Pipeline crack repairing method and system, electronic equipment and computer program product

Abstract

The invention provides a pipeline crack repairing method and system, electronic equipment and a computer program product, and relates to the technical field of computers. The pipeline crack repairing method comprises the steps that pipeline cracks in a pipeline digital twinning model are recognized, the positions and the crack types of the pipeline cracks are determined, and the pipeline digital twinning model is determined based on a pipeline of a target area; determining the danger level of the pipeline crack according to the position and the crack type of the pipeline crack; based on the danger level and crack features extracted for the pipeline cracks, determining a repair priority for sorting the pipeline cracks; and generating a control instruction based on the repair priority, the control instruction being used for controlling the unmanned aerial vehicle to repair the pipeline crack of the target area.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and in particular to a method, system, electronic device, and computer program product for repairing pipe cracks. Background Technology

[0002] In the operation and maintenance of urban underground pipe networks, industrial pipeline systems, and long-distance pipelines, pipelines are prone to cracks due to corrosion, fatigue, stress concentration, or external damage. If not repaired promptly, these cracks can lead to major safety accidents such as leaks, explosions, or environmental pollution. Traditional pipeline inspection and repair mainly rely on manual inspection and excavation, which suffers from high operational risks, low efficiency, high costs, and long downtime, especially in complex, narrow, or high-risk pipeline environments. While pipeline robots have been applied in recent years, most only possess visual inspection capabilities and lack the ability to intelligently identify cracks, assess risks, and autonomously repair them. Summary of the Invention

[0003] This disclosure provides a method, system, electronic device, and computer program product for repairing pipe cracks.

[0004] According to one aspect of this disclosure, a method for repairing pipeline cracks is provided, comprising: identifying pipeline cracks in a pipeline digital twin model, determining the location and type of the pipeline cracks, wherein the pipeline digital twin model is determined based on pipelines in a target area; determining the hazard level of the pipeline cracks based on the location and type of the pipeline cracks; determining a repair priority for prioritizing the pipeline cracks based on the hazard level and crack features extracted for the pipeline cracks; and generating control commands based on the repair priorities, the control commands being used to control a drone to repair the pipeline cracks in the target area.

[0005] Based on a pipeline crack repair method, this approach achieves precise perception of pipeline defects by automatically identifying the location and type of cracks in a digital twin model of the pipeline built around the target area. The hazard level is assessed by combining location information and crack type, and a comprehensive analysis of crack characteristics is performed to scientifically determine repair priorities, ensuring that high-risk defects are addressed first. Control commands are generated based on these priorities, driving unmanned aerial vehicles (UAVs) to autonomously execute repair tasks. This achieves closed-loop management of the entire process, from defect identification, risk assessment, decision prioritization to unmanned repair, significantly improving the intelligence, precision, and efficiency of pipeline operation and maintenance, effectively preventing major leakage accidents, and ensuring the safe and stable operation of the pipeline network system.

[0006] According to at least one embodiment of the present disclosure, a method for repairing pipe cracks, wherein the pipe digital twin model is determined based on the pipe in a target area, includes: determining the pipe digital twin model based on the pipe geometric model of the target area and defect parameters of the pipe cracks in a pipe crack dataset, wherein the pipe geometric model is determined based on the three-dimensional point cloud of the target area, and the defect parameters are determined based on the three-dimensional model of the pipe cracks.

[0007] According to at least one embodiment of the pipe crack repair method of this disclosure, a pipe digital twin model is determined based on the pipe geometric model of the target area and the defect parameters of the pipe crack in the pipe crack dataset. The method includes: fusing the defect parameters of the pipe crack in the pipe crack dataset of the target area with the pipe geometric model to determine the pipe defect model; and adding labels containing the pipe's material properties, stress distribution and / or defect type to the pipe defect model to determine the pipe digital twin model.

[0008] According to at least one embodiment of the present disclosure, a method for repairing pipe cracks, wherein the pipe geometric model is determined based on a three-dimensional point cloud of the target region, includes: performing semantic segmentation on the three-dimensional point cloud of the target region to determine a first three-dimensional point cloud containing pipe structural features; fusing the first three-dimensional point cloud and pipe temperature data to determine a second three-dimensional point cloud containing pipe temperature data; and determining a pipe geometric model based on the second three-dimensional point cloud.

[0009] According to at least one embodiment of the present disclosure, the method for repairing pipe cracks includes a defect parameter determined based on a three-dimensional model of the pipe crack. The three-dimensional model of the pipe crack is determined by the following steps: determining a defect region containing the pipe crack based on a three-dimensional point cloud of a target region; scanning and detecting the defect region to determine the surface geometric data and internal defect features of the corresponding pipe crack; and determining a three-dimensional model of the pipe crack based on the surface geometric data and internal defect features.

[0010] According to at least one embodiment of the present disclosure, a method for repairing pipe cracks, based on a three-dimensional point cloud of a target area, determines a defect area containing the pipe crack, including: determining the average temperature of a target point within a corresponding target range for the three-dimensional point cloud of the target area, wherein the target point is a point located in the pipe structure in the three-dimensional point cloud; determining the temperature difference of the target point based on the difference between the temperature of the target point and the average temperature; and determining the defect area containing the pipe crack based on the temperature difference of the target point.

[0011] A method for repairing pipe cracks according to at least one embodiment of the present disclosure, comprising: extracting features from the pipe cracks to determine crack features; determining a risk score for the pipe cracks based on crack features, the risk level, and / or constraints; and determining a repair priority for ranking the pipe cracks based on the risk score.

[0012] According to at least one embodiment of the pipe crack repair method of this disclosure, based on the risk score, a repair priority for ranking the pipe cracks is determined, including: when the risk score is greater than or equal to a first threshold, determining the corresponding pipe crack as a first repair priority; when the risk score is less than the first threshold but greater than or equal to a second threshold, determining the corresponding pipe crack as a second repair priority, wherein the first threshold is greater than the second threshold, and the repair order of the second repair priority is after the repair order of the first repair priority; when the risk score is less than the second threshold, determining the corresponding pipe crack as a third repair priority, wherein the repair order of the third repair priority is after the repair order of the second repair priority.

[0013] According to at least one embodiment of the present disclosure, the method for repairing pipe cracks further includes, after repairing the pipe cracks,: inspecting the repaired pipe cracks to determine the surface smoothness of the repaired pipe cracks, the curing state of the repair material, and / or the bonding quality between the repair material and the pipe wall; and generating a pipe repair report based on the surface smoothness, the curing state of the repair material, and / or the bonding quality.

[0014] According to at least one embodiment of the pipe crack repair method of the present disclosure, when determining the hazard level of the pipe crack, the method further includes: assessing the pipe crack based on the location and type of the crack, and determining a repair method for the pipe crack, wherein the repair method is used to generate the control command.

[0015] According to at least one embodiment of the pipeline crack repair method of this disclosure, before identifying the pipeline crack in the pipeline digital twin model, the method includes: acquiring the starting pose of the pipeline when a drone enters the target area, the starting pose being used to determine the spatial coordinates of the drone relative to a global reference coordinate system; determining a three-dimensional point cloud map of the pipeline based on the starting pose and data from a pipeline geographic information system; determining the detection path of the drone based on the three-dimensional point cloud map; and determining environmental perception data of the pipeline based on the detection path for scanning and detecting the pipeline, the environmental perception data being used to construct a pipeline digital twin model.

