A robot for repairing leaking pipes and a control method for its application.

Abstract

This application discloses a control method for a robot used in pipeline leak repair and its application, relating to the field of intelligent robot technology. The robot for pipeline leak repair includes a robot control mechanism that acquires pipeline inspection information and repair task requirements, plans the robot's travel path and operating mode; predicts a first predicted movement speed and a second predicted operating speed of the robot in the repair area based on the travel path and operating mode; generates a control parameter table for adjusting the robot's operating efficiency; obtains a pipeline environment reference range representing the threshold of pipeline environment changes; optimizes a preset robot operation control model based on the pipeline environment reference range and the control parameter table; controls the robot to perform pipeline repair operations based on the robot operation control model; and adjusts the robot's operating state based on real-time pipeline environment data. This application improves the efficiency of pipeline leak repair.

Description

Technical Field

[0001] This application relates to the field of intelligent robot technology, and in particular to a robot for repairing pipe leaks and a control method for its application. Background Technology

[0002] In urban operations and industrial production, pipeline systems play a crucial role in transporting fluid media (such as water, oil, and gas). However, due to various reasons such as pipe aging, corrosion, and external damage, pipe leaks occur frequently. For example, existing sewage pipe corridors are typically quite long, formed by connecting multiple pipes end-to-end with connectors. Over long-term operation, leaks may occur at the connection points between two pipes. Because sewage pipe corridors are long, and the number and density of pipes are high, inspection and maintenance are quite difficult.

[0003] The existing inspection and maintenance methods are mostly based on manual inspection and maintenance, which rely on the inspection and experience judgment of the staff. The staff need to inspect the pipeline regularly, and after discovering the leak, they will manually repair it. The manual repair method is time-consuming and laborious, and it is difficult to find hidden or small leaks. Therefore, the existing pipeline leak repair efficiency is low and needs to be improved. Summary of the Invention

[0004] To improve the efficiency of pipe leak repair, this application provides a robot for pipe leak repair and a control method for its application.

[0005] Firstly, the objective of this invention is achieved through the following technical solution:

[0006] A robot for repairing pipe leaks includes a repair robot and a robot control mechanism.

[0007] The robot control mechanism acquires pipeline inspection information and maintenance task requirements, and plans the travel path and operation mode of the maintenance robot based on the pipeline inspection information and maintenance task requirements.

[0008] Based on the travel path and operation mode, predict the first predicted moving speed and the second predicted operation speed of the maintenance robot in the operation and maintenance area;

[0009] Based on the maintenance task requirements, the first predicted moving speed and the second predicted working speed, a control parameter table is generated to adjust the working efficiency of the maintenance robot.

[0010] Obtain a pipeline environment reference range representing the threshold of pipeline environment change, and optimize the preset robot operation control model based on the pipeline environment reference range and the control parameter table;

[0011] The robot operation control model is used to control the maintenance robot to perform pipeline maintenance operations; real-time pipeline environment data is acquired and input into the maintenance robot operation control model to adjust the operation status of the maintenance robot.

[0012] By adopting the above technical solution, an intelligent pipeline maintenance robot with high maintenance efficiency and high-quality completion of maintenance tasks is provided. Specifically, by acquiring pipeline inspection information (including but not limited to pipeline material, diameter, curvature, and pipeline inspection images) and maintenance task requirements (including but not limited to maintenance type, required tools, and maintenance point location), the robot control mechanism can intelligently plan the robot's travel path and operation mode, which is conducive to improving maintenance efficiency. At the same time, by predicting the robot's first moving speed and second operating speed in the sewage pipe corridor, the robot's working structure can be adjusted according to the actual situation, generating a control parameter table for adjusting the robot's working rhythm and operating efficiency, so that the robot can complete high-quality maintenance tasks within a suitable time. By acquiring real-time pipeline environment data and making timely adjustments, adjustments can be made quickly in case of unexpected situations, ensuring the smooth progress of maintenance work. Unlike the existing technology that relies on manual inspection to find maintenance points before maintenance, this application significantly improves the efficiency and reliability of pipeline leak repair through the prediction of the robot's moving speed and optimization of operating efficiency, strong environmental adaptability, and real-time dynamic response mechanism.

[0013] In a preferred embodiment of this application: the step of acquiring pipeline inspection information and maintenance task requirements, and planning the travel path and operation mode of the maintenance robot based on the pipeline inspection information and the maintenance task requirements, specifically includes:

[0014] Obtain pipeline inspection information, including pipeline material, pipeline diameter, pipeline curvature, and location of obstructions within the pipeline; the maintenance task requirements include maintenance type, maintenance location, and required tools.

[0015] The travel path of the maintenance robot is planned based on the pipe diameter, the pipe curvature, and the location of obstacles inside the pipe; the operation mode of the maintenance robot is determined based on the maintenance type, the maintenance location, and the required tools.

[0016] By adopting the above technical solution, this application integrates pipeline inspection information (including pipeline material, diameter, curvature and obstacle location) with maintenance task requirements (such as maintenance type, location and required tools) to accurately plan the travel path and operation mode of the maintenance robot. Unlike the problem of inaccurate robot operation caused by the lack of precise path planning in the prior art, the maintenance robot of this application improves maintenance efficiency and success rate.

