Intelligent inspection methods based on tethered drones and ground inspection robots

Abstract

This invention discloses an intelligent inspection method based on a tethered drone and a ground inspection robot, belonging to the field of drone inspection technology. This invention analyzes the drone's motion intentions in real time and enables the ground robot to intelligently switch between stationary locking and synchronous movement states. This ensures that the ground robot can follow the drone at all times while it is inspecting along the pipeline axis, thus maintaining the length of the composite cable within the optimal safety range, greatly improving the system's operating distance and safety. Furthermore, this invention overcomes the limitations of traditional offline planning, establishing a closed-loop adaptive control process that uses newly collected data during the inspection process to update the environmental model in real time. Based on the updated model and changes in the robot's own position, the safe length constraint of the cable is recalculated and adjusted. This gives the method the ability to learn and make autonomous decisions in a dynamically changing environment, significantly improving the intelligence level and environmental adaptability of the inspection.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) inspection technology, and specifically relates to an intelligent inspection method based on tethered UAVs and ground inspection robots. Background Technology

[0002] In critical infrastructure sectors such as oil, natural gas, electricity, and transportation, regular inspections of large, long-distance structures are crucial for ensuring stable operation and public safety. With the development of automation technology, the use of unmanned equipment for inspection tasks has become an industry trend. Tethered drones, connected to ground inspection equipment via composite cables, can obtain uninterrupted energy supply and high-speed data communication, making them suitable for long-term hovering observation and close-range operations, such as reading equipment meters or detecting gas leaks. This combination integrates the endurance of ground platforms with the flexibility of aerial platforms, and can generally replace traditional high-risk manual operations. However, in most solutions, the role of the ground robot is relatively passive, usually only serving as a fixed anchor point. This necessitates the deployment of extremely long cables when drones perform large-scale inspections of axially extending pipelines. These lengthy cables not only affect the drone's flight attitude and positioning accuracy due to their own weight and wind force, but also increase the risk of the cables being hooked or entangled by surrounding obstacles, severely limiting the system's operational range and safety. Furthermore, traditional methods typically rely on one-time offline path planning, where the entire path is set based on a known environmental model before the task begins. This rigid execution method cannot cope with uncertainties in actual operation, such as the deviation between the initial model and the actual scenario, and newly discovered obstacles during the inspection process. It lacks the ability to dynamically adjust and self-correct according to real-time environmental changes during task execution, resulting in poor robustness and autonomy of the system. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides an intelligent inspection method based on tethered drones and ground inspection robots to solve the aforementioned technical problems.

[0004] The intelligent inspection method based on tethered drones and ground inspection robots includes the following steps:

[0005] Based on the initial position of the ground inspection robot and the pre-acquired three-dimensional geometric model of the inspection object, the lower limit constraint of the composite cable is calculated, and the working length of the composite cable for deploying the tethered UAV to perform the inspection task is set according to the lower limit constraint; the working length is greater than or equal to the lower limit constraint.

[0006] During the inspection mission, collaborative control is activated. The collaborative control is used to switch the ground inspection robot between a stationary locked state and a synchronous moving state based on the motion vector of the tethered UAV.

[0007] After the ground inspection robot is displaced due to performing the synchronous movement state, the three-dimensional geometric model of the inspection object is updated using the three-dimensional data collected by the tethered UAV during the inspection task. Based on the updated three-dimensional geometric model and the current position of the ground inspection robot, the lower limit constraint of the composite cable is recalculated.

[0008] Preferably, when the target object is a horizontal pipe, the specific switching method of the collaborative control is as follows:

[0009] The movement of the tethered drone around the horizontal pipe is identified as a movement type that triggers a stationary locking state.

[0010] The movement of the tethered drone along the extension of the horizontal pipe is identified as a movement type that triggers a synchronized movement state.

[0011] Preferably, the movement of the tethered drone around the horizontal pipe is a zigzag circling path, in which a preset safe distance is always maintained between the center point of the tethered drone and the side wall surface of the horizontal pipe.

[0012] The path specifically includes: alternating between clockwise and counterclockwise directions to circle the horizontal pipe, and moving along the axial direction of the horizontal pipe after completing one circle.