[0016] The pipe crack repair method according to at least one embodiment of the present disclosure further includes: obtaining the endpoint pose of the UAV at the current moment; determining the return path of the UAV based on the three-dimensional point cloud map, the battery level of the UAV, the starting pose and the endpoint pose; and controlling the UAV to return to the starting point based on the return path.

[0017] The method for repairing pipe cracks according to at least one embodiment of the present disclosure further includes: determining the fault type of the UAV based on its operating status; releasing the positioning buoy of the UAV when the fault type is power failure; shutting off the corresponding valve of the UAV when the fault type is material leakage; and storing the data collected by the UAV into a local cache when the fault type is communication interruption.

[0018] According to another aspect of this disclosure, a system for repairing pipe cracks based on unmanned aerial vehicles (UAVs) is provided, for performing the pipe crack repair method as described in any of the preceding claims, comprising: a perception module for acquiring multimodal perception data of the pipe in a target area; a data processing module for establishing a digital twin model of the pipe using the multimodal perception data, and generating control commands based on the digital twin model of the pipe; a repair module for repairing the pipe crack according to the control commands; an energy management module for providing energy to the UAV; and an environmental adaptation module for providing the UAV with impact resistance, waterproofing, and / or resistance to water flow disturbance.

[0019] According to at least one embodiment of the present disclosure, a system for repairing pipe cracks based on unmanned aerial vehicles (UAVs) includes a sensing module comprising: a multispectral imaging module, including a lidar, a structured light camera, an infrared thermal imager, and a near-infrared camera, for scanning and detecting pipe cracks; and a three-dimensional scanning module, including an ultrasonic probe, for detecting the internal structure of the pipe cracks.

[0020] According to at least one embodiment of the present disclosure, a system for repairing pipe cracks based on unmanned aerial vehicles (UAVs) includes a repair module comprising: a robotic arm consisting of multiple rotary joints connected in series, the rotary joints being driven by brushless servo motors; and an adhesive injection system including a two-component metering pump and an adhesive injection head.

[0021] According to at least one embodiment of the present disclosure, a system for repairing pipe cracks based on a drone includes an energy management module comprising: a main power source consisting of a lithium polymer battery pack; and a wireless charging interface for charging the drone.

[0022] According to at least one embodiment of the present disclosure, a system for repairing pipe cracks based on unmanned aerial vehicles (UAVs) includes an environmental adaptation module comprising: a protective device including an IP68-rated housing and a wing with an impact-resistant design; and a stabilizing device including a retractable stabilizing wing.

[0023] According to another aspect of this disclosure, an electronic device is provided, comprising: a memory storing execution instructions; and a processor executing the execution instructions stored in the memory, causing the processor to perform a pipe crack repair method according to any embodiment of this disclosure.

[0024] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements a method for repairing pipe cracks according to any embodiment of this disclosure. Attached Figure Description

[0025] The accompanying drawings illustrate exemplary embodiments of the present disclosure and, together with the description thereof, serve to explain the principles of the present disclosure. These drawings are included to provide a further understanding of the present disclosure and are incorporated in and constitute a part of this specification.

[0026] Figure 1 This is a schematic block diagram of a system for repairing pipe cracks based on an application of a pipe crack repair method according to one embodiment of the present disclosure.

[0027] Figure 2 This is a schematic diagram of the overall process of a pipe crack repair method according to one embodiment of the present disclosure.

[0028] Figure 3 This is a schematic diagram of the process of acquiring environmental perception data based on a detection path in a pipeline crack repair method according to one embodiment of the present disclosure.

[0029] Figure 4 This is a schematic flowchart illustrating the process of generating a pipeline repair report in a pipeline crack repair method according to one embodiment of the present disclosure.

[0030] Figure 5 This is a flowchart illustrating the process of determining the return path of a drone in a pipe crack repair method according to one embodiment of the present disclosure.

[0031] Figure 6 This is a schematic flowchart of an abnormal fault handling process in a pipeline crack repair method according to one embodiment of the present disclosure.

[0032] Figure 7 This is a schematic flowchart illustrating the process of determining a digital twin model of a pipeline in a pipeline crack repair method according to one embodiment of the present disclosure.

[0033] Figure 8 This is a schematic flowchart illustrating the process of determining the geometric model of a pipeline in a pipeline crack repair method according to one embodiment of the present disclosure.

[0034] Figure 9This is a flowchart illustrating the process of determining a three-dimensional model of a pipe crack in a pipe crack repair method according to one embodiment of the present disclosure.

[0035] Figure 10 This is a schematic flowchart illustrating the process of determining a defective region containing a pipe crack in a pipe crack repair method according to one embodiment of the present disclosure.

[0036] Figure 11 This is a flowchart illustrating the process of determining repair priority in a pipe crack repair method according to one embodiment of the present disclosure.

[0037] Figure 12 This is a schematic block diagram of a pipe crack repair device according to one embodiment of the present disclosure.

[0038] Figure 13 This is a schematic structural block diagram of an electronic device according to one embodiment of the present disclosure. Detailed Implementation

[0039] The present disclosure will now be described in further detail with reference to the accompanying drawings and examples. It should be understood that the specific examples described herein are for illustrative purposes only and are not intended to limit the scope of the disclosure. Furthermore, it should be noted that, for ease of description, only the parts relevant to the present disclosure are shown in the accompanying drawings.

[0040] It should be noted that, where there is no conflict, the embodiments and features described in this disclosure can be combined with each other. The technical solutions of this disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0041] In complex application scenarios such as urban drainage networks and petrochemical pipelines, pipelines are often in environments with high humidity, water, narrow bends, or even partial submersion. Cracks frequently occur in stress concentration areas such as welds and elbows, and their forms are diverse (e.g., longitudinal microcracks, circumferential penetrating cracks), making it difficult for traditional manual or semi-automatic equipment to access and perform precise repairs. While existing robots can detect cracks, they lack the ability to comprehensively assess crack type and location, and cannot automatically generate matching repair methods based on scenario differences such as whether straight pipe sections are suitable for adhesive injection or weld areas require flexible sealing. Furthermore, in scenarios with multiple consecutive cracks, robots often operate blindly according to a preset sequence, without dynamically calculating repair priorities based on crack depth, propagation trends, and other characteristics, which can easily lead to delays in treating high-risk cracks.

[0042] To address this issue, this disclosure proposes a method for repairing pipeline cracks. The method involves accurately identifying the location and type of pipeline cracks using a digital twin model of the pipeline constructed based on the target area. The hazard level is assessed by combining the stress environment at the location with the crack type characteristics, and a comprehensive score is generated by integrating crack features such as depth and length to scientifically generate a repair priority ranking. Based on the repair priority, control commands are generated for unmanned aerial vehicles (UAVs), driving the UAV to accurately reach the target location carrying the repair payload and perform repair operations such as adhesive injection and spraying. This significantly improves the intelligence level of defect handling in complex pipeline systems, effectively solving problems such as reliance on manual interpretation, chaotic repair sequence, and delayed response in traditional operation and maintenance, and greatly improving repair efficiency, safety, and resource utilization.

[0043] To facilitate description and make the technical solutions of this disclosure easier to understand, the terminology of this disclosure will be explained before describing the technical solutions of this disclosure.

[0044] A digital twin model is a technology that accurately maps a physical entity to a digital model in a virtual space. It utilizes information such as physical models, sensor data, and operational history to integrate simulation processes involving multiple disciplines, multiple physical quantities, and multiple scales, comprehensively reflecting the entire lifecycle of the physical entity.