[0017] In a preferred embodiment of this application, the step of predicting the first predicted movement speed and the second predicted operation speed of the maintenance robot within the pipeline based on the travel path and operation mode specifically includes:

[0018] Based on the pipe diameter, the pipe curvature, and the travel path, predict the first predicted moving speed of the maintenance robot inside the pipe;

[0019] Based on the repair type, the required tools, and the operation mode, a second predicted operation speed of the repair robot at the work point is predicted.

[0020] By adopting the above technical solution, the problem of inaccurate time management caused by the inability to accurately predict the working speed of maintenance robots is effectively solved, thereby optimizing time scheduling and improving the working efficiency of maintenance robots. This application adopts a first predicted moving speed of maintenance robots based on pipe diameter, curvature, and travel path, and a second predicted working speed based on maintenance type, required tools, and work mode. Therefore, the process can pre-evaluate the performance of the robot in different environments, so as to predict the working time of maintenance robots and carry out efficient maintenance time management of maintenance robots.

[0021] In a preferred embodiment of this application, the step of generating a control parameter table for adjusting the operating efficiency of the maintenance robot based on the maintenance task requirements, the first predicted moving speed, and the second predicted operating speed specifically includes:

[0022] The total time required for the maintenance task is calculated based on the maintenance location, the first predicted movement speed, and the second predicted operation speed.

[0023] Based on the total time consumed and the maintenance task requirements, the target operating efficiency of the maintenance robot is obtained;

[0024] Based on the target work efficiency, the first predicted moving speed, and the second predicted work speed, the moving speed and work speed parameters of the maintenance robot are adjusted to generate a control parameter table for adjusting the work efficiency of the maintenance robot.

[0025] By adopting the above technical solution, in order to flexibly adjust the robot's working strategy according to the actual situation, this application adopts a method to calculate the total time based on the maintenance position, the first predicted moving speed and the second predicted working speed, and adjusts the moving speed and working speed parameters of the maintenance robot accordingly to generate a control parameter table, which effectively solves the problem of resource waste or time delay caused by fixed parameter settings, thereby maximizing the working efficiency of the maintenance robot.

[0026] Secondly, the objective of this invention is achieved through the following technical solution:

[0027] A control method for a robot used in pipe leak repair, applied to a robot for pipe leak repair as described above, the control method comprising:

[0028] Obtain pipeline inspection information and maintenance task requirements;

[0029] Based on the pipeline inspection information and maintenance task requirements, plan the travel path and operation mode of the maintenance robot;

[0030] Based on the planned travel path and operation mode, the first predicted movement speed and the second predicted operation speed of the maintenance robot in the operation and maintenance area are predicted.

[0031] Based on the maintenance task requirements, the first predicted moving speed and the second predicted working speed, a control parameter table is generated to adjust the working efficiency of the maintenance robot.

[0032] Obtain a pipeline environment reference range representing the threshold of pipeline environment change, and optimize the preset robot operation control model based on the pipeline environment reference range and the control parameter table;

[0033] The optimized robot operation control model is used to control the maintenance robot to perform pipeline maintenance operations.

[0034] During maintenance operations, real-time pipeline environment data is acquired; the real-time pipeline environment data is input into the robot operation control model to dynamically adjust the operation status of the maintenance robot.

[0035] By adopting the above technical solution, this application integrates pipeline inspection information based on the actual pipeline maintenance situation and maintenance task requirements, plans the travel path and operation mode of the maintenance robot in real time, and predicts the first predicted movement speed and the second predicted operation speed of the maintenance robot in the operation area. Therefore, the process can accurately set the best working parameters for the robot.

[0036] In a preferred embodiment of this application: the pipeline inspection information includes low-resolution images of damaged pipelines; after acquiring the pipeline inspection information, the method further includes:

[0037] Determine the forward repair relationship between multiple low-resolution damaged pipe images and the first intact pipe model, and determine the reverse verification relationship between the first intact pipe model and multiple repaired pipe images.

[0038] Based on the image recognition network corresponding to the positive repair relationship and the mechanical operation model corresponding to the reverse verification relationship, a maintenance robot operation control model is constructed, wherein the image recognition network includes a repair accuracy loss function and the machine operation model includes an operation accuracy loss function.

[0039] Obtain a training dataset in which the forward repair relationship and the reverse verification relationship have been determined; based on the training dataset, train the operation control model of the maintenance robot, wherein the repair accuracy loss function and the operation accuracy loss function are jointly optimized during training;

[0040] The multiple low-resolution images of the damaged pipe sections to be repaired are input into the trained operation control model of the repair robot to obtain high-resolution images of the repaired pipes.

[0041] By adopting the above technical solution, the forward repair relationship and reverse verification relationship between multiple low-resolution damaged pipeline images and the first intact pipeline model are determined. Based on this, a maintenance robot operation control model including repair accuracy loss function and operation accuracy loss function is constructed to accurately simulate and optimize the maintenance process, thereby improving the quality of pipeline repair.

[0042] In a preferred embodiment of this application: training the maintenance robot operation control model based on the training dataset includes:

[0043] Based on the training dataset, a training high-resolution repaired pipe image, a training low-resolution damaged pipe image, and a training intact pipe model are determined; wherein, the training low-resolution damaged pipe image and the training intact pipe model satisfy a positive repair relationship, and the training intact pipe model and the training high-resolution repaired pipe image satisfy a reverse verification relationship.

