Non-excavation repair method for local collapse of pipeline based on robot technology

Abstract

The invention relates to a trenchless repair method for local collapse of a pipeline based on a robot technology. The trenchless repair method comprises the following steps: S1, collecting a pipe wall image in real time and transmitting the pipe wall image to a detection platform for preprocessing; s2, on the basis of the image recognition model, recognizing a local collapse area of the pipeline, and marking a collapse position; s3, laser point cloud and ultrasonic data of the local collapse area of the pipeline are obtained and transmitted to the detection platform, and a three-dimensional model of the local collapse area of the pipeline is constructed; s4, according to the three-dimensional model of the local collapse area of the pipeline, obtaining detection damage characteristics, and evaluating the collapse degree; s5, grading analysis is carried out based on the collapse degree evaluation result, and a repair strategy is made; and S6, according to the repair strategy, the pipeline repair robot executes automatic repair, equipment parameters are dynamically adjusted in combination with real-time sensing data, and the optimal repair effect is ensured. State monitoring and automatic repairing of the whole life cycle of the pipeline are achieved.

Description

Technical Field

[0001] The present invention relates to the field of robots, and in particular to a trenchless repair method for local pipeline collapse based on robot technology. Background Art

[0002] With the rapid development of urban infrastructure and the extensive application of industrial pipelines, pipelines, as important carriers of energy, resources and information transmission, have a direct impact on the reliable operation of public services and industrial production due to their safety and stability. However, during long-term operation, pipelines are susceptible to damage such as local collapse, loss of roundness, corrosion perforation, etc. due to factors such as material aging, corrosion, external mechanical pressure, and geological disasters.

[0003] Traditional pipeline repair methods rely on large-scale excavation, manual inspection and on-site processing and repair. The operation is complex, the cycle is long, the interference is large and the cost is high. Especially in high-density urban areas or deep buried pipeline scenarios, its application has obvious limitations. In addition, the traditional method lacks high-precision data support and is prone to detection errors or insufficient repair problems. Summary of the invention

[0004] In order to solve the above problems, the purpose of the present invention is to provide a trenchless repair method for local pipeline collapse based on robotics technology, so as to realize status monitoring and automated repair of the pipeline throughout its life cycle.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A trenchless repair method for local pipeline collapse based on robot technology includes the following steps: S1: The image acquisition device carried by the patrol robot collects the pipe wall image in real time and transmits it to the detection platform for preprocessing; S2: The detection platform identifies the local collapsed area of ​​the pipeline based on the image recognition model and marks the collapsed location; S3: Based on the laser radar and ultrasonic sensor carried by the patrol robot, the laser point cloud and ultrasonic data of the local pipeline collapse area are obtained and transmitted to the detection platform to construct a three-dimensional model of the local pipeline collapse area; S4: According to the three-dimensional model of the local collapsed area of ​​the pipeline, the detection damage characteristics are obtained and the collapse degree is evaluated; S5: Conduct graded analysis based on the collapse assessment results and formulate a repair strategy; S6: According to the repair strategy, the pipeline repair robot performs automated repair, including material laying and UV operation, and combines real-time sensor data to dynamically adjust equipment parameters to ensure the best repair effect.

[0006] Furthermore, the preprocessing is specifically as follows: Use the perspective transformation matrix to correct the image tilt or deformation caused by camera movement: ; Where (x, y) is the pixel coordinate of the input image. is the output image coordinate; is the scale factor of the homogeneous coordinates after perspective transformation; is the perspective transformation matrix; The perspective transformation matrix H must satisfy the correspondence between the input point (x, y) and the output point (x′, y′). Substitute the mapping of the four pairs of points to establish a linear equation system and solve H: Get the coordinates of the four key points of the image before transformation (x1, y1), (x2, y2), (x3, y3), (x4, y4); set the coordinates of the four key points of the target image (x1′, y1′), (x2′, y2′), (x3′, y3′), (x4′, y4′); use OpenCV to calculate the transformation matrix ; Perform perspective correction on the image according to the calculated perspective transformation matrix H; The image grayscale value is evenly distributed through histogram equalization, the grayscale histogram of the original image is statistically analyzed, the cumulative distribution function is calculated, and the grayscale mapping is performed to obtain the enhanced image; the enhanced image is processed using Gaussian filtering to eliminate noise interference.

