is output.

[0009] Furthermore, aligning 2D image data and 3D point cloud data through a hybrid calibration method includes the following steps: Determine the transformation matrix between 2D image data and 3D point cloud data through feature matching; The transformation matrix is used to map the 2D image data and 3D point cloud data to the same coordinate system.

[0010] Furthermore, the point cloud data of the door gap area is collected by a 3D camera, and noise filtering and denoising operations are performed.

[0011] Furthermore, the culling and expansion of point cloud data include: Eliminate some invalid or duplicate data points; The missing point cloud is reasonably expanded based on the geometric characteristics of the door gap area.

[0012] Furthermore, statistical filtering includes: analyzing the distribution characteristics of point cloud data and removing noise points that deviate from the normal range according to a preset threshold.

[0013] Furthermore, analyzing the residual glue position based on the changes in depth and normal vectors includes the following steps: Compare the depth and normal vector baseline features of the normal door gap area; Detect sudden depth changes or abnormal direction changes of normal vectors in local areas and determine them as the location of residual glue.

[0014] Furthermore, the moving speed is negatively correlated with the depth of the door gap.

[0015] The present invention also discloses an automatic leak repairing robot, which adopts the leak repairing method and comprises: a walking trolley, an anti-interference module, a feeding module and a spraying module; The trolley moves along the length of the carriage. Anti-collision radars are installed at the front and rear ends of the trolley. A recognition camera is installed on the trolley to identify the carriage number. The anti-interference module is set on the top of the walking trolley, and is used to block external light; The feeding module and the spraying module are arranged on the traveling trolley, and the output end of the feeding module is connected to the input end of the spraying module; The feeding module outputs the leak-sealing material to the spraying module, and the spraying module sprays the leak-sealing material onto the gap of the vehicle door.

[0016] Furthermore, the anti-interference module includes: The main frame is installed on the top of the walking trolley. A plurality of rotating shafts are symmetrically arranged on both sides of the main frame. The rotating shafts are rotatably connected to the main frame; A wing plate is provided on the rotating shaft, and the wing plate can rotate along with the rotating shaft; Adjacent rotating shafts on the same side of the main frame are connected via a universal joint; A pair of driving motors, the output ends of which are respectively connected to the rotating shafts on both sides of the main support, can drive the corresponding rotating shafts to rotate.

[0017] Furthermore, the feeding module comprises: an insulation box, a temperature control device and a feeding barrel; The thermal insulation box and temperature control device are arranged on the traveling trolley; The temperature control device is connected to the interior of the insulation box to control the temperature inside the insulation box; The feeding barrel is arranged in the heat preservation box and is connected with the spraying module. The feeding barrel stores and outputs the leak repairing material.

[0018] Furthermore, the spraying module includes: a six-axis robotic arm, a spray gun, and a collection camera; The six-axis robotic arm is set on the walking trolley; The spray gun and acquisition camera are fixed at the output end of the six-axis robotic arm; The acquisition camera collects door gap information; The six-axis robotic arm drives the spray gun to spray at the gap of the car door.

[0019] According to the embodiments of the present invention, the automatic leak repairing robot and the leak repairing method thereof adopt machine vision perception technology, innovate the program algorithm of 2D camera images and 3D radar point cloud data, and realize accurate identification of the length, width and position information of the door gap of the carriage; through noise filtering, data collection and intelligent compensation and other processing methods, interference factors such as dirt, residual glue and ambient light changes in the door gap of the carriage are solved, thereby effectively improving the recognition rate of the door gap of the carriage.

[0020] It is to be understood that both the foregoing general description and the following detailed description are exemplary, and are intended to provide further explanation of the technology as claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1Flowchart of a leak repairing method of an automatic leak repairing robot according to an embodiment of the present invention; Figure 2 A first flow chart of the method for recognizing door gap information and residual glue information by the automatic leak repairing robot according to an embodiment of the present invention; Figure 3 A second flow chart of the method for recognizing door gap information and residual glue information by the automatic leak repairing robot according to an embodiment of the present invention; Figure 4 2 is a schematic structural diagram of an automatic leak repairing robot according to an embodiment of the present invention; Figure 5 is a side view schematic diagram of an automatic leak repairing robot according to an embodiment of the present invention; Figure 6 Schematic diagram of an anti-interference module of an automatic leak repairing robot according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings to further illustrate the present invention.

[0023] First, combine Figures 1 to 6 An automatic leak repairing robot and a leak repairing method thereof according to an embodiment of the present invention are described, which are used for repairing leaks in door seams of carriages and have a wide range of application scenarios.