[0013] Preferably, the synchronized movement of the ground inspection robot is achieved in the following way:

[0014] The projection of the motion vector of the tethered drone onto the axial direction of the horizontal pipe is analyzed to determine the axial displacement data;

[0015] The motion control commands for the ground inspection robot are generated based on the axial displacement data; the axial displacement data includes the displacement direction and displacement distance.

[0016] Preferably, the process of generating motion control commands for the ground inspection robot based on the axial displacement data further includes the following steps:

[0017] Before the tethered drone moves, its position data relative to the ground inspection robot is acquired; the position data includes its orientation, angle and distance relative to the ground inspection robot.

[0018] The relative positions of the ground inspection robot and the tethered drone are obtained based on location data; the relative position relationship is: the coordinates of the tethered drone in a three-dimensional coordinate system with the ground inspection robot as the origin;

[0019] After the tethered drone position is displaced, a navigation path is generated using a preset inertial navigation system based on the axial displacement data, with the aim of enabling the ground inspection robot to reproduce the displacement of the tethered drone in the axial direction of the horizontal pipeline.

[0020] The navigation path is used as the motion control command for the ground inspection robot.

[0021] Preferably, the calculation of the lower limit constraint of the composite cable specifically includes the following steps:

[0022] Based on the radius of the horizontal pipe and the preset safety distance, the distance of half a circumference of the tethered drone around the surface of the horizontal pipe is used as the first release amount;

[0023] Based on the initial position of the ground inspection robot and the pre-acquired three-dimensional geometric model of the inspection object, the second expansion is obtained;

[0024] The lower limit constraint of the composite cable is obtained by summing the lengths of the first expansion and the second expansion.

[0025] Preferably, the ground inspection robot is positioned within the orthographic projection of the horizontal pipe axis onto the ground below it.

[0026] The process of obtaining the second volume specifically includes the following steps:

[0027] The distance between the bottom of the horizontal pipe and the ground inspection robot in the vertical direction is used as the second measurement.

[0028] Preferably, when the ground inspection robot deviates laterally from the orthographic projection of the horizontal pipe axis onto the ground below it,

[0029] The process of obtaining the second volume specifically includes the following steps:

[0030] In a three-dimensional coordinate system, with the position of the ground inspection robot as the origin and the cable straightening as a constraint, the geometric tangent point from the origin to the surface of the horizontal pipe is calculated.

[0031] Calculate the distance between the origin and the geometric tangent point, and use this distance as the second scaling factor.

[0032] Preferably, when there is an obstruction between the ground inspection robot and the horizontal pipe...

[0033] The process of obtaining the second volume specifically includes the following steps:

[0034] In a three-dimensional coordinate system, with the position of the ground inspection robot as the origin, the three-dimensional geometric model of the obstruction is obtained, and the coordinates of all its constituent vertices are extracted.

[0035] For any vertex, the sum of the shortest path lengths to the origin and the surface of the horizontal pipe is calculated to obtain the candidate detour path length.

[0036] Among all the calculated candidate detour path lengths, the one with the smallest value is selected as the second release.

[0037] Preferably, the inspection task involves at least one of using a camera to read device meter readings or using a gas sensor to detect gas leak concentration.

[0038] The beneficial effects of this invention are as follows: By analyzing the movement intentions of the UAV in real time and enabling the ground robot to intelligently switch between stationary locking and synchronous movement states, this invention ensures that the ground robot can follow the UAV at all times while it is inspecting along the pipeline axis, thereby always maintaining the length of the composite cable within the optimal safety range, greatly improving the system's operating distance and safety. Furthermore, this invention overcomes the limitations of traditional offline planning, establishing a closed-loop adaptive control process that uses newly collected data during the inspection process to update the environmental model in real time. Based on the updated model and changes in the robot's own position, it recalculates and adjusts the cable's safe length constraints. This gives it the ability to learn and make autonomous decisions in a dynamically changing environment, significantly improving the intelligence level and environmental adaptability of the inspection. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a flowchart illustrating the steps of the intelligent inspection method based on tethered drones and ground inspection robots provided by the present invention.