[0007] Furthermore, the image recognition model is used to identify the local collapsed area of ​​the pipeline and mark the collapsed location, as follows: Based on the preprocessed image data, the collapsed area is located using the object detection algorithm YOLO; YOLO divides the input image into S×S grids, and predicts B candidate boxes for each grid, as well as the confidence and category information corresponding to each candidate box; Each candidate box contains the following attributes: the center coordinates of the candidate box (a, b); the width w and height h of the candidate box; the confidence score C, which indicates whether the box contains the target and the accuracy of the box; the classification probability, that is, the probability that the target belongs to the collapsed box; Model training is performed using YOLO’s loss function; The candidate box of the identified collapsed area is converted into pixel coordinates, the position of the collapsed area is marked, and the position coordinate information and visual image are provided.

[0008] Furthermore, the loss function of YOLO consists of three parts: L=L coord +L conf +Lclass ; ; ; ; Among them, L, L coord , L conf and L class They are YOLO's loss function, coordinate loss, confidence loss, and classification loss; is the center coordinate of the predicted candidate box; and They are the predicted candidate box width and height respectively; Indicates the probability that the target belongs to category c; The probability that the predicted target belongs to category c (collapse); is an indicator variable, indicating that the jth bounding box in the i-th grid contains the object; is an indicator variable, indicating that the jth bounding box in the i-th grid does not contain the object; and is the weight parameter; is the true confidence of the i-th grid, is the prediction confidence of the i-th grid.

[0009] Furthermore, the S3 is specifically: The laser radar scans the pipeline to generate point cloud data, obtains the depth information of the pipeline through the ultrasonic sensor, and transmits the data to the detection platform through wireless communication; Based on the detection results of the target detection algorithm YOLO, the two-dimensional image bounding box of the collapsed area is determined. According to the external parameter calibration of the image and the lidar, the image coordinates are projected to the lidar coordinate system. The three-dimensional point cloud area is obtained by projection, and the point cloud data is cropped to retain the data of the collapsed area. The point cloud is reconstructed into a three-dimensional surface, and the discrete points are converted into a continuous surface. According to the surface normal vector of the point cloud, Poisson reconstruction is performed to generate a smooth surface. The optimization equation is as follows: ; Among them, ρ is the distribution of point cloud density field; is the Laplace operator; u represents the smooth curve of the reconstructed surface; The depth information collected by ultrasound is integrated into the point cloud data to supplement the depth value of the laser point cloud and construct a three-dimensional model of the local collapse area of ​​the pipeline.

[0010] Furthermore, the S4 is specifically: The three-dimensional model of the local collapsed area of ​​the pipeline is aligned with the original complete pipeline model, and the standard cylindrical surface is fitted by the least square error; ; in, is the cumulative error, is the amount of point cloud data, is a standard surface point, For the The location of the actual sampling points; Extract geometric and morphological features from the 3D model of the local collapsed area of ​​the pipeline as the basis for assessing the extent of the collapse damage, including the degree of out-of-roundness of the pipeline. , collapse depth and the volume of the collapsed area ; ; ; ; in, is the maximum radius of the cross section; is the minimum radius of the cross section; is the average radius of the standard section; For sampling point The height value of is the original pipe surface to be fitted; Represents the three-dimensional spatial integral of the entire collapsed area collapse; Based on the extracted geometric and morphological features, the degree of collapse is comprehensively quantified: ; in, , and is the weight coefficient, and the collapse degree is divided into mild collapse, moderate collapse and severe collapse according to DI; is the original pipe volume, is the collapse degree value.