[0044] Multiple training low-resolution damaged pipe images are input into the maintenance robot operation control model. The image recognition network outputs a test recovery model based on the multiple training low-resolution damaged pipe images. The mechanical operation model outputs multiple test-repaired pipe images based on the test repair model.

[0045] Based on the trained intact pipeline model and the test repair model, the repair accuracy loss function is determined;

[0046] Based on the training high-resolution repaired pipeline image and the test repaired pipeline image, the operation accuracy loss function is determined;

[0047] Based on the repair accuracy loss function and the operation accuracy loss function, the total maintenance loss function is determined;

[0048] The total maintenance loss function is optimized to train the maintenance robot operation control model until the maintenance robot operation control model is fully trained.

[0049] By adopting the above technical solutions, the generalization ability and problem-solving ability of the maintenance robot operation control model are improved. This application adopts a method based on training datasets to determine the training of high-definition repaired pipe images, low-definition damaged pipe images, and intact pipe models, and trains the maintenance robot operation control model by jointly optimizing the repair accuracy loss function and the operation accuracy loss function. This makes the maintenance robot operation control model have good adaptability and accuracy to various pipe damage conditions.

[0050] In a preferred embodiment of this application: the mechanical operation model outputs multiple segments of the tested and repaired pipeline images based on the test and repair model, including:

[0051] The test repair model is input into the mechanical operation model to obtain multiple segments of the pipeline after the test repair. The mechanical operation model is a mechanical model of the simulated repair process constructed based on the robot operation path and the application parameters of the repair tool.

[0052] By adopting the above technical solution, a method is used to generate a pipeline image after test repair by inputting the test repair model into a mechanical model constructed based on the robot operation path and repair tool application parameters. This allows for dynamic adjustment of the repair strategy according to the specific circumstances of the actual damage to the pipeline, enabling the maintenance robot to flexibly cope with different types of damage and optimize the repair results.

[0053] In a preferred embodiment of this application: the step of inputting the test repair model into the mechanical operation model to obtain multiple segments of the tested and repaired pipeline images includes:

[0054] The damage type and location information corresponding to the low-resolution damaged pipe image input to the image recognition network, and the test repair model output by the image recognition network are input to the mechanical operation model to obtain multiple segments of the tested and repaired pipe images. The mechanical operation model adjusts the robot operation path according to the damage type, location information and application parameters of the repair tool.

[0055] By adopting the above technical solution, the method of inputting the damage type and location information corresponding to the low-resolution damaged pipe image used for training the image recognition network, and inputting the test repair model output by the image recognition network into the mechanical operation model, can dynamically adjust the robot's operation path based on the specific damage situation (including damage type and location) and the application parameters of the repair tool. Through precise positioning of the damaged pipe location, it is beneficial to improve the operation efficiency of the maintenance robot.

[0056] In a preferred example, the first predicted moving speed v m With the second predicted operation speed ωj The calculation formula is:

[0057] v m =200 / D eq +50×(1-α / 1+α) (1)

[0058]

[0059] Among them, D eq The equivalent diameter of the pipe is α; the obstacle density index is α = N / S. r Where N is the number of obstacles, S r For a specified unit area based on the travel path; T j Tool torque requirements for repair tools (unit: N·m); D j S is the diameter of the pipe at the work point; j This refers to the contact area.

[0060] By adopting the above technical solution, factors such as the equivalent diameter of the pipe, the obstacle density index, the tool torque requirements of the maintenance tools, the pipe diameter at the work point, and the contact area are considered. This allows for the accurate calculation of the maintenance robot's speed parameters at different stages based on actual environmental conditions and task requirements. This enables better prediction of the entire maintenance task's time requirement and facilitates the rational planning of work schedules.

[0061] In summary, this application includes at least one of the following beneficial technical effects:

[0062] 1. A method integrating pipeline inspection information and maintenance task requirements is adopted. Based on this information, the travel path and operation mode of the maintenance robot are planned, and its first predicted movement speed and second predicted operation speed in the work area are predicted. This allows for the precise setting of the optimal working parameters for the robot, effectively solving the problem of low maintenance efficiency caused by the lack of detailed planning in existing technologies.

[0063] 2. Determine the forward repair relationship and reverse verification relationship between multiple low-resolution damaged pipe images and the first intact pipe model, and based on this, construct a maintenance robot operation control model that includes a repair accuracy loss function and an operation accuracy loss function to accurately simulate and optimize the maintenance process. Attached Figure Description

[0064] Figure 1 This is a schematic diagram illustrating an application scenario of a robot for repairing pipe leaks, according to one embodiment of this application.

[0065] Figure 2 This is a schematic cross-sectional view of two repair schemes for a robot used for repairing pipe leaks in one embodiment of this application when a pipe leak occurs;

[0066] Figure 3 This is a schematic cross-sectional view of two repair schemes for a robot used for repairing pipe leaks in one embodiment of this application when a pipe crack failure occurs;

[0067] Figure 4 This is a schematic cross-sectional view of a glue injection and sealing scheme for a robot used in pipe leak repair when a pipe is partially detached, according to one embodiment of this application.