[0045] The pipeline crack repair method disclosed herein can be widely applied to various complex scenarios such as urban underground pipe networks, petrochemical pipelines, nuclear power plant cooling systems, cross-sea water pipelines, and long-distance oil and gas trunk lines. In urban drainage systems, drones can enter combined sewer systems, autonomously navigate in high water levels and strong current environments, identify and prioritize the repair of circumferential cracks located at manhole connections, preventing leakage and groundwater pollution. In petrochemical plants, thermal imaging fusion detection can be performed on weld cracks in high-temperature and high-pressure pipelines. After assessing the hazard level, corrosion-resistant materials can be automatically injected to avoid production shutdowns and excavation. In cross-sea or underwater tunnel pipelines, drones, with the aid of digital twin models and water-resistant stabilizing wings, can complete long-distance inspections and emergency sealing under communication-limited conditions, and automatically return to the entrance buoy for positioning when communication is lost. In high-radiation areas such as nuclear power plant cooling loops, drones can replace manual labor to enter dangerous sections for inspection and repair, with data cached locally and securely uploaded via wireless links. In the construction of smart cities, access to the city's CIM (City Information Modeling) platform can be used to achieve real-time visual monitoring and predictive maintenance of the health status of the city's pipeline network, support remote dispatching, emergency early warning and life cycle management by government regulatory departments, and form a new paradigm of intelligent and unmanned pipeline operation and maintenance.

[0046] Figure 1 A schematic block diagram of a system for repairing pipe cracks based on an embodiment of this disclosure is shown. Figure 1The system 100 shown, which repairs pipe cracks based on drones, includes: a perception module 110 for acquiring multimodal perception data of the pipes in the target area; a data processing module 120 for establishing a digital twin model of the pipes using the multimodal perception data and generating control commands based on the digital twin model; a repair module 130 for repairing pipe cracks according to the control commands; an energy management module for providing energy to the drone; and an environmental adaptation module 140 for providing the drone with functions such as impact resistance, waterproofing, and / or resistance to water flow disturbance.

[0047] The system acquires multimodal perception data of the pipeline through the perception module, constructs a digital twin model of the pipeline through the data processing module, and generates control commands to drive the repair module to perform crack repair operations. At the same time, the energy management module provides continuous energy support for the UAV, and the environmental adaptation module endows it with the ability to resist impact, water, and water flow disturbance. This enables the UAV to autonomously perceive, make decisions, and repair in complex pipeline environments based on the digital twin model, ensuring the reliability of the system under harsh working conditions and the accurate execution of repair operations.

[0048] Preferably, the perception module integrates lidar, structured light camera, infrared thermal imager and near-infrared camera through a multispectral imaging module to achieve multimodal surface scanning and identification of pipeline cracks. At the same time, the ultrasonic probe in the three-dimensional scanning module detects the internal structural features of the cracks, thereby comprehensively acquiring the apparent morphology and deep structural information of pipeline cracks in different spectra and imaging dimensions, providing complete and accurate perception data support for subsequent modeling and repair.

[0049] Preferably, the repair module achieves flexible movement through a robotic arm composed of multiple rotary joints connected in series. Each rotary joint is driven by a brushless servo motor to ensure high-precision and high-response attitude control. The glue injection system integrates a two-component metering pump and a glue injection head, which can accurately measure and mix the two component materials to achieve stable and controllable glue injection operations. Thus, under the precise positioning of the robotic arm, reliable filling and sealing of pipe cracks are completed.

[0050] Preferably, the energy management module provides stable power to the drone through a main energy source consisting of a lithium polymer battery pack, supporting its long-term operation. It is also equipped with a wireless charging interface to enable contactless charging of the drone at a designated location, thereby ensuring the system's continuous power supply and reusability.

[0051] Preferably, the environmental adaptability module achieves waterproof and dustproof protection through a shell with an IP68 protection rating, combined with an external structure protected by wings with an impact-resistant design, thereby improving the system's tolerance in complex environments. At the same time, the retractable stabilizer wing adjusts the attitude, enhancing the stability of the UAV in fluid environments, thus ensuring the safe operation and reliable operation of the equipment under harsh conditions.

[0052] Figure 2 A schematic diagram illustrating the overall flow of a pipe crack repair method according to one embodiment of this disclosure is shown. Figure 1 The method M200 shown includes steps S210 to S240. This method M200 can be performed by a repair device (e.g., a drone).

[0053] In step S210, pipe cracks in the digital twin model of the pipeline are identified, and the location and type of the cracks are determined. The digital twin model of the pipeline is determined based on the pipeline in the target area.

[0054] Based on the actual structure and scanning data of the pipeline in the target area, multimodal sensor data is integrated to construct a digital twin model of the pipeline. Anomalies on the pipeline surface are detected and analyzed in the digital twin model, and the precise spatial location and crack type (such as longitudinal, circumferential and weld cracks) of the pipeline cracks are identified, providing basic data support for subsequent risk assessment, prioritization and control command generation.

[0055] Preferably, the aforementioned pipeline digital twin model is determined based on the pipeline geometric model of the target area and the defect parameters of pipeline cracks concentrated in the pipeline crack dataset. Specifically, the pipeline geometric model is determined based on the 3D point cloud of the target area, and the defect parameters are determined based on the 3D model of the pipeline cracks. This achieves an accurate mapping of the pipeline digital twin model to the real physical environment of the pipeline in the target area. The pipeline geometric model realistically reproduces the shape, orientation, and structural features of the pipeline, while the defect parameters accurately characterize the location, size, and morphology of the cracks. The fusion of these two aspects enables the pipeline digital twin model to not only possess geometric fidelity but also dynamically reflect the detailed state of pipeline defects. This provides a high-precision, computable, and traceable virtual benchmark for subsequent crack identification, risk assessment, repair decisions, and autonomous UAV operations.

[0056] In one specific embodiment, prior to identifying pipe cracks in the digital twin model of the pipeline, the following may be included: Figure 3 Steps S310 to S340 are shown.

[0057] In step S310, the starting pose of the pipe into the target area of ​​the UAV is obtained. The starting pose is used to determine the spatial coordinates of the UAV relative to the global reference coordinate system.

[0058] Using UWB (Ultra-Wideband) positioning technology, the UAV accurately locates its position and acquires initial coordinates and starting pose when entering the manhole. The initial coordinates refer to the absolute position coordinates relative to the global reference coordinate system, initially determined by the UAV through the UWB system upon entering the manhole, typically expressed in the form of (x0, y0, z0). Here, x0 and y0 represent the UAV's two-dimensional position on the horizontal plane, usually with the center of the manhole opening as the reference origin; z0 represents the UAV's vertical height relative to the manhole opening plane (i.e., the initial height in the depth direction).

[0059] In step S320, a three-dimensional point cloud map of the pipeline is determined based on the starting point pose and the data from the pipeline geographic information system.

[0060] Load pipeline geographic information system (GIS) data containing pipe diameter, direction, and manhole location information to provide basic data support for subsequent navigation. Utilize laser SLAM (simultaneous localization and mapping) technology to generate a 3D point cloud map in real time.

[0061] In step S330, the detection path of the UAV is determined based on the three-dimensional point cloud map.

[0062] Based on a 3D point cloud map, a semantic segmentation algorithm is used to identify feature elements (such as flanges and branch pipes) within the pipeline, perform dynamic obstacle avoidance detection and path planning, and determine the detection path for the UAV.