[0011] Furthermore, the S5 is specifically: For mild collapse, epoxy resin is sprayed on the surface of the collapsed area to fill small cracks or depressions, improve the quality of the inner surface of the pipe, and UV curing coating materials are sprayed by robots and quickly cured in the collapsed area; For moderate collapse, use a glass fiber reinforced composite patch to fit inside the collapsed area for enhanced fixation, and use a resin-based filling material to expand and fill the cracks in the collapse; In case of severe collapse, install an expandable composite liner pipe in the collapsed area, apply pressure to make the liner material fit against the original pipe wall, and restore the bearing capacity of the collapsed position. Use a steel sleeve outer clamp to repair pipes with large cracks and a roundness loss rate exceeding the critical value.

[0012] Furthermore, the S6 is specifically: Based on the repair strategy, pipeline repair robots are used to complete automated repair operations, including material laying and UV / heat curing operations; Adjust the repair robot's repair parameters through real-time sensor data combined with robot feedback; After the robot repair is completed, the real-time point cloud data is compared with the point cloud data before repair, and a post-repair report is generated, recording the repair quality, time and cost.

[0013] Furthermore, the repair parameters of the repair robot are adjusted by combining the real-time sensor data with the robot feedback, as follows: Adjust the spray direction and spray head movement speed through real-time sensor data combined with robot feedback; ; Among them, v spray is the spraying speed; Q is the total amount of spraying material; A zone is the area of ​​the repaired area; h desired is the target spraying thickness; If abnormal temperature and humidity are detected, adjust the output power of the UV lamp: ; Among them, P adjust is the adjusted power; P initial is the initial power; T optimal , T current is the target and current temperature.

[0014] A trenchless repair system for local pipeline collapse based on robotics technology includes a processor, a memory, and a computer program stored in the memory. When the processor executes the computer program, it specifically performs the steps of the trenchless repair method for local pipeline collapse based on robotics technology as described above.

[0015] The present invention has the following beneficial effects: 1. The present invention combines image recognition technology to effectively reduce the interference of complex factors in the pipeline detection environment on the visual system, ensure the accuracy and high reliability of the detection results, and accurately identify the location and range of local collapse of the pipeline; 2. The present invention uses an image recognition model to quickly locate the collapsed area, mark its specific location, and integrates the high-precision spatial measurement of laser radar and the thickness detection capabilities of ultrasound to form a fine-grained and accurate local pipeline collapse model; 3. The damaged area of ​​the present invention is graded and evaluated according to the degree of collapse (mild, moderate, severe), and a targeted repair plan is formulated to avoid waste of resources. After the repair is completed, the inspection robot scans the repair area again, and compares the three-dimensional model before and after the repair through point cloud and image to evaluate the repair effect; a repair quality assessment report is automatically generated based on the data, recording the key parameters involved in the repair, such as time, cost, laying thickness and error range, to ensure the controllability and traceability of the repair quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0017] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments: refer to Figure 1 In this embodiment, a trenchless repair method for local pipeline collapse based on robot technology is provided, including the following steps: S1: The image acquisition device carried by the patrol robot collects the pipe wall image in real time and transmits it to the detection platform for preprocessing; S2: The detection platform identifies the local collapsed area of ​​the pipeline based on the image recognition model and marks the collapsed location; S3: Based on the laser radar and ultrasonic sensor carried by the patrol robot, the laser point cloud and ultrasonic data of the local pipeline collapse area are obtained and transmitted to the detection platform to construct a three-dimensional model of the local pipeline collapse area; S4: According to the three-dimensional model of the local collapsed area of ​​the pipeline, the detection damage characteristics are obtained and the collapse degree is evaluated; S5: Conduct graded analysis based on the collapse assessment results and formulate a repair strategy; S6: According to the repair strategy, the pipeline repair robot performs automated repair, including material laying and UV operation, and dynamically adjusts equipment parameters in combination with real-time sensor data (pipeline environmental parameters, fitting conditions) to ensure the best repair effect.