[0041] Figure 2 This is a schematic diagram of the tethered drone's reversible circling path, which is part of the intelligent inspection method based on tethered drones and ground inspection robots provided by this invention.

[0042] Figure 3 This is a schematic diagram showing the relative positions of the ground inspection robot and the tethered drone in the intelligent inspection method based on the tethered drone and the ground inspection robot provided by the present invention. Detailed Implementation

[0043] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0044] The following disclosure provides many different embodiments or examples for implementing different structures of the present invention. To simplify the disclosure of the present invention, specific examples of components and arrangements are described below. Of course, these are merely examples and are not intended to limit the present invention.

[0045] The embodiments of the invention will now be described in detail with reference to the accompanying drawings.

[0046] like Figure 1 As shown, the intelligent inspection method based on tethered drones and ground inspection robots includes the following steps:

[0047] Based on the initial position of the ground inspection robot and the pre-acquired three-dimensional geometric model of the inspection object, the lower limit constraint of the composite cable is calculated, and the working length of the composite cable for deploying the tethered UAV to perform the inspection task is set according to the lower limit constraint; the working length is greater than or equal to the lower limit constraint.

[0048] During the inspection mission, collaborative control is activated. The collaborative control is used to switch the ground inspection robot between a stationary locked state and a synchronous moving state based on the motion vector of the tethered UAV.

[0049] After the ground inspection robot is displaced due to performing the synchronous movement state, the three-dimensional geometric model of the inspection object is updated using the three-dimensional data collected by the tethered UAV during the inspection task. Based on the updated three-dimensional geometric model and the current position of the ground inspection robot, the lower limit constraint of the composite cable is recalculated.

[0050] Before a mission begins, the ground inspection robot, acting as a mobile base station, loads a pre-acquired global 3D geometric model of the inspection object and combines it with its initial position coordinates determined by its own LiDAR or GPS. It then calculates the initial lower limit constraint of the composite cable. This lower limit constraint value is sent to the tethered drone, whose onboard automatic cable deployment and retraction system sets a working length greater than or equal to this constraint. After the mission begins, the drone flies along the path. At this time, the motion vector resolved by its onboard inertial navigation system is transmitted back to the ground robot's server in real time. The server then switches the robot's drive system to a stationary lock state, i.e., applies braking to the wheels. When the drone begins long-distance movement, its motion vector is resolved to trigger a synchronous movement state. In this state, the server aims to maintain a constant relative position between the drone and the robot, driving the robot to follow synchronously based on the drone's axial displacement data. For example, after the ground inspection robot has moved 10 meters synchronously, the drone uses its onboard laser ranging system to perform a new scan of the pipe section ahead, discovering a support structure not included in the original 3D model. The newly acquired 3D point cloud data is immediately transmitted back and used by the server to update the global 3D geometric model. Subsequently, a dynamic replanning process is triggered. Based on the robot's updated current position and the 3D model that now includes new obstacles, the lower limit constraint of the cable is recalculated, and the updated value is transmitted to the UAV to adjust the cable length. Through intelligent switching between synchronous movement and stationary locking, the problems of flight attitude instability and snagging risk caused by the need to release ultra-long cables due to long-distance axial movement of the UAV are solved. This allows the ground robot to continuously provide optimal support as a dynamic anchor point, greatly expanding the effective operating range of a single inspection. Furthermore, by using the newly acquired data to update the 3D model during task execution and dynamically adjusting the cable length constraint based on this closed-loop control mechanism, the ability to learn online and adapt autonomously in unknown or changing environments is achieved. This overcomes the rigidity of traditional offline planning and significantly improves the robustness and intelligence level of inspection operations.

[0051] More specifically, when the target object is a horizontal pipe, the specific switching method of the collaborative control is as follows:

[0052] The movement of the tethered drone around the horizontal pipe is identified as a movement type that triggers a stationary locking state.

[0053] The movement of the tethered drone along the extension of the horizontal pipe is identified as a movement type that triggers a synchronized movement state.