[0063] In step S340, the pipeline is scanned and detected based on the detection path to determine the environmental perception data of the pipeline. The environmental perception data is used to construct a digital twin model of the pipeline.

[0064] Based on the detection path, a drone equipped with multiple sensors is controlled to travel along the inside of the pipeline, simultaneously acquiring lidar point clouds, visible light images, infrared thermal images, and pose data. The acquired multimodal sensor data is denoised, registered, and spatiotemporally aligned, then fused to generate 3D environmental information in a global reference coordinate system. Pipeline geometric features (such as pipe diameter and elbow locations), surface conditions, and abnormal areas are extracted and structured into environmental perception data containing spatial coordinates, texture, temperature, and other attributes, for subsequent construction of a digital twin model of the pipeline.

[0065] By acquiring the starting pose of the UAV entering the pipeline and fusing it with the pipeline's GIS data, precise positioning of the UAV in a global reference coordinate system was achieved. Based on this positioning information, a high-precision 3D point cloud map was constructed, and a detection path adapted to the actual pipeline structure was planned to ensure coverage of key areas. Subsequently, scanning and detection were carried out along this detection path to acquire environmental perception data containing geometric, visual, and thermal features. This provided a high-quality, spatially aligned data foundation for building a realistic, complete, and mappable digital twin model of the pipeline, improving the accuracy of the pipeline digital twin model and the reliability of subsequent autonomous inspection and repair operations.

[0066] In step S220, the hazard level of the pipe crack is determined based on the location and type of the crack.

[0067] Based on the identified locations (such as straight pipe sections, bends, welds, etc.) and crack types (such as longitudinal cracks, circumferential cracks, weld cracks, etc.) of pipeline cracks in the pipeline digital twin model, and combined with risk assessment rules or weight models, the risk assessment of cracks in different locations and types is carried out to determine their impact on the integrity of the pipeline structure. The resulting classification results, i.e., the risk level, provide a quantitative basis for subsequent repair priority ranking.

[0068] In one specific embodiment, determining the hazard level of a pipeline crack further includes: assessing the pipeline crack based on its location and type, determining a repair method, and using this repair method to generate control commands. For example, the assessment is performed by matching the location and type of the pipeline crack with a pre-defined repair strategy rule base. If the crack is located in a straight pipe section and is a surface micro-crack, adhesive injection repair is recommended. If it is located at an elbow or weld and is a circumferential crack, composite material wrapping or mechanical clamping is used for sealing. If it is a deep crack or a penetrating crack approaching 50% of the wall thickness, high-strength structural reinforcement is required. Based on the assessment results, a specific repair method is determined, including the process, material type, and operating parameters. This repair method is then converted into structured control commands containing the target location, execution actions, tool selection, and process parameters to drive the UAV to perform precise repair operations.

[0069] The above repair method follows the standard glue injection process, including pre-injection, main glue injection and pressure holding, and glue removal (to prevent dripping).

[0070] In step S230, a repair priority is determined based on the hazard level and the crack features extracted for the pipe cracks to prioritize the pipe cracks.

[0071] Based on the hazard level of the pipeline crack (e.g., length, depth, propagation trend, and / or pressure level of the pipe segment) and the crack features extracted for the pipeline crack, a weighted scoring method, rule engine, or priority matrix algorithm is used to quantify and score each pipeline crack, and a repair order list is generated according to the score, thereby forming an executable repair priority sequence, providing a sorting basis for the generation of subsequent control commands.

[0072] In step S240, control commands are generated based on the repair priority. These commands are used to control the UAV to repair the pipe cracks in the target area.

[0073] Based on repair priorities, and combining the digital twin model of the pipeline in the target area with the current location of the UAV, the optimal operation path is planned, and a sequence of control commands containing information such as target coordinates, repair methods, and process parameters is generated. The control commands are sent to the UAV through the communication link, driving it to autonomously navigate to the crack location, and calling on the robotic arm and repair module to complete the precise operation, realizing the mapping from virtual decision-making to physical execution.

[0074] For example, a visual servo control system is used to acquire real-time images of the relative position of the robotic arm and the pipe crack using a camera. Image feedback is used to adjust the robotic arm's movement, achieving dynamic tracking and high-precision positioning of the target location. Simultaneously, a force-position hybrid control algorithm is used to adjust the contact force of the end effector based on force sensor feedback when approaching and contacting the pipe wall, avoiding collisions or poor adhesion and ensuring operational stability. The delivery of repair material is performed by a screw pump, with a PID control algorithm adjusting the pump speed in real-time and precisely controlling the material output based on preset flow parameters, ensuring the uniformity and continuity of the adhesive injection process. This achieves coordinated operation of precise robotic arm positioning and stable material delivery.

[0075] In one specific embodiment, after repairing the pipe crack, the following may be included: Figure 4 Steps S410 to S420 are shown.

[0076] In step S410, the repaired pipe crack is inspected to determine the surface smoothness of the repaired pipe crack, the curing state of the repair material, and / or the bonding quality between the repair material and the pipe wall.

[0077] The repair surface is scanned using a structured light camera or laser profilometer, and surface smoothness is calculated through three-dimensional morphology analysis. Infrared thermal imagers are used to detect the temperature distribution and heat capacity characteristics of the repair material, and its curing state is determined by combining this with time-series data. Simultaneously, an ultrasonic probe emits sound waves into the repair area, and the adhesion quality between the repair material and the pipe wall is analyzed by the intensity and time difference of the echo signals to identify any debonding or voids. The above test results are fused and feature extracted to form quantitative indicators for smoothness deviation, curing degree, and adhesion integrity, which are used for subsequent repair quality assessment.

[0078] In step S420, a pipe repair report is generated based on surface smoothness, the curing state of the repair material, and / or the bonding quality.

[0079] Based on the surface flatness values ​​obtained from the inspection, the curing status of the repair material, and the bonding quality assessment results, each indicator is standardized and compared against thresholds. If the flatness is within the allowable tolerance range, the material is fully cured, and there is no debonding, the repair is deemed qualified. If any indicator exceeds the standard, it is marked as requiring rework, and the defect type is noted. The pipeline digital twin model, inspection data, judgment results, before-and-after images, and location information are integrated to generate a structured pipeline repair report. The format includes a title, repair number, inspection time, details of each indicator, comprehensive conclusions, and recommendations. The report is stored in PDF or JSON format or uploaded to the operation and maintenance management platform for personnel review and traceability.

[0080] By inspecting the repaired pipe cracks, the surface smoothness, curing state of the repair material, and adhesion quality to the pipe wall are obtained, enabling a multi-dimensional quantitative assessment of the repair effect. Based on the inspection results, a pipe repair report is generated, containing judgments on various indicators. This objectively verifies whether the repair quality meets standards, promptly identifies potential defects, ensures the reliability and durability of the repair, and improves the traceability and safety of pipeline operation and maintenance.

[0081] In one specific embodiment, after repairing the pipe crack, the process may further include: Figure 5 Steps S510 to S530 are shown.

[0082] In step S510, the final pose of the UAV at the current moment is obtained.

[0083] The drone's position and pose information inside the pipeline is acquired in real time through the positioning system (such as SLAM algorithm, inertial measurement unit IMU, fusion of odometry and lidar or visual sensor). After completing the current detection or repair task, the drone's final position coordinates and attitude angle in the global reference coordinate system are recorded as the endpoint pose at the current moment.

[0084] In step S520, the return path of the UAV is determined based on the 3D point cloud map, the battery level of the UAV, the starting pose and the ending pose.