[0068] Figure 5 This is a flowchart of a control method for a robot applied to pipe leak repair in one embodiment of this application;

[0069] Figure 6 This is another flowchart of a control method for a robot applied to pipe leak repair in one embodiment of this application. Detailed Implementation

[0070] The present application will be further described in detail below with reference to the accompanying drawings.

[0071] In one embodiment, such as Figure 1 As shown, this application discloses a robot for repairing pipe leaks. The robot for repairing pipe leaks includes a repair robot and a robot control mechanism. The robot control mechanism acquires pipe inspection information and repair task requirements, and plans the travel path and operation mode of the repair robot based on the pipe inspection information and repair task requirements. Based on the travel path and operation mode, it predicts a first predicted movement speed and a second predicted operation speed of the repair robot in the repair area. Based on the repair task requirements, the first predicted movement speed, and the second predicted operation speed, it generates a control parameter table for adjusting the operation efficiency of the repair robot. It acquires a pipe environment reference range representing the threshold of pipe environment changes, and optimizes a preset robot operation control model based on the pipe environment reference range and the control parameter table. Based on the robot operation control model, it controls the repair robot to perform pipe repair operations. It acquires real-time pipe environment data and inputs the real-time pipe environment data into the repair robot operation control model to adjust the operation state of the repair robot.

[0072] In this embodiment, the maintenance robot includes a mobile mechanism (such as a tracked or wheeled chassis), a work tool module (such as a sealant sprayer, welding device, camera, etc.), a sensor module (such as a lidar, infrared sensor, pressure sensor), and a communication module; the robot control mechanism consists of a central processing unit (CPU / GPU), a storage unit, a communication interface, and control algorithm software, which is responsible for receiving data, executing path planning, and generating control commands.

[0073] Specifically, the robot control mechanism acquires pipeline inspection information through a multimodal sensor group (including a laser scanner, infrared thermal imaging, and industrial cameras). This information includes pipeline material, diameter, curvature, and the location of obstacles within the pipeline. Maintenance task requirements include maintenance type, maintenance location, and required tools. Based on the pipeline diameter, curvature, and obstacle location, the robot control mechanism plans the robot's path. Based on the maintenance type, location, and required tools, the robot's operating mode is determined. Maintenance tools include a sealant gun, welding equipment, and grinding tools. The path optimization strategy includes prioritizing the expansion area, applying safety constraints, and smoothing the path. Prioritizing the expansion area involves increasing the radius by 30% around the maintenance point. Safety constraints include a minimum distance of ≥3mm from the pipeline wall (based on real-time feedback from the laser radar and using a Sigmoid interpolation algorithm for path smoothing, with an improved A* algorithm for path planning). The operating mode selection uses a decision tree model based on the maintenance type, location, and required tools.

[0074] For example, operating modes for different maintenance types and operating patterns, such as Figure 2-4 As shown, repair types include pipe dripping and pipe bursting. For pipe dripping, the work methods include sealing with adhesive and sealing with adhesive tape. Pipe bursting is further divided into pipe cracks, partial pipe detachment, and complete pipe detachment. For pipe cracks, the work methods include sealing with adhesive tape, injecting adhesive into the joints, replacing the pipe with adhesive, and replacing the pipe with adhesive-injected connections. For partial pipe detachment, the work methods include injecting adhesive into the joints, replacing the pipe with adhesive, and replacing the pipe with adhesive-injected connections. For complete pipe detachment, the work methods include replacing the pipe with adhesive and replacing the pipe with adhesive-injected connections. Figure 2 This is a schematic cross-sectional diagram illustrating two repair solutions: glue injection and adhesive tape wrapping, when a pipe leak occurs. Figure 3 This is a schematic cross-sectional diagram of two repair solutions for pipe crack sealing: glue injection and adhesive tape wrapping.

[0075] Figure 4 This is a schematic cross-sectional view of the glue injection and sealing scheme when the pipeline is partially detached. Other operation types and maintenance schemes will not be described in detail in this application. Specific maintenance schemes can be added or removed according to actual needs.

[0076] Specifically, based on the travel path and operation mode, a first predicted movement speed and a second predicted operation speed of the maintenance robot within the pipeline are predicted. This includes predicting the first predicted movement speed of the maintenance robot within the pipeline based on the pipeline diameter, pipeline curvature, and travel path; and predicting the second predicted operation speed of the maintenance robot at the work point based on the maintenance type, required tools, and operation mode. In this embodiment, the first predicted movement speed v... 移动 =D管道 / C 弯曲 ×K 路径 × Reference speed; The reference speed is set according to the robot's maximum design speed (e.g., the default value is 0.5m / s); D 管道 C is the pipe diameter. 弯曲 For pipe curvature, K 路径 For the travel path, if the pipe curvature exceeds the threshold (e.g., radius of curvature < 5 meters) or C 弯曲 =1.5, then the speed will be further reduced; the second predicted operating speed v 作业 =T 工具 / d 作业点 ×A 接触 ×Benchmark operating speed, T 工具 The required tool torque, i.e., the maximum torque of the tool (e.g., a welding gun torque of 15 N·m), d 作业点 Let A be the pipe diameter. 接触 The contact area between the tool and the pipe (e.g., the contact area of ​​the welding nozzle is 5 cm²). 2 The baseline operating speed is the tool's default speed (e.g., the default speed of a welding gun is 0.2 m / s).