[0018] In this implementation, the preprocessing is specifically as follows: Use the perspective transformation matrix to correct the image tilt or deformation caused by camera movement: ; Where (x, y) is the pixel coordinate of the input image. is the output image coordinate; is the scale factor of the homogeneous coordinates after perspective transformation; is the perspective transformation matrix; The perspective transformation matrix H must satisfy the correspondence between the input point (x, y) and the output point (x′, y′). Substitute the mapping of the four pairs of points to establish a linear equation system and solve H: Get the coordinates of the four key points of the image before transformation (x1, y1), (x2, y2), (x3, y3), (x4, y4); set the coordinates of the four key points of the target image (x1′, y1′), (x2′, y2′), (x3′, y3′), (x4′, y4′); use OpenCV to calculate the perspective transformation matrix ; According to the calculated perspective transformation matrix H, the image Perform perspective correction.

[0019] The image grayscale value is evenly distributed through histogram equalization, the grayscale histogram of the original image is statistically analyzed, the cumulative distribution function is calculated, and the grayscale mapping is performed to obtain the enhanced image; the enhanced image is processed using Gaussian filtering to eliminate noise interference.

[0020] In this implementation, based on the image recognition model, the local collapse area of ​​the pipeline is identified and the collapse location is marked, as follows: Based on the preprocessed image data, the collapsed area is located using the object detection algorithm YOLO; YOLO divides the input image into S×S grids, each of which predicts B candidate boxes, the confidence level (whether there is an object) and category information (such as the "collapse" of the pipeline) corresponding to each candidate box; Each candidate box contains the following attributes: the center coordinates of the candidate box (a, b); the width w and height h of the candidate box; the confidence score C, which indicates whether the box contains the target and the accuracy of the box; the classification probability, that is, the probability that the target belongs to the collapsed box; Model training is performed using YOLO’s loss function; The candidate box of the identified collapsed area is converted into pixel coordinates, the position of the collapsed area is marked, and the position coordinate information and visual image are provided.

[0021] In this implementation, YOLO's loss function consists of three parts: L=L coord +L conf +L class ; ; ; ; Among them, L, L coord , L conf and L class They are YOLO's loss function, coordinate loss, confidence loss, and classification loss; is the center coordinate of the predicted candidate box; and They are the predicted candidate box width and height respectively; Indicates the probability that the target belongs to category c; The probability that the predicted target belongs to category c (collapse); is an indicator variable, indicating that the jth bounding box in the i-th grid contains the object; is an indicator variable, indicating that the jth bounding box in the i-th grid does not contain the object; and is the weight parameter; is the true confidence of the i-th grid, is the prediction confidence of the i-th grid; In this implementation, S3 is specifically: The laser radar scans the pipeline to generate point cloud data, obtains the depth information of the pipeline through the ultrasonic sensor, and transmits the data to the detection platform through wireless communication; Based on the detection results of the target detection algorithm YOLO, the two-dimensional image bounding box of the collapsed area is determined. According to the external parameter calibration of the image and the lidar, the image coordinates are projected to the lidar coordinate system. The three-dimensional point cloud area is obtained by projection, and the point cloud data is cropped to retain the data of the collapsed area. The point cloud is reconstructed into a three-dimensional surface, and the discrete points are converted into a continuous surface. According to the surface normal vector of the point cloud, Poisson reconstruction is performed to generate a smooth surface. The optimization equation is as follows: ; Among them, ρ is the distribution of point cloud density field; is the Laplace operator; u represents the smooth curve of the reconstructed surface; The depth information collected by ultrasound is integrated into the point cloud data to supplement the depth value of the laser point cloud and construct a three-dimensional model of the local collapse area of ​​the pipeline.