[0085] Based on a 3D point cloud map, combined with the drone's starting pose and ending pose after mission completion, the optimal path back to the starting point from the current location is planned. Simultaneously, the drone's real-time battery level is read to assess remaining range. If the battery is sufficient, a path is planned based on the detected route; if the battery is insufficient, the shortest energy-efficient path is planned. The path planning process incorporates obstacle locations and pipeline topology to ensure the path is feasible and safe. Finally, a return path is generated, including a waypoint sequence, attitude adjustment commands, and energy management strategies.

[0086] In step S530, the drone is controlled to return to the starting point based on the return path.

[0087] Based on the waypoint sequence and attitude commands in the generated return path, the deviation between the UAV's current position and the return path is calculated in real time, and SLAM positioning and path tracking are performed using a 3D point cloud map. Heading correction is achieved by adjusting motor speed and control surfaces, navigating point by point to each intermediate waypoint, and stabilization devices are activated to enhance attitude control in curves or complex pipe sections. Simultaneously, battery level and environmental changes are continuously monitored, triggering emergency hovering or docking at the nearest facility in case of anomalies. Finally, the UAV is guided safely and stably back to the starting point along the planned path, and upon arrival, a shutdown or charging command is executed, completing the closed-loop return operation.

[0088] By acquiring the drone's current destination pose, combined with a 3D point cloud map, starting pose, and real-time battery power, and taking into account both spatial location and energy status, a safe and efficient return path is planned. Based on this path, the drone achieves autonomous navigation and closed-loop control, ensuring that it can reliably and accurately return to the starting point in complex pipeline environments, thereby improving the overall autonomy, safety, and mission completion rate of the system.

[0089] In one specific embodiment, it may also include, for example: Figure 6 Steps S610 to S620 are shown.

[0090] In step S610, the type of malfunction of the UAV is determined based on its operating status.

[0091] Based on the operational status of the drone, real-time analysis is performed using a fault diagnosis rule base or machine learning model. When a parameter exceeds the normal threshold range (such as motor overcurrent, abnormal attitude angle fluctuations, or sudden voltage drops), a fault detection mechanism is triggered. Multi-parameter correlation analysis is then used to determine the fault type.

[0092] The above operating status includes motor current, gyroscope and accelerometer readings, battery voltage, communication signal strength and / or sensor output, etc.

[0093] For example, multi-level sensor redundancy verification is set up to improve perception reliability through cross-comparison of multi-source data (such as IMU and visual odometry, LiDAR and structured light). At the same time, a heartbeat packet monitoring mechanism is established, with a communication timeout threshold set at 500ms. If no response is received for a continuous period of time, it is judged as abnormal.

[0094] In step S620, when the fault type is power failure, the positioning buoy of the UAV is released; when the fault type is material leakage, the corresponding valve of the UAV is shut off; when the fault type is communication interruption, the data collected by the UAV is stored in the local cache.

[0095] When a power failure is detected in the drone, the positioning buoy release mechanism is immediately triggered, causing it to detach from the drone body and float to the surface of the liquid in the pipeline or on the ground. It then transmits its location information wirelessly (e.g., via Bluetooth or RF) for easy location and rescue. When a material leak occurs, the control system automatically closes the corresponding valve connected to the storage tank within 50ms to prevent further leakage of repair materials, which could lead to waste or secondary pollution. When a communication interruption occurs, the drone encrypts and stores the currently collected environmental perception data, point cloud data, images, and mission status information to its local solid-state storage unit, ensuring this data is cached for at least 8 hours. It then enters a low-power standby mode, automatically uploading the cached data once communication is restored, ensuring that critical information is not lost.

[0096] By monitoring the operational status of drones in real time, the system accurately identifies fault types such as power failure, material leakage, and communication interruption, and executes corresponding emergency response measures for different faults. In the event of power failure, a positioning buoy is released for precise location, facilitating rapid search and rescue. In the event of material leakage, the corresponding valve is promptly shut off to prevent resource waste and environmental impact. In the event of communication interruption, collected data is stored in a local cache to ensure data integrity. This enhances the operational safety, mission reliability, and fault tolerance of drones in complex pipeline environments, ensuring the system's controllability and recoverability under abnormal conditions.

[0097] Furthermore, the pipeline crack repair method disclosed above is based on a digital twin model of the pipeline constructed in the target area. This model accurately identifies the location and type of the crack, then combines the location and crack type information to assess the crack's hazard level. Finally, it integrates crack characteristics such as crack length and depth for comprehensive analysis, scientifically determining the repair priority. Based on the repair priority, control commands are generated to drive the drone to perform repair operations in an orderly manner. This achieves closed-loop management from defect identification and risk assessment to repair decision-making and autonomous execution, significantly improving the intelligence, precision, and efficiency of pipeline operation and maintenance.

[0098] Regarding step S210, in some embodiments of this disclosure, determining the pipeline digital twin model based on the pipeline geometric model of the target area and the defect parameters of pipeline cracks in the pipeline crack dataset may include, for example: Figure 7 Steps S710 to S720 are shown.

[0099] In step S710, the defect parameters of the pipe cracks in the target area's pipe crack dataset are fused with the pipe geometric model to determine the pipe defect model.

[0100] The defect parameters in the target area pipeline crack dataset are spatially registered and fused with the pipeline geometric model constructed based on 3D point cloud. By aligning coordinates, the defect parameters of each pipeline crack are mapped to the corresponding position in the pipeline geometric model. Surface fitting or mesh deformation techniques are used to generate geometric depressions or fracture features on the model surface that are consistent with the actual defect morphology, forming a pipeline defect model containing real defect information. This achieves accurate reconstruction of the spatial distribution and morphology of cracks in physical pipelines in a virtual environment.

[0101] In step S720, labels containing material properties, stress distribution, and / or defect types of the pipeline are added to the pipeline defect model to determine the pipeline digital twin model.

[0102] Structured labeling information, including pipeline material properties (such as material type, elastic modulus, and yield strength), stress distribution data obtained from finite element analysis (such as high-stress areas and fatigue regions), and defect types (such as cracks, corrosion, and deformation), is associated with the constructed pipeline defect model through spatial coordinate and attribute mapping. Data fusion technology is then used to embed these labels into the corresponding regions of the pipeline defect model, forming a multi-dimensional data layer encompassing geometric morphology, defect features, and physical properties. Finally, all information is integrated to generate a complete digital twin model of the pipeline, enabling it not only to possess 3D visualization capabilities but also to support advanced functions such as risk assessment, life prediction, and intelligent decision-making.

[0103] For example, based on the 3D point cloud of the pipeline surface constructed from data collected by a structured light camera, a continuous and smooth mesh surface model, i.e., the pipeline geometric model, is generated through point cloud denoising, registration, and meshing algorithms, completing the basic geometric modeling. Defect regions identified jointly by structured light and ultrasonic probes are located within the mesh surface model. Combining surface morphology and internal depth information, their 3D geometric shape and internal extension structure are reconstructed, achieving refined modeling of the defect regions. Subsequently, material properties, stress distribution, and defect type labels are added to each region of the mesh surface model, constructing a pipeline digital twin model with multi-level information. A dynamic update mechanism is established by integrating historical detection data and real-time sensor information, supporting version iteration and health status tracking of the pipeline digital twin model over time. Finally, defect features such as crack length, depth, and orientation are extracted and visualized as highlighted areas in the pipeline digital twin model. Combined with clickable attribute labels and 3D cross-sectional views, this achieves intuitive presentation of defect information and interactive parameter querying.