[0077] Specifically, based on the maintenance task requirements, the first predicted moving speed, and the second predicted operating speed, a control parameter table is generated to adjust the operating efficiency of the maintenance robot. This includes calculating the total maintenance task time based on the maintenance location, the first predicted moving speed, and the second predicted operating speed. In this embodiment, the formula for calculating the total time is T. 总 =L 路径 / v 移动 +N 作业点 / v 作业 L 路径 The total path length required for the current maintenance task is given; based on the total time and maintenance task requirements, the target operating efficiency of the maintenance robot is obtained, where the target operating efficiency refers to the completion time threshold set according to the maintenance task requirements (e.g., an emergency task needs to be completed within 5 minutes), and the target operating efficiency E is... 目标 The calculation formula is: E 目标 =T 总 / T 需求 ×100%, where T 需求 The desired completion time is set in advance; based on the target work efficiency, the first predicted moving speed, and the second predicted work speed, the moving speed and work speed parameters of the maintenance robot are adjusted to generate a control parameter table for adjusting the work efficiency of the maintenance robot.

[0078] The various modules in the aforementioned robot for repairing pipe leaks can be implemented entirely or partially through software, hardware, or a combination thereof; the modules can be embedded in the processor of a computer device in hardware form or independent of the processor, or they can be stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to the modules.

[0079] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed. That is, the internal structure of the robot for pipe leak repair can be divided into different functional units or modules to complete all or part of the functions described above.

[0080] In one embodiment, such as Figure 5 As shown, a control method for a robot used in pipe leak repair is provided. This control method, applied to a robot used in pipe leak repair, and the data processing method applied to a multi-functional integrated service terminal specifically include the following steps:

[0081] S1: Obtain pipeline inspection information and maintenance task requirements.

[0082] S2: Based on pipeline inspection information and maintenance task requirements, plan the travel path and operation mode of the maintenance robot.

[0083] S3: Based on the planned travel path and operation mode, predict the first predicted movement speed and the second predicted operation speed of the maintenance robot in the operation and maintenance area.

[0084] In this embodiment, the first predicted moving speed refers to the theoretical speed at which the robot travels along the planned path, taking into account the dynamic value of the pipe geometry and mechanical transmission characteristics; the second predicted working speed refers to the rotational speed at which the robot performs maintenance actions at the work point, which is related to the tool characteristics and contact state; the travel path refers to the optimal moving route of the robot from its current position to the maintenance point.

[0085] Furthermore, in this embodiment, the first predicted moving speed v m With the second predicted operating speed ω j The calculation formula is:

[0086] v m =200 / D eq +50×(1-α / 1+α) (1)

[0087]

[0088] Among them, D eqThe equivalent diameter of the pipe is α; the obstacle density index is α = N / S. r Where N is the number of obstacles, S r For a specified unit area based on the travel path; T j Tool torque requirements for repair tools (unit: N·m); D j S is the diameter of the pipe at the work point; j This refers to the contact area.

[0089] S4: Based on the maintenance task requirements, the first predicted moving speed, and the second predicted working speed, generate a control parameter table for adjusting the working efficiency of the maintenance robot.

[0090] In this embodiment, the control parameter table refers to an optimized configuration table that includes adjustments to robot speed and tool parameters.

[0091] S5: Obtain the pipeline environment reference range representing the threshold of pipeline environment change, and optimize the preset robot operation control model based on the pipeline environment reference range and the control parameter table.

[0092] Specifically, the pipeline environment reference range refers to the safety threshold of pipeline environment parameters (such as temperature ≤50℃, pressure ≤10MPa); the robot operation control model refers to the control algorithm based on historical data and real-time input. The robot operation control model in this application refers to the intelligent control model of LSTM neural network.

[0093] S6: Control the maintenance robot to perform pipeline maintenance operations based on the optimized robot operation control model.

[0094] S7: During the maintenance operation, acquire real-time pipeline environment data; input the real-time pipeline environment data into the robot operation control model to dynamically adjust the operation status of the maintenance robot.

[0095] In this embodiment, real-time pipeline environmental data refers to environmental parameters (such as temperature, pressure, and obstacle changes) continuously collected during the operation. For example, if an obstacle is detected moving in front of the path, the robot operation control model replans the path, reduces the moving speed, or adjusts the tool parameters, such as pausing welding to avoid the obstacle.

[0096] In one embodiment, such as Figure 6 As shown, the pipeline inspection information includes low-resolution images of damaged pipelines; after step S1, a control method for a robot applied to pipeline leak repair further includes:

[0097] S11: Determine the forward repair relationship between multiple low-resolution damaged pipe images and the first intact pipe model, and determine the reverse verification relationship between the first intact pipe model and multiple repaired pipe images.

[0098] In this embodiment, the first low-resolution damaged pipe image refers to a low-resolution or poor-quality pipe defect image (such as a blurry image caused by insufficient lighting or sensor limitations); the first intact pipe model refers to an ideal pipe structure model generated by an algorithm, such as a pipe geometric model without cracks or leaks.