[0022] In this implementation, S4 is specifically: The three-dimensional model of the local collapsed area of ​​the pipeline is aligned with the original complete pipeline model, and the standard cylindrical surface is fitted by the least square error; ; in, is the cumulative error, is the amount of point cloud data, is a standard surface point, For the The location of the actual sampling points; Extract geometric and morphological features from the 3D model of the local collapsed area of ​​the pipeline as the basis for assessing the extent of the collapse damage, including the degree of out-of-roundness of the pipeline. , collapse depth and the volume of the collapsed area ; ; ; ; in, is the maximum radius of the cross section; is the minimum radius of the cross section; is the average radius of the standard section; For sampling point The height value of is the original pipe surface to be fitted; Represents the three-dimensional spatial integral of the entire collapsed area collapse; Based on the extracted geometric and morphological features, the degree of collapse is comprehensively quantified: ; in, , and is the weight coefficient, and the collapse degree is divided into mild collapse, moderate collapse and severe collapse according to DI; is the original pipe volume, is the collapse degree value.

[0023] Mild collapse: loss of roundness < 10%, collapse depth < 10%, volume deviation < 10%.

[0024] Moderate collapse: loss of roundness 10%-30%, collapse depth 10%-30%, volume deviation 10%-30%.

[0025] Severe collapse: loss of roundness > 30%, or wall thickness thinning > 30%, or collapse volume deviation > 30%.

[0026] In this implementation, S5 is specifically: For mild collapse, corrosion-resistant, high-strength epoxy resin is sprayed on the surface of the collapsed area to fill small cracks or depressions and improve the quality of the inner surface of the pipe. UV-curable coating materials are sprayed by robots and quickly cured in the collapsed area; For moderate collapse, use glass fiber reinforced composite (GFRP) patches to fit inside the collapsed area for enhanced fixation, and use resin-based filling materials to expand and fill the cracks in the collapse; In case of severe collapse, install an expandable composite liner pipe in the collapsed area, apply pressure to make the liner material fit against the original pipe wall, and restore the bearing capacity of the collapsed position. Use a steel sleeve outer clamp to repair pipes with large cracks and a roundness loss rate exceeding the critical value.

[0027] In this implementation, S6 specifically includes: Based on the repair strategy, pipeline repair robots are used to complete automated repair operations, including material laying and UV / heat curing operations; Adjust the repair robot's repair parameters through real-time sensor data combined with robot feedback; After the robot repair is completed, the real-time point cloud data is compared with the point cloud data before repair, and a post-repair report is generated, recording the repair quality, time and cost.

[0028] In this implementation, the repair parameters of the repair robot are adjusted by combining real-time sensor data with robot feedback, as follows: Adjust the spray direction and spray head movement speed through real-time sensor data combined with robot feedback; ; Among them, v spray is the spraying speed; Q is the total amount of spraying material; A zone is the area of ​​the repaired area; h desired is the target spraying thickness; If abnormal temperature and humidity are detected, adjust the output power of the UV lamp: ; Among them, P adjust is the adjusted power; P initial is the initial power; T optimal , T current is the target and current temperature.

[0029] In this embodiment, a trenchless repair system for partial collapse of a pipeline based on robotic technology is also provided, comprising a processor, a memory, and a computer program stored on the memory. When the processor executes the computer program, it specifically performs the steps of the trenchless repair method for partial collapse of a pipeline based on robotic technology as described above.

[0030] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0031] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0032] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0033] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0034] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any technician familiar with the profession may use the above disclosed technical content to change or modify it into an equivalent embodiment with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention still belongs to the protection scope of the technical solution of the present invention.