[0024] like Figures 1 to 6As shown, the leak repairing method of the automatic leak repairing robot of the embodiment of the present invention, in this embodiment, the leak repairing robot includes two groups and four sets, which perform fully automatic leak repairing operations on multiple carriages on one side at the same time. Before the leak repairing operation, the leak repairing robot is in the waiting area for work. After the leak repairing starts, the following steps are performed: S1, the leak repairing robot moves along the track to the target carriage, locates the carriage position and identifies the carriage identification number through the scanning device, and the scanning device is a laser radar; S2, calls the pre-stored carriage door parameters according to the carriage identification number; S3, moves to the door position according to the carriage door parameters, and identifies the door gap information and residual glue information; S4, adjusts the moving speed according to the door gap information and residual glue information to spray and repair the door gap, and the moving speed is negatively correlated with the depth of the door gap. Using machine vision technology, the location and depth of door gaps requiring patching are identified. A robotic arm then moves to the required location at varying speeds to accommodate varying gap depths. The deeper the gap, the slower the arm moves. To fill the gaps in train carriage doors with foam glue, the location of the gap must first be identified. Furthermore, the gaps vary in depth; deeper areas require more foam glue, while shallower areas require less. The robotic arm moves more slowly to apply more foam glue to deeper gaps. It moves more quickly to apply less foam glue to shallower areas. During normal operations, the gaps in carriage doors are often previously patched, so glue already exists. If the glue is partially damaged, cracked, or detached, it requires re-spraying. If the glue is intact, no re-spraying is required. S5: After patching the current door, the robot moves to the next door and repeats S3 and S4 until the entire carriage is patched. After the robot completes patching a car door, the arm returns to its original position and moves to the next car door, repeating the same process in S3 and S4. This process continues until all the door seams in a car have been patched. The robot then returns to S1 and begins patching the next car door, repeating the process in a cycle. After the robot completes patching all planned cars, it returns to its standby position at the end of the track to await the next round of instructions.

[0025] Furthermore, in this embodiment, the door gap information includes the door gap length, door gap width, door gap depth and the precise position of the door gap.

[0026] Further, if Figures 2 and 3As shown, in this embodiment, identifying door gap information and residual glue information includes the following steps: capturing image data using a 2D camera and performing noise filtering and enhancement processing to ensure data quality; using the YOLO algorithm to perform preliminary target detection on the preprocessed image to identify the approximate location of the door gap; and optimizing the YOLO detection results based on the RGB channel gradient changes in the residual glue area using the Transformer codec mechanism to identify the residual glue location. The combination of the YOLO algorithm and the Transformer codec mechanism achieves efficient preprocessing and high-precision image recognition of the train door gap dataset. This combination not only improves target detection accuracy but also significantly increases processing speed, making it suitable for real-time detection.

[0027] Further, if Figure 3As shown, in this embodiment, identifying door gap information and residual glue information also includes the following steps: collecting point cloud data of the door gap area using a 3D TOF camera and performing preprocessing, including noise filtering and denoising operations, to remove interference information introduced by ambient light changes or camera noise to ensure data quality for subsequent processing; aligning 2D image data and 3D point cloud data using a hybrid calibration method to ensure that the two different types of data can be fused and processed in the same coordinate system, providing a basis for subsequent comprehensive analysis; and performing culling, expansion, statistical filtering, and gridding on the point cloud data to extract depth and normal vector features. In this embodiment, culling and expansion of point cloud data include: removing invalid or duplicate data points; and reasonably expanding missing point clouds based on the geometric characteristics of the door gap area to ensure data integrity and continuity and avoid detection errors caused by missing data. In this embodiment, statistical filtering includes: analyzing the distribution characteristics of the point cloud data and removing noise points that deviate from the normal range based on a preset threshold; and further optimizing the quality of the point cloud data to provide more accurate data support for subsequent feature extraction. The optimized point cloud data is gridded to extract depth and normal features. By analyzing the geometric structure of the point cloud data, depth and normal information for each point is extracted. These features effectively reflect the shape and surface condition of the door gap area. The location of the residual glue is analyzed based on changes in depth and normal, and the coordinate information of the door gap and the residual glue is output. In this embodiment, analyzing the residual glue location based on changes in depth and normal includes the following steps: comparing the depth and normal baseline features of a normal door gap area; detecting sudden changes in depth or abnormal changes in the normal direction in a local area and identifying them as the location of residual glue. Because the presence of residual glue causes local changes in depth and normal, comparing the features of the normal door gap area allows the location of the residual glue to be accurately identified, maintaining high recognition accuracy even in complex conditions such as dirt or changes in ambient light. Based on this analysis, the detection results are updated, and the world coordinate positions of the door gap and the residual glue are output, providing important reference for subsequent processing or repair work. Through this continuous detection process, the system maintains continuous and stable operation, effectively improving the recognition rate and stability of door gap detection.