[0104] By spatially fusing defect parameters from a pipeline crack dataset with a high-precision pipeline geometric model, a pipeline defect model reflecting the actual location and morphology of defects is constructed. Furthermore, multi-dimensional labeling information such as material properties, stress distribution, and defect type is integrated to generate a pipeline digital twin model with physical semantics and engineering characteristics. This achieves a complete mapping from geometric morphology to functional attributes, not only improving the refinement and realism of the pipeline digital twin model but also supporting intelligent analysis, risk assessment, and visual management of pipeline structural health status. This provides a high-fidelity, computable digital foundation for subsequent repair decisions and long-term monitoring.

[0105] Regarding step S210, in some other embodiments of this disclosure, the pipeline geometry model is determined based on the three-dimensional point cloud of the target region, and may include, for example: Figure 8 Steps S810 to S830 are shown.

[0106] In step S810, semantic segmentation is performed on the three-dimensional point cloud of the target region to determine the first three-dimensional point cloud containing pipeline structure features.

[0107] The 3D point cloud of the target area is preprocessed to remove noise and outliers. Then, a deep learning-based semantic segmentation model is used to classify and identify different objects by combining the spatial coordinates, normal vectors, curvature, and other geometric features of the point cloud. Based on the preset category labels, the point cloud belonging to the pipeline body (such as pipe walls, straight pipe sections, elbows, tees, etc.) is classified into one category. The first 3D point cloud containing the complete pipeline structural features is extracted, and non-structural interference point clouds such as supports, sediments, and cables are removed, achieving accurate separation and preservation of the main pipeline structure.

[0108] In step S820, the first three-dimensional point cloud and the pipeline temperature data are fused to determine a second three-dimensional point cloud containing the pipeline temperature data.

[0109] The first 3D point cloud is spatiotemporally aligned with the synchronously acquired pipeline temperature data. Using the pose information of the infrared thermal imager, the 2D temperature image is mapped to the corresponding region of the 3D point cloud. Coordinate transformation and interpolation algorithms are used to match each temperature value to its corresponding point cloud coordinates. For points without a direct correspondence, the nearest neighbor or inverse distance weighting method is used to assign temperature attributes, generating attribute information for each point including spatial coordinates, color, normal vector, and temperature value. This constructs a second 3D point cloud containing pipeline temperature data, achieving a fusion representation of geometric structure and thermal field information.

[0110] In step S830, the pipeline geometric model is determined based on the second three-dimensional point cloud.

[0111] The second 3D point cloud is denoised, simplified, and its normal vectors estimated. The discrete point cloud is then converted into a continuous triangular mesh surface using Poisson reconstruction or Delaunay triangulation algorithms. Mesh optimization (such as smoothing and hole repair) enhances the integrity and smoothness of the pipe geometry model, preserving structural features such as straight pipes, elbows, and tees. Simultaneously, temperature attributes from the second 3D point cloud are mapped to corresponding mesh vertices or faces, generating a pipe geometry model that integrates spatial morphology and temperature information.

[0112] By semantically segmenting the 3D point cloud of the target region, a first 3D point cloud containing pipeline structural features is accurately extracted, effectively eliminating non-structural interference such as supports and sediments. Then, pipeline temperature data is fused to generate a second 3D point cloud that combines geometric morphology and thermal field information. Finally, a pipeline geometric model is constructed based on this fused data, accurately reconstructing the pipeline's 3D structure and integrating temperature distribution features, achieving a collaborative expression of multimodal information.

[0113] Regarding step S210, in some other embodiments of this disclosure, the defect parameters are determined based on a three-dimensional model of the pipe crack, wherein the three-dimensional model of the pipe crack can be determined by including, for example... Figure 9 Steps S910 to S930 are determined as shown.

[0114] In step S910, based on the three-dimensional point cloud of the target area, the defect area containing the pipe crack is determined.

[0115] For example, it may include, Figure 10 Steps S1010 to S1030 are shown.

[0116] In step S1010, for the three-dimensional point cloud of the target area, the average temperature of the target point within the corresponding target range is determined, where the target point is the point in the three-dimensional point cloud located in the pipe structure.

[0117] For the 3D point cloud of the target area, target points belonging to the pipeline structure are extracted through semantic segmentation or geometric classification. Temperature data acquired by an infrared thermal imager is spatiotemporally aligned with the 3D point cloud, and each temperature value is associated with its corresponding point cloud coordinates through coordinate mapping. For each target point, a neighborhood range with a specified radius (e.g., 10 mm) is defined centered on it, and all neighboring points and their temperature values ​​within this range are searched. The average temperature of all points within this neighborhood is calculated as the average temperature of the target point within the corresponding target range, thus obtaining a smoothed temperature distribution reflecting the local thermal field characteristics.

[0118] In step S1020, the temperature difference of the target point is determined based on the difference between the temperature of the target point and the average temperature.

[0119] For each target point located on the pipeline structure, first obtain its actual temperature value mapped by an infrared thermal imager, and calculate its average temperature within a specified neighborhood range; then subtract its corresponding average temperature from the actual temperature of the point to obtain the temperature difference of the target point.

[0120] In step S1030, the defect area containing the pipe crack is determined based on the temperature difference at the target point.

[0121] Based on the temperature difference calculation results of each target point, all points on the pipeline structure in the 3D point cloud are traversed to identify anomalous points whose absolute temperature difference exceeds the target threshold (e.g., ≥2℃). Spatial clustering analysis is performed on these anomalous points to merge adjacent anomalous point groups, forming independent thermal anomaly regions. Secondary verification is then performed using point cloud geometric features (e.g., curvature, indentation depth) to eliminate false thermal anomalies caused by reflection, occlusion, or contamination. Finally, regions with concentrated temperature differences that conform to the thermal-mechanical coupling characteristics of cracks are identified as defect regions containing pipeline cracks.

[0122] For a target point located on the pipeline structure, the average temperature within its neighborhood is calculated, and the temperature difference between that point and the local mean is determined. This effectively eliminates the influence of ambient background temperature fluctuations and enhances sensitivity to minor thermal anomalies. Based on point-by-point temperature difference analysis, areas where the temperature difference significantly deviates from the normal range are identified. Combining spatial clustering and geometric feature verification, the location of potential cracks is accurately determined. The integration of three-dimensional geometric structure and thermal field distribution information improves the accuracy and anti-interference capability of defect detection.

[0123] In step S920, the defect area is scanned and detected to determine the surface geometric data and internal defect characteristics of the corresponding pipe crack.

[0124] For the defective area, a structured light camera is used to perform a high-precision 3D scan to acquire geometric data of the crack surface, including length, width, depth, and surface morphology. Simultaneously, an ultrasonic phased array probe is activated to emit sound waves into the crack area. By analyzing the time difference and energy attenuation of the received echo signals, the internal defect characteristics of the pipe crack are determined, such as its internal extension depth, direction, and whether it penetrates the pipe wall.

[0125] In step S930, a three-dimensional model of the pipe crack is determined based on surface geometric data and internal defect characteristics.

[0126] Surface geometric data acquired by structured light and internal defect features detected by ultrasonic testing are spatially registered and fused in a unified global reference coordinate system to form a multimodal defect dataset, comprehensively characterizing the three-dimensional structure and damage level of pipeline cracks. Based on the multimodal defect dataset, a three-dimensional model of the pipeline crack is generated, including the crack's full length, depth, orientation, and wall thickness penetration.