Claims

1. A trenchless repair method for local pipeline collapse based on robot technology, characterized in that: The following steps are involved: S1: The image acquisition device carried by the patrol robot collects the pipe wall image in real time and transmits it to the detection platform for preprocessing; S2: The detection platform identifies the local collapsed area of ​​the pipeline based on the image recognition model and marks the collapsed location; S3: Based on the laser radar and ultrasonic sensor carried by the patrol robot, the laser point cloud and ultrasonic data of the local pipeline collapse area are obtained and transmitted to the detection platform to construct a three-dimensional model of the local pipeline collapse area; S4: According to the three-dimensional model of the local collapsed area of ​​the pipeline, the detection damage characteristics are obtained and the collapse degree is evaluated; S5: Conduct graded analysis based on the collapse assessment results and formulate a repair strategy; S6: According to the repair strategy, the pipeline repair robot performs automated repair, including material laying and UV operation, and dynamically adjusts equipment parameters in combination with real-time sensor data to ensure the best repair effect; The S4 is specifically: The three-dimensional model of the local collapsed area of ​​the pipeline is aligned with the original complete pipeline model, and the standard cylindrical surface is fitted by the least square error; ； in, is the cumulative error, is the amount of point cloud data, is a standard surface point, For the The location of the actual sampling points; Extract geometric and morphological features from the 3D model of the local collapsed area of ​​the pipeline as the basis for assessing the extent of the collapse damage, including the degree of out-of-roundness of the pipeline. , collapse depth and the volume of the collapsed area ; ； ； ； in, is the maximum radius of the cross section; is the minimum radius of the cross section; is the average radius of the standard section; For sampling point The height value of is the original pipe surface to be fitted; Represents the three-dimensional spatial integral of the entire collapsed area collapse; Based on the extracted geometric and morphological features, the degree of collapse is comprehensively quantified: ； in, , and is the weight coefficient, and the collapse degree is divided into mild collapse, moderate collapse and severe collapse according to DI; is the original pipe volume, is the collapse degree value.

2. The trenchless repair method for local pipeline collapse based on robot technology according to claim 1 is characterized in that: The pre-processing is specifically as follows: Use the perspective transformation matrix to correct the image tilt or deformation caused by camera movement: ； Where (x, y) is the pixel coordinate of the input image. is the output image coordinate; is the scale factor of the homogeneous coordinates after perspective transformation; is the perspective transformation matrix; The perspective transformation matrix H must satisfy the correspondence between the input point (x, y) and the output point (x′, y′). Substitute the mapping of the four pairs of points to establish a linear equation system and solve H: Get the coordinates of the four key points of the image before transformation (x1, y1), (x2, y2), (x3, y3), (x4, y4); set the coordinates of the four key points of the target image (x1′, y1′), (x2′, y2′), (x3′, y3′), (x4′, y4′); use OpenCV to calculate the transformation matrix ; Perform perspective correction on the image according to the calculated perspective transformation matrix H; The image grayscale value is evenly distributed through histogram equalization, the grayscale histogram of the original image is statistically analyzed, the cumulative distribution function is calculated, and the grayscale mapping is performed to obtain the enhanced image; the enhanced image is processed using Gaussian filtering to eliminate noise interference.

3. The trenchless repair method for local pipeline collapse based on robot technology according to claim 1 is characterized in that: Based on the image recognition model, the local collapsed area of ​​the pipeline is identified and the collapsed location is marked, as follows: Based on the preprocessed image data, the collapsed area is located using the object detection algorithm YOLO; YOLO divides the input image into S×S grids, and predicts B candidate boxes for each grid, as well as the confidence and category information corresponding to each candidate box; Each candidate box contains the following attributes: the center coordinates of the candidate box (a, b); the width w and height h of the candidate box; the confidence score C, which indicates whether the box contains the target and the accuracy of the box; the classification probability, that is, the probability that the target belongs to the collapsed box; Model training is performed using YOLO’s loss function; The candidate box of the identified collapsed area is converted into pixel coordinates, the position of the collapsed area is marked, and the position coordinate information and visual image are provided.