[0099] Specifically, the forward repair relationship refers to the mapping function from a low-resolution damaged image to a pipe model in its intact state. The forward repair relationship trains an image recognition network using labeled data (such as the correspondence between known damaged areas and the ideal model), enabling it to learn the mapping from low-quality images to the ideal model. The reverse verification relationship refers to the physical simulation process from the first intact pipe model to multiple repaired pipe images, used to verify the accuracy of the repair results. The reverse verification relationship refers to constructing a reverse model based on physical rules (such as materials mechanics and fluid dynamics) to ensure that the generated repaired images conform to actual physical conditions.

[0100] S12: Based on the image recognition network corresponding to the positive repair relationship and the mechanical operation model corresponding to the negative verification relationship, a maintenance robot operation control model is constructed, wherein the image recognition network includes a repair accuracy loss function and the machine operation model includes an operation accuracy loss function.

[0101] In this embodiment, the image recognition network is a deep learning model, such as U-Net; the image recognition network is connected in series with the mechanical operation model to form an end-to-end control model, namely the maintenance robot operation control model.

[0102] Specifically, the formula for calculating the accuracy loss function is as follows:

[0103]

[0104] Where N corresponds to the number of image samples with positive inpainting relationships, ||I pred -I gt ||2 is the pixel-level L2 norm, used to predict image I. pred With real image I gt (Here referring to low-resolution damaged images) pixel-level Euclidean distance; α is the loss weight coefficient, a balance factor between pixel loss and perceptual loss, with values ​​between [0.3, 0.7], and Per is the feature matching loss based on a pre-trained convolutional network (such as VGG19). Specifically, I is extracted. pred with I gt The VGG feature maps of the two images are used to calculate the L2 norm difference for each layer of feature maps, and then a weighted sum is obtained: i.e. L refers to the feature map layer, α l γ1 and γ2 are the weight coefficients of the current layer, and are predefined weight coefficient values.

[0105] The formula for calculating the operational accuracy loss function is:

[0106]

[0107] Where M is the number of action sequences in the mechanical operation dataset that satisfies the reverse verification relation; ||F pred -F real ||1 is used to satisfy the L1 norm of the mechanical response and to define the predicted torque F. pred With actual torque F real The sum of absolute errors, torque considers the three-dimensional force vector along the X, Y, and Z axes, F pred =[F x F y F z ] T β is the loss weighting coefficient, a balance factor between mechanical loss and smoothing penalty, with a value between [0.5, 0.9]. For is the force smoothing penalty, used to define the smoothness index of torque changes between adjacent actions. The formula for calculating For is... Where F avg is the average torque value, and T is the length of the action sequence.

[0108] S13: Obtain the training dataset with the established positive repair relationship and reverse verification relationship; based on the training dataset, train the operation control model of the maintenance robot, wherein the repair accuracy loss function and the operation accuracy loss function are jointly optimized during training.

[0109] In this embodiment, the Adam optimizer is used with a learning rate of 0.001 to jointly optimize the network parameters of the repair precision loss function and the operation accuracy loss function.

[0110] Specifically, step S13 also includes:

[0111] S131: Based on the training dataset, determine the training high-resolution repaired pipe image, the training low-resolution damaged pipe image, and the training intact pipe model; wherein, the training low-resolution damaged pipe image and the training intact pipe model satisfy a positive repair relationship, and the training intact pipe model and the training high-resolution repaired pipe image satisfy a negative verification relationship.

[0112] S132: Input multiple training low-resolution damaged pipe images into the maintenance robot operation control model. The image recognition network outputs a test recovery model based on the multiple training low-resolution damaged pipe images. The mechanical operation model outputs multiple test and repaired pipe images based on the test repair model.

[0113] Specifically, the damage type and location information corresponding to the low-resolution damaged pipe image trained by the input image recognition network, as well as the test repair model output by the image recognition network, are input into the mechanical operation model to obtain multiple pipe images after test repair. The mechanical operation model adjusts the robot operation path according to the damage type, location information, and application parameters of the repair tool.

[0114] In this embodiment, the mechanical operation model includes mechanical simulation based on finite element analysis (FEA) or a dynamic model, combined with robot kinematics (such as joint angles and movement speed) and tool mechanical parameters (such as heat conduction of the welding gun and adhesion of the sealant) to simulate the repair process. That is, based on the geometric data of the test repair model, the interaction between the tool and the pipe (such as the heat-affected zone during welding) is simulated, and the surface morphology of the repaired pipe (such as smoothness and stress distribution) is calculated. The post-repair image refers to the image after the simulation results are converted into a high-cleanliness image. The surface of the repaired pipe image is free of cracks and meets the relevant repair process standards.

[0115] The process involves inputting a test repair model into a mechanical operation model to obtain multiple test-repaired pipeline images. This includes: inputting the damage type and location information corresponding to the low-resolution damaged pipeline images trained by the image recognition network, and inputting the test repair model output by the image recognition network into the mechanical operation model to obtain multiple test-repaired pipeline images. The mechanical operation model adjusts the robot's operation path based on the damage type, location information, and application parameters of the repair tool. For example, the damage type (e.g., crack width, corrosion depth) and location coordinates are extracted from the low-resolution training images, and the repair results (e.g., three-dimensional coordinates after crack closure) are output by the image recognition network. The damage type (e.g., cracks need to be precisely aligned), location (e.g., coordinates X, Y, Z), and tool parameters (e.g., radius of the welding gun) are input into the path planning algorithm to obtain the motion trajectory of the robot's end effector (e.g., avoiding obstacles, aligning with the repair area). Through an improved A* algorithm, combined with the contact angle between the tool and the pipeline (e.g., the welding gun needs to be perpendicular to the crack) and the maximum load of the tool (e.g., torque not exceeding 15 N·m), the path is adjusted based on mechanical problems found in the simulation (e.g., stress concentration areas) (e.g., increasing the number of welding layers).