[0127] For example, a multi-line LiDAR is used to scan the interior of the pipeline to acquire a 3D point cloud. By coaxially mounting the LiDAR with an infrared thermal imager, synchronizing the time, and calibrating the intrinsic and extrinsic parameters, the data acquired by both are ensured to be strictly aligned in spatial viewpoint and time axis. The temperature data from the infrared image is mapped to the coordinates of the 3D point cloud, generating a 3D point cloud with temperature attributes. Areas with a temperature difference exceeding 2°C are identified as potential defects, i.e., defect areas. For these key areas, a structured light camera is used for sub-millimeter-level 3D surface reconstruction, while ultrasonic phased array technology is employed to accurately measure crack depth. The extrinsic parameter relationship between the structured light camera and the ultrasonic probe is obtained through calibration, unifying these two sets of data into a global reference coordinate system. ICP and other registration algorithms are used to spatially align the structured light surface model with the locations of internal defects detected by ultrasonic testing, establishing a correspondence between surface morphology and internal cracks. Finally, surface geometric features (such as pits and deformations) and internal defect features (depth, orientation, and extension direction) are fused to construct a multimodal defect dataset containing multi-source information, thereby determining the 3D model of the pipeline crack.

[0128] By identifying and extracting defect regions containing pipeline cracks from the 3D point cloud of the target area, precise location of anomalies is achieved. Multimodal scanning detection of the defect regions acquires surface geometric data and internal defect features of the cracks, comprehensively reflecting their morphology and structural damage. The surface and internal information are then fused to construct a high-precision 3D model of the pipeline crack, realizing a complete closed loop from macroscopic location to microscopic modeling, significantly improving the accuracy and completeness of defect characterization.

[0129] Regarding step S230, in some embodiments of this disclosure, it may include, for example... Figure 11 Steps S1110 to S1130 are shown.

[0130] In step S1110, feature extraction is performed on the pipeline crack to determine the crack characteristics of the pipeline crack.

[0131] By analyzing 3D point cloud data, images, and ultrasonic testing data of the pipeline crack area, its geometric and physical features are extracted. The length, width, depth, and orientation of the crack are calculated using point cloud curvature, surface depression depth, and contour fitting. Crack morphology is identified through image texture analysis. Internal extension depth and crack angle are obtained by combining ultrasonic echo signals. Simultaneously, location information (such as distance from the weld and orientation of the pipe segment) and propagation trend are statistically analyzed. Finally, these multi-dimensional parameters are structured to form a crack feature profile of the pipeline crack, encompassing geometric dimensions, spatial attributes, and morphological characteristics.

[0132] In step S1120, a risk score for the pipeline crack is determined based on crack characteristics, hazard level, and / or constraint conditions.

[0133] Based on the extracted crack features, hazard levels, and constraints, a weighted scoring model is used to calculate the risk score. After normalizing each feature parameter, a linear weighted sum is performed to output the comprehensive risk score for each pipeline crack, quantifying its degree of hazard.

[0134] The constraints mentioned above include the repair time for the pipe crack, the drone's endurance, and environmental factors.

[0135] In step S1130, a repair priority is determined based on the risk score to prioritize the pipe cracks.

[0136] In one specific embodiment, when the risk score is greater than or equal to a first threshold, the corresponding pipeline crack is determined as the first repair priority. When the risk score is less than the first threshold but greater than or equal to a second threshold, the corresponding pipeline crack is determined as the second repair priority, where the first threshold is greater than the second threshold, and the repair order of the second repair priority follows that of the first repair priority. When the risk score is less than the second threshold, the corresponding pipeline crack is determined as the third repair priority, and the repair order of the third repair priority follows that of the second repair priority. Through clear threshold division and priority order, the scientific nature, standardization, and feasibility of repair decisions are improved, ensuring the safety and operational efficiency of pipeline maintenance.

[0137] For example, multi-dimensional features of all pipeline cracks, including depth, length, and propagation trend, are extracted to assess the hazard level. A comprehensive priority assessment system is constructed by combining the preset time consumption of different repair processes, pipeline structural constraints (such as the impact of bends and valve positions on flight paths), UAV endurance (battery or fuel limitations), and environmental factors (such as the impact of wind speed and temperature on flight stability). A weighted scoring method is used to calculate the risk score (RS) for each pipeline crack. Based on the score results, cracks are divided into three levels: high-risk (urgent), medium-risk (planned), and low-risk (observation). The repair order is determined by prioritizing high-risk cracks, thus determining the repair priority. The repair priority is used for subsequent multi-objective optimized path planning to ensure that the UAV completes critical repair tasks efficiently and safely in limited resources and complex environments.

[0138] Preferably, the formula for the risk score calculated based on the weighted scoring method is as follows: Where RS represents the risk score, , , Here are the characteristic values ​​for each crack feature: depth represents the depth of the pipe crack, length represents the length of the pipe crack, stress represents the stress level of the pipe crack, and the subscript i represents the i-th pipe crack.

[0139] The technical solution of this invention will be further described below using an urban pipeline application scenario as an example. Those skilled in the art should understand that it can also be applied to other application scenarios besides urban pipelines.

[0140] Drones equipped with sensors such as lidar and infrared thermal imagers scan urban pipelines in target areas to acquire high-density 3D point clouds and multimodal environmental data. Based on the 3D point cloud data, noise is removed and mesh reconstruction is performed to generate a pipeline geometric model that accurately reflects the pipeline's orientation, diameter, bends, and other structural features.

[0141] Simultaneously, semantic segmentation and anomaly detection are performed on the point cloud to locate suspected crack areas. The surface morphology and internal defect features are extracted by fusing structured light and ultrasonic detection data to construct a three-dimensional model of the pipeline crack. Then, the defect parameters such as the length, depth, and direction of the crack are calculated to form a pipeline crack dataset.

[0142] By spatially aligning and mapping the defect parameters in the pipeline crack dataset with the pipeline geometric model, and integrating information such as material properties and stress distribution, a digital twin model of the pipeline containing real defect information is constructed.

[0143] By using a digital twin model of the pipeline, the location and type of pipeline cracks are identified, thereby determining the hazard level of the cracks. Based on the hazard level and the crack features extracted from the cracks, the repair priority for each crack is determined, generating a repair sequence. Subsequently, control commands are generated according to the repair priority to guide a drone to repair each pipeline crack.

[0144] After the repair is completed, the surface flatness, curing status and bonding quality are checked again, a repair report is generated and the pipeline digital twin model is updated, realizing the fully automated process from data acquisition, modeling and analysis to intelligent decision-making and closed-loop repair.

[0145] Based on any of the above embodiments, this disclosure also provides a pipe crack repair device.

[0146] Figure 12 This is a schematic block diagram of a pipe crack repair device according to one embodiment of the present disclosure.

[0147] like Figure 12 As shown, the pipe crack repair device includes: The identification module 1202 is used to identify pipeline cracks in the pipeline digital twin model and determine the location and type of the pipeline cracks. The rating determination module 1204 determines the hazard level of the pipeline crack based on its location and type. The sorting module 1206 determines the repair priority for sorting pipeline cracks based on the hazard level and crack features extracted for the pipeline cracks. The instruction generation module 1208 generates control instructions based on the repair priority. These control instructions are used to control the UAV to repair the pipe cracks in the target area.

[0148] The aforementioned pipeline crack repair device can be in the form of computer software, and each module of the aforementioned pipeline crack repair device can be implemented through computer software modules.