[0028] In this embodiment, since there may be interference factors such as surface dirt, residual glue, and changes in ambient light in the door gaps of the vehicle compartment, the recognition accuracy is significantly improved by integrating multiple technologies, ensuring the stability and reliability of the system under complex working conditions.

[0029] Furthermore, in this embodiment, aligning 2D image data and 3D point cloud data through a hybrid calibration method includes the following steps: determining the transformation matrix between the 2D image data and the 3D point cloud data through feature matching; and mapping the 2D image data and the 3D point cloud data to the same coordinate system using the transformation matrix.

[0030] like Figures 4-6 As shown, in this embodiment, the present invention further discloses an automatic leak repairing robot, which adopts the above leak repairing method and includes: a walking car 1, an anti-interference module 2, a feeding module and a spraying module.

[0031] Specifically, if Figures 4-6 As shown, in this embodiment, the walking trolley 1 moves along the length direction of the carriage. The walking trolley 1 is a self-propelled trolley that uses a servo motor as a power source and can accurately control the moving speed. Anti-collision radars 11 are installed at the front and rear ends of the walking trolley 1. An identification camera 12 is installed on the walking trolley 1. The identification camera 12 is used to identify the carriage number so as to retrieve the door gap information in the database, which helps the walking trolley 1 to quickly locate the door position; the anti-interference module 2 is arranged on the top of the walking trolley 1, and the anti-interference module 2 is used to block external light. Since there is a certain distance between the robot and the carriage, strong sunlight will affect the recognition effect of the identification camera 12, so the anti-interference module 2 is required to block it to reduce the interference of external light; the feeding module and the spraying module are arranged on the walking trolley 1, and the output end of the feeding module is connected to the input end of the spraying module; the feeding module outputs the leak-proof material to the spraying module, and the spraying module sprays the leak-proof material to the door gap. The leak-proof material is sealant, filler, etc.

[0032] Further, if Figures 4-6 As shown, in this embodiment, the anti-interference module 2 includes: a main body bracket 21, the main body bracket 21 is installed on the top of the walking trolley 1, and a number of rotating shafts 22 are symmetrically arranged on both sides of the main body bracket 21, and the rotating shafts 22 are rotatably connected to the main body bracket 21; a wing plate 23 is provided on the rotating shaft 22, and the wing plate 23 can rotate with the rotating shaft 22; the adjacent rotating shafts 22 on the same side of the main body bracket 21 are connected by a universal joint 24; a pair of driving motors 25, the output ends of the pair of driving motors 25 are respectively connected to the rotating shafts 22 on both sides of the main body bracket 21, and can drive the corresponding rotating shafts 22 to rotate, and the wing plates 23 are opened or closed by rotating to adapt to different light conditions. A long brush 26 is also provided on the outside of the wing plate 23 to achieve flexible contact with the car body.

[0033] Further, if Figures 4-6 As shown, in this embodiment, the feeding module includes an insulated box 31, a temperature control device 32, and a feeding barrel 33. The insulated box 31 and the temperature control device 32 are mounted on the traveling carriage 1. The temperature control device 32 is connected to the interior of the insulated box 31 to control the temperature therein. The feeding barrel 33 is located within the insulated box 31 and is connected to the spraying module. The feeding barrel 33 stores and delivers the leak-proof material. In this embodiment, the temperature control device 32 is an air conditioner, and two feeding barrels 33 are used to meet the feeding requirements of a full eight-hour shift.

[0034] Further, if Figures 4-6As shown, in this embodiment, the spraying module includes: a six-axis robotic arm 41, a spray gun 42 and a collection camera 43; the six-axis robotic arm 41 is arranged on the walking vehicle 1, and the six-axis robotic arm 41 has six rotating joints, which can realize complex spatial movement and positioning, and can move and operate in a large range; the spray gun 42 and the collection camera 43 are fixed at the output end of the six-axis robotic arm 41; the collection camera 43 collects the door gap information; the six-axis robotic arm 41 drives the spray gun 42 to align with the door gap for spraying.

[0035] Above, refer to Figures 1 to 6 An automatic leak-patching robot and its leak-patching method according to an embodiment of the present invention are described. Based on machine vision perception technology, an innovative algorithm for integrating 2D camera images and 3D radar point cloud data is used to accurately identify the length, width, and position information of door gaps in carriages. Interference factors such as dirt, residual glue, and changes in ambient light in door gaps are resolved through noise filtering, data acquisition, and intelligent compensation, effectively improving the recognition rate of door gaps in carriages.

[0036] It should be noted that, in this specification, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, the elements defined by the phrase "comprising..." do not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the elements.