[0116] For example, the correspondence between damage types and tool parameters includes:

[0117] Crack → Welding gun (3000W power, 0.2m / s speed).

[0118] Corrosion → Sealant spraying (pressure 5MPa, coverage 100%).

[0119] S133: Determine the repair accuracy loss function based on the trained intact state pipeline model and the test repair model.

[0120] In this embodiment, the accuracy loss function L is repaired. im =MSE(tested and repaired model, trained and perfected model) + λ im ×Structural similarity loss), MSE refers to mean squared error, which measures the difference in geometric structure. Structural similarity loss is used to evaluate the shape matching degree (such as curvature consistency) between the predicted model and the real model. For example: if the difference in curvature between the predicted model and the real model is 0.1, and the MSE is 0.01, then L im =0.01+0.2×s0.1=0.03.

[0121] S134: Determine the operational accuracy loss function based on the training high-resolution repaired pipe image and the test repaired pipe image.

[0122] In this embodiment, the operation accuracy loss function L me =SSIM(test restored image, training high-resolution image) + λ me ×Physical Constraint Loss), SSIM refers to the Structural Similarity Index, which evaluates the matching degree between image texture and structure (in this embodiment, SSIM ≥ 0.9 is set as acceptable). Physical constraint loss is used to ensure that the repair result conforms to physical rules (such as the welding temperature not exceeding the melting point of the material). For example, if SSIM is 0.85 and physical constraint loss is 0.05, then L me =1-0.85+0.05=0.2.

[0123] S135: Determine the total maintenance loss function based on the repair accuracy loss function and the operation accuracy loss function.

[0124] Specifically, the total maintenance loss function refers to the weighted loss combining repair accuracy and operational accuracy, used for joint optimization of model parameters; the formula for calculating the total maintenance loss function is: L me =μ1×L im +μ2×L me The weights are set according to the task requirements. For example, μ1 = 0.7 and μ2 = 0.3, and both μ1 and μ2 conform to the dynamic adjustment mechanism.

[0125] S136: Optimize the total maintenance loss function to train the maintenance robot operation control model until the maintenance robot operation control model is trained.

[0126] S14: Input the multiple low-resolution images of the damaged pipe sections to be repaired into the trained maintenance robot operation control model to obtain high-resolution images of the repaired pipes.

[0127] In this embodiment, the second low-resolution damaged pipe image refers to a low-quality image of the actual pipe to be repaired; the high-resolution pipe image refers to the repaired image generated by the model, which is used to guide the robot to perform repair operations.

[0128] Specifically, an image recognition network analyzes low-resolution images to generate an ideal pipe model (such as the three-dimensional structure after crack closure). A mechanical operation model then generates a high-resolution image of the repaired pipe (such as the crack-free pipe surface) based on this ideal model. This high-resolution image is transmitted to the robot control system to guide the tool in performing repair operations (such as locating the crack and controlling the sealant spraying trajectory). The mechanical operation model in this application simulates the welding repair process using finite element analysis (FEA) to generate the repaired surface image.

[0129] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0130] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0131] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A robot for repairing leaking pipes, characterized in that, Including maintenance robots and robot control mechanisms, The robot control mechanism acquires pipeline inspection information and maintenance task requirements, and plans the robot's travel path and operation mode based on the pipeline inspection information and maintenance task requirements; the maintenance task requirements include maintenance type, maintenance location, and required tools; Based on the travel path and operation mode, predict the first predicted moving speed and the second predicted operation speed of the maintenance robot in the operation and maintenance area; Based on the maintenance task requirements, the first predicted moving speed and the second predicted working speed, a control parameter table is generated to adjust the working efficiency of the maintenance robot. Obtain a pipeline environment reference range representing the threshold of pipeline environment change, and optimize the preset robot operation control model based on the pipeline environment reference range and the control parameter table; The robot operation control model is used to control the maintenance robot to perform pipeline maintenance operations; real-time pipeline environment data is acquired and input into the robot operation control model to adjust the operation status of the maintenance robot. The step of generating a control parameter table for adjusting the operating efficiency of the maintenance robot based on the maintenance task requirements, the first predicted moving speed, and the second predicted operating speed specifically includes: The total time required for the maintenance task is calculated based on the maintenance location, the first predicted movement speed, and the second predicted operation speed. Based on the total time consumed and the maintenance task requirements, the target operating efficiency of the maintenance robot is obtained; Based on the target work efficiency, the first predicted moving speed, and the second predicted work speed, the moving speed and work speed parameters of the maintenance robot are adjusted to generate a control parameter table for adjusting the work efficiency of the maintenance robot.