[0149] The specific implementation process of the functions and roles of each module in the above-mentioned pipeline crack repair device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0150] This disclosure also provides an electronic device 1000. Figure 13 A schematic diagram of the hardware implementation using the processing system is shown.

[0151] The hardware structure of electronic device 1000 can be implemented using a bus architecture. The bus architecture can include any number of interconnect buses and bridges, depending on the specific application and overall design constraints of the hardware. Bus 1100 connects various circuits including one or more processors 1200, memory 1300, and / or hardware modules. Bus 1100 can also connect various other circuits 1400 such as peripherals, voltage regulators, power management circuits, external antennas, etc. Bus 1100 can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Component (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, only one connection line is used in this figure, but this does not indicate that there is only one bus or one type of bus.

[0152] For ease of explanation, certain steps of the above method are described in relation to modules. It should be understood that the corresponding module performing one or more steps of the above method may be one or more hardware modules specifically configured to perform the corresponding step, or implemented by a processor configured to perform the corresponding step, or stored in a computer-readable medium for implementation by a processor, or implemented by some combination thereof.

[0153] This disclosure also provides a readable storage medium storing a computer program that, when executed by a processor, is used to implement the methods described above. A "readable storage medium" can be any means capable of containing, storing, communicating, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples of a readable storage medium include: an electrical connection with one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable read-only memory (CDROM), etc.

[0154] This disclosure also provides a computer program product, the methods of which can be implemented wholly or partially through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented wholly or partially as a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instructions are loaded and executed, all or part of the processes or functions of this disclosure are performed.

[0155] Computer programs or instructions can be stored in a readable storage medium or transferred from one readable storage medium to another. For example, the computer program or instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The readable storage medium can be any available medium capable of access, or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; an optical medium, such as a digital video optical disc; or a semiconductor medium, such as a solid-state drive. The computer-readable storage medium can be a volatile or non-volatile storage medium, or it can include both volatile and non-volatile types of storage media.

[0156] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied 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.

[0157] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0158] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0159] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0160] In the description of this specification, the references to terms such as "one embodiment / mode," "some embodiments / modes," "example," "specific example," or "some examples," etc., refer to specific features, structures, or characteristics described in connection with that embodiment / mode or example, which are included in at least one embodiment / mode or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment / mode or example. Moreover, the specific features, structures, or characteristics described may be combined in any suitable manner in one or more embodiments / modes or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments / modes or examples described in this specification, as well as the features of different embodiments / modes or examples.

[0161] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this disclosure, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0162] Those skilled in the art should understand that the above embodiments are merely for illustrating the present disclosure and are not intended to limit the scope of the disclosure. Those skilled in the art can make other changes or modifications based on the above disclosure, and these changes or modifications still fall within the scope of the present disclosure.

Claims

1. A method for repairing pipe cracks, characterized in that, include: Pipeline cracks in a pipeline digital twin model are identified to determine the location and type of the cracks, wherein the pipeline digital twin model is determined based on pipelines in a target area; The hazard level of the pipe crack is determined based on its location and type. Based on the hazard level and the crack features extracted for the pipe cracks, a repair priority is determined for ranking the pipe cracks. as well as Based on the repair priority, control commands are generated, which are used to control the UAV to repair the pipe cracks in the target area.

2. The method for repairing pipe cracks as described in claim 1, characterized in that, The pipeline digital twin model is based on pipeline determination in the target area and includes: A digital twin model of the pipeline is determined based on the pipeline geometry model of the target area and the defect parameters of the pipeline cracks in the pipeline crack dataset. The pipeline geometry model is determined based on the three-dimensional point cloud of the target area, and the defect parameters are determined based on the three-dimensional model of the pipeline cracks.

3. The method for repairing pipe cracks as described in claim 2, characterized in that, Based on the pipeline geometric model of the target area and the defect parameters of pipeline cracks in the pipeline crack dataset, a digital twin model of the pipeline is determined, including: The defect parameters of the pipe cracks in the target area's pipe crack dataset are fused with the pipe geometric model to determine the pipe defect model. Add labels containing the pipe's material properties, stress distribution, and / or defect types to the pipe defect model to determine the pipe digital twin model.

4. The method for repairing pipe cracks as described in claim 2, characterized in that, The pipeline geometry model is determined based on the three-dimensional point cloud of the target region, including: Semantic segmentation is performed on the 3D point cloud of the target region to determine the first 3D point cloud containing pipeline structure features; The first three-dimensional point cloud and the pipeline temperature data are fused to determine a second three-dimensional point cloud containing the pipeline temperature data; Based on the second 3D point cloud, the geometric model of the pipeline is determined.

5. The method for repairing pipe cracks as described in claim 2, characterized in that, The defect parameters are determined based on a three-dimensional model of the pipe crack, wherein the three-dimensional model of the pipe crack is determined through the following steps: Based on the 3D point cloud of the target area, the defect area containing the pipe crack is identified; The defective area is scanned and detected to determine the surface geometry data and internal defect characteristics of the corresponding pipe crack; Based on the surface geometry data and internal defect characteristics, a three-dimensional model of the pipe crack was determined.

6. The method for repairing pipe cracks as described in claim 5, characterized in that, Based on the 3D point cloud of the target region, the defect region containing the pipe crack is identified, including: For the three-dimensional point cloud of the target area, determine the average temperature of the target point within the corresponding target range, where the target point is the point located in the pipe structure in the three-dimensional point cloud; The temperature difference at the target point is determined based on the difference between the temperature at the target point and the average temperature. Based on the temperature difference at the target point, the defect area containing the pipe crack is determined.

7. The method for repairing pipe cracks as described in claim 1, characterized in that, Based on the crack features extracted from the pipe cracks and the hazard level, a repair priority is determined for ranking the pipe cracks, including: Feature extraction is performed on the pipeline crack to determine the crack characteristics of the pipeline crack; Based on the crack characteristics, the hazard level, and / or constraints, a risk score for the pipeline crack is determined. Based on the risk score, a repair priority is determined for ranking the pipe cracks; The determination of repair priorities for ranking the pipe cracks based on the risk score includes: When the risk score is greater than or equal to the first threshold, the corresponding pipeline crack is determined to be the first repair priority; When the risk score is less than the first threshold and greater than or equal to the second threshold, the corresponding pipeline crack is determined to be the second repair priority. The first threshold is greater than the second threshold, and the repair order of the second repair priority is after the repair order of the first repair priority. When the risk score is less than the second threshold, the corresponding pipeline crack is determined to be the third repair priority, and the repair order of the third repair priority is after the repair order of the second repair priority.

8. A system for repairing pipe cracks based on unmanned aerial vehicles (UAVs), characterized in that, A method for performing the repair of a pipe crack as described in any one of claims 1 to 7, comprising: The perception module is used to acquire multimodal perception data of the pipeline in the target area; The data processing module is used to establish a digital twin model of the pipeline using the multimodal sensing data, and to generate control commands based on the digital twin model of the pipeline. A repair module is used to repair pipe cracks according to the control commands; The energy management module is used to provide energy for the drone; An environmental adaptation module is used to provide the UAV with impact resistance, waterproofing, and / or resistance to water flow disturbances.

9. An electronic device, characterized in that, include: The memory stores execution instructions; as well as A processor that executes execution instructions stored in the memory, causing the processor to perform the pipe crack repair method according to any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for repairing pipe cracks as described in any one of claims 1 to 7.

Images