[0037] Although the present invention has been described in detail through the above preferred embodiments, it should be understood that the above description is not intended to limit the present invention. After reading the above description, various modifications and substitutions of the present invention will become apparent to those skilled in the art. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A leak repairing method using an automatic leak repairing robot, characterized in that: The following steps are included: S1, the leak repair robot moves along the track to the target carriage, locates the carriage position and recognizes the carriage identification number through the scanning device; S2. Calling pre-stored carriage door parameters according to the carriage identification number; S3. Move to the door position according to the door parameters and identify the door gap information and residual glue information; S4. Adjust the moving speed according to the door gap information and residual glue information to spray and seal the door gap; S5. After completing the current door leak repair, move to the next door and execute S3 and S4 in a loop until the entire carriage leak repair operation is completed.

2. The leak repairing method of the automatic leak repairing robot according to claim 1, characterized in that: The door gap information includes the door gap length, door gap width, door gap depth and the precise position of the door gap.

3. The leak repairing method of the automatic leak repairing robot according to claim 1, characterized in that: The identification of door gap information and residual glue information includes the following steps: Collect image data through 2D camera and perform preprocessing; Use the YOLO algorithm to perform preliminary target detection on the preprocessed image and identify the approximate location of the door gap; Based on the gradient changes of the RGB channels in the residual glue area, the YOLO detection results are optimized through the Transformer encoding and decoding mechanism to identify the residual glue location.

4. The leak repairing method of the automatic leak repairing robot according to claim 3, characterized in that: The identification of door gap information and residual glue information further includes the following steps: Use a 3D camera to collect point cloud data of the door gap area and perform preprocessing; Align 2D image data and 3D point cloud data through a hybrid calibration method; Eliminate, expand, statistically filter, and grid the point cloud data to extract depth and normal vector features; The residual glue position is analyzed based on the changes in depth and normal vector, and the coordinate information of the door gap and residual glue is output.

5. The leak repairing method of the automatic leak repairing robot according to claim 4, characterized in that: The hybrid calibration method for aligning 2D image data and 3D point cloud data comprises the following steps: Determine the transformation matrix between 2D image data and 3D point cloud data through feature matching; The 2D image data and the 3D point cloud data are mapped to the same coordinate system using the transformation matrix.

6. The leak repairing method of the automatic leak repairing robot according to claim 4, characterized in that: Analyzing the residual glue position according to the change of depth and normal vector includes the following steps: Compare the depth and normal vector baseline features of the normal door gap area; Detect sudden depth changes or abnormal direction changes of normal vectors in local areas and determine them as the location of residual glue.

7. An automatic leak repairing robot, characterized in that: The leak repair method according to any one of claims 1 to 6 comprises: a walking trolley, an anti-interference module, a feeding module and a spraying module; The walking trolley moves along the length of the carriage. Anti-collision radars are installed at the front and rear ends of the walking trolley. An identification camera is installed on the walking trolley to identify the carriage number. The anti-interference module is arranged on the top of the walking trolley, and the anti-interference module is used to block external light; The feeding module and the spraying module are arranged on the walking trolley, and the output end of the feeding module is connected to the input end of the spraying module; The feeding module outputs the leak-sealing material to the spraying module, and the spraying module sprays the leak-sealing material onto the gap of the vehicle door.

8. The automatic leak repairing robot according to claim 7, characterized in that: The anti-interference module includes: A main body bracket, the main body bracket is mounted on the top of the walking trolley, and a plurality of rotating shafts are symmetrically provided on both sides of the main body bracket, and the rotating shafts are rotatably connected to the main body bracket; The rotating shaft is provided with a wing plate, and the wing plate can rotate along with the rotating shaft; The adjacent rotating shafts on the same side of the main frame are connected via a universal joint; A pair of driving motors, wherein the output ends of the pair of driving motors are respectively connected to the rotating shafts on both sides of the main frame, and can drive the corresponding rotating shafts to rotate.

9. The automatic leak repairing robot according to claim 7, characterized in that: The feeding module comprises: an insulation box, a temperature control device and a feeding barrel; The thermal insulation box and the temperature control device are arranged on the walking trolley; The temperature control device is in communication with the interior of the heat preservation box to control the temperature inside the heat preservation box; The material supply barrel is arranged in the heat preservation box and is connected to the spraying module. The material supply barrel stores and outputs the leak-proof material.

10. The automatic leak repairing robot according to claim 7, characterized in that: The spraying module includes: a six-axis robotic arm, a spray gun and a collection camera; The six-axis robotic arm is arranged on the walking trolley; The spray gun and the acquisition camera are fixed to the output end of the six-axis robotic arm; The acquisition camera acquires door gap information; The six-axis robotic arm drives the spray gun to aim at the door gap for spraying.

Images