2. The robot for repairing leaking pipes according to claim 1, characterized in that, The process of acquiring pipeline inspection information and maintenance task requirements, and planning the travel path and operation mode of the maintenance robot based on the pipeline inspection information and maintenance task requirements, specifically includes: Obtain pipeline inspection information, which includes pipeline material, pipeline diameter, pipeline curvature, and the location of obstacles inside the pipeline; The travel path of the maintenance robot is planned based on the pipe diameter, the pipe curvature, and the location of obstacles inside the pipe. The operating mode of the maintenance robot is determined based on the maintenance type, the maintenance location, and the required tools.

3. The robot for repairing leaking pipes according to claim 2, characterized in that, The step of predicting the first predicted movement speed and the second predicted operation speed of the maintenance robot within the pipeline based on the travel path and operation mode specifically includes: Based on the pipe diameter, the pipe curvature, and the travel path, predict the first predicted moving speed of the maintenance robot inside the pipe; Based on the repair type, the required tools, and the operation mode, a second predicted operation speed of the repair robot at the work point is predicted.

4. A control method for a robot used in pipe leak repair, characterized in that, The control method, applied to a robot for pipe leak repair as described in any one of claims 1-3, comprises: Obtain pipeline inspection information and maintenance task requirements; Based on the pipeline inspection information and maintenance task requirements, plan the travel path and operation mode of the maintenance robot; Based on the planned travel path and operation mode, the first predicted movement speed and the second predicted operation speed of the maintenance robot in the operation and maintenance area are predicted. Based on the maintenance task requirements, the first predicted moving speed and the second predicted working speed, a control parameter table is generated to adjust the working efficiency of the maintenance robot. Obtain a pipeline environment reference range representing the threshold of pipeline environment change, and optimize the preset robot operation control model based on the pipeline environment reference range and the control parameter table; The optimized robot operation control model is used to control the maintenance robot to perform pipeline maintenance operations. During maintenance operations, real-time pipeline environment data is acquired; the real-time pipeline environment data is input into the robot operation control model to dynamically adjust the operation status of the maintenance robot.

5. The control method for a robot applied to pipeline leak repair according to claim 4, characterized in that, The pipeline inspection information includes low-resolution images of damaged pipelines. After acquiring the pipeline inspection information, the method further includes: Determine the forward repair relationship between multiple low-resolution damaged pipe images and the first intact pipe model, and determine the reverse verification relationship between the first intact pipe model and multiple repaired pipe images. Based on the image recognition network corresponding to the positive repair relationship and the mechanical operation model corresponding to the reverse verification relationship, a robot operation control model is constructed, wherein the image recognition network includes a repair accuracy loss function and the mechanical operation model includes an operation accuracy loss function. Obtain a training dataset in which the forward repair relationship and the reverse verification relationship have been determined; based on the training dataset, train the robot operation control model, wherein the repair accuracy loss function and the operation accuracy loss function are jointly optimized during training; The multiple low-resolution images of the damaged pipe sections to be repaired are input into the trained robot operation control model to obtain high-resolution images of the repaired pipes.

6. The control method for a robot applied to pipeline leak repair according to claim 5, characterized in that, The step of training the robot operation control model based on the training dataset includes: Based on the training dataset, a training high-resolution repaired pipe image, a training low-resolution damaged pipe image, and a training intact pipe model are determined; wherein, the training low-resolution damaged pipe image and the training intact pipe model satisfy a positive repair relationship, and the training intact pipe model and the training high-resolution repaired pipe image satisfy a reverse verification relationship. Multiple training low-resolution damaged pipe images are input into the robot operation control model. The image recognition network outputs a test repair model based on the multiple training low-resolution damaged pipe images. The mechanical operation model outputs multiple test repair pipe images based on the test repair model. Based on the trained intact pipeline model and the test repair model, the repair accuracy loss function is determined; Based on the training high-resolution repaired pipeline image and the test repaired pipeline image, the operation accuracy loss function is determined; Based on the repair accuracy loss function and the operation accuracy loss function, the total maintenance loss function is determined; The total maintenance loss function is optimized to train the robot operation control model until the robot operation control model is fully trained.

7. The control method for a robot applied to pipeline leak repair according to claim 6, characterized in that, The mechanical operation model outputs multiple segments of the tested and repaired pipeline images based on the test and repair model, including: The test repair model is input into the mechanical operation model to obtain multiple segments of the pipeline after the test repair. The mechanical operation model is a mechanical model of the simulated repair process constructed based on the robot operation path and the application parameters of the repair tool.

8. The control method for a robot applied to pipeline leak repair according to claim 7, characterized in that, The step of inputting the test repair model into the mechanical operation model to obtain multiple segments of the tested and repaired pipeline images includes: The damage type and location information corresponding to the low-resolution damaged pipe image input to the image recognition network, and the test repair model output by the image recognition network are input to the mechanical operation model to obtain multiple segments of the tested and repaired pipe images. The mechanical operation model adjusts the robot operation path according to the damage type, location information and application parameters of the repair tool.

9. The control method for a robot applied to pipeline leak repair according to claim 5, characterized in that, First predicted movement speed With the second predicted operation speed The calculation formula is: in, The equivalent diameter of the pipe; The obstacle density index, Where N is the number of obstacles. For a specified unit area based on the travel path; The torque requirement for maintenance tools, in N·m; The diameter of the pipe at the work point; This refers to the contact area.

Images