4. The trenchless repair method for local pipeline collapse based on robot technology according to claim 3 is characterized in that: The loss function of YOLO consists of three parts: L=L coord +L conf +L class ； ; ； ； Among them, L, L coord , L conf and L class They are YOLO's loss function, coordinate loss, confidence loss, and classification loss; is the center coordinate of the predicted candidate box; and They are the predicted candidate box width and height respectively; Indicates the probability that the target belongs to category c; The probability that the predicted target belongs to category c; is an indicator variable, indicating that the jth bounding box in the i-th grid contains the object; is an indicator variable, indicating that the jth bounding box in the i-th grid does not contain the object; and is the weight parameter; is the true confidence of the i-th grid, is the prediction confidence of the i-th grid.

5. The trenchless repair method for local pipeline collapse based on robot technology according to claim 1 is characterized in that: The S3 is specifically: The laser radar scans the pipeline to generate point cloud data, obtains the depth information of the pipeline through the ultrasonic sensor, and transmits the data to the detection platform through wireless communication; Based on the detection results of the target detection algorithm YOLO, the two-dimensional image bounding box of the collapsed area is determined. According to the external parameter calibration of the image and the lidar, the image coordinates are projected to the lidar coordinate system. The three-dimensional point cloud area is obtained by projection, and the point cloud data is cropped to retain the data of the collapsed area. The point cloud is reconstructed into a three-dimensional surface, and the discrete points are converted into a continuous surface. According to the surface normal vector of the point cloud, Poisson reconstruction is performed to generate a smooth surface. The optimization equation is as follows: ； Among them, ρ is the distribution of point cloud density field; is the Laplace operator; u represents the smooth curve of the reconstructed surface; The depth information collected by ultrasound is integrated into the point cloud data to supplement the depth value of the laser point cloud and construct a three-dimensional model of the local collapse area of ​​the pipeline.

6. The trenchless repair method for local pipeline collapse based on robot technology according to claim 1 is characterized in that: The S5 is specifically: For mild collapse, epoxy resin is sprayed on the surface of the collapsed area to fill small cracks or depressions, improve the quality of the inner surface of the pipe, and UV curing coating materials are sprayed by robots and quickly cured in the collapsed area; For moderate collapse, use a glass fiber reinforced composite patch to fit inside the collapsed area for enhanced fixation, and use a resin-based filling material to expand and fill the cracks in the collapse; In case of severe collapse, install an expandable composite liner pipe in the collapsed area, apply pressure to make the liner material fit against the original pipe wall, and restore the bearing capacity of the collapsed position. Use a steel sleeve outer clamp to repair pipes with large cracks and a roundness loss rate exceeding the critical value.

7. The trenchless repair method for local pipeline collapse based on robot technology according to claim 1 is characterized in that: The S6 is specifically: Based on the repair strategy, pipeline repair robots are used to complete automated repair operations, including material laying and UV / heat curing operations; Adjust the repair robot's repair parameters through real-time sensor data combined with robot feedback; After the robot repair is completed, the real-time point cloud data is compared with the point cloud data before repair, and a post-repair report is generated, recording the repair quality, time and cost.

8. The trenchless repair method for local pipeline collapse based on robot technology according to claim 7 is characterized in that: The repair parameters of the repair robot are adjusted by combining real-time sensor data with robot feedback, as follows: Adjust the spray direction and spray head movement speed through real-time sensor data combined with robot feedback; ； Among them, v spray is the spraying speed; Q is the total amount of spraying material; A zone is the area of ​​the repaired area; h desired is the target spraying thickness; If abnormal temperature and humidity are detected, adjust the output power of the UV lamp: ； Among them, P adjust is the adjusted power; P initial is the initial power; T optimal , T current is the target and current temperature.

9. A non-excavation repair system for local pipeline collapse based on robotics technology, characterized in that: The method comprises a processor, a memory and a computer program stored in the memory. When the processor executes the computer program, the method specifically performs the steps in a trenchless repair method for local collapse of a pipeline based on robotic technology as described in any one of claims 1 to 8.

Images