[0054] The high-performance server of the ground inspection robot stores the axis vector of the inspection object. When the tethered drone performs a task, its onboard inertial navigation system calculates its motion vector in three-dimensional space in real time at a high frequency and transmits it back to the server via optical fiber embedded in a composite cable. The server's built-in motion analysis module performs real-time vector calculations between the received motion vector and the stored pipe axis vector. In a specific embodiment, when the drone performs close-range observation or circling the pipe wall, the principal component of its motion vector will continuously be approximately perpendicular to the pipe axis vector. At this time, the motion analysis module determines that its motion type is "circling" and immediately sends a "stationary lock" command to the robot's drive system, that is, applies electromagnetic braking to the wheels to ensure the absolute stability of the ground base station. This mechanism provides a tethered anchor point for drones to perform hovering operations that require high precision, such as reading meters and capturing high-definition images, avoiding observation failures caused by base station shaking or cable dragging. When the UAV completes a circumferential inspection of a pipe segment and needs to move 5 meters along the pipe axis to the next inspection point, the principal component of its motion vector will be approximately parallel to the pipe axis vector. At this point, the motion analysis module determines that its motion type is "axial displacement" and triggers "synchronous movement state". In this state, the control system directly converts the UAV's axial displacement data into navigation path instructions for the ground robot, driving the robot to follow synchronously.

[0055] like Figure 2 As shown, more specifically, the movement of the tethered drone around the horizontal pipe is a zigzag circling path, in which the center point of the tethered drone and the side wall surface of the horizontal pipe are always kept at a preset safe distance.

[0056] The path specifically includes: alternating between clockwise and counterclockwise directions to circle the horizontal pipe, and moving along the axial direction of the horizontal pipe after completing one circle.

[0057] During execution, the drone launches from the 12 o'clock position directly above the pipe, first performing a 360-degree clockwise circle, passing the 6 o'clock position directly below the pipe, and returning to the 12 o'clock position directly above the pipe. During this process, its onboard laser rangefinder continuously measures the distance between the drone and the pipe wall at high frequency. Based on this real-time feedback, the flight control system performs closed-loop control to precisely maintain a preset safety distance, such as 0.5 meters, ensuring that the drone can collect complete data on half of the pipe's surface using its onboard camera without colliding with the pipe. This ensures safety while obtaining continuous, stable, and high-quality visual or sensor data of the target area. After completing the first circle, the drone translates along the pipe's axis by a preset distance, such as 1 meter. This preset distance can be set according to the diameter of the drone's detection range; in one embodiment, the preset translation distance is equal to the diameter of the detection range.

[0058] The second circling method is then executed, which involves a 360-degree counter-clockwise semi-circular flight. Starting from the 12 o'clock position directly above the pipe, it passes the 6 o'clock position directly below the pipe and returns to the 12 o'clock position directly above the pipe. This process also utilizes laser ranging for closed-loop distance control. The next circling is done clockwise, and so on. This reverse motion counteracts the torsional stress that might have been caused to the composite cable during the first clockwise motion, thus completely solving the problems of cable accumulation, knotting, and even self-entanglement caused by continuous unidirectional circling in existing technologies. In summary, the zigzag circling path of this invention achieves comprehensive data acquisition of the inspection target without blind spots, ensures the health status of the tethered cable, and improves the reliability and long-term autonomous operation capability of the automated inspection system.

[0059] More specifically, the synchronized movement of the ground inspection robot is achieved in the following way:

[0060] The projection of the motion vector of the tethered drone onto the axial direction of the horizontal pipe is analyzed to determine the axial displacement data;

[0061] The motion control commands for the ground inspection robot are generated based on the axial displacement data; the axial displacement data includes the displacement direction and displacement distance.

[0062] The ground robot server pre-loads a 3D model of the inspection object and extracts its central axis unit direction vector. When the collaborative control system switches to "synchronous movement state," the server begins to receive and process the 3D motion vectors transmitted back from the tethered UAV's inertial navigation system at high frequency in real time. In one implementation, when the UAV flies positively along the pipeline axis and the main component of its output motion vector is aligned with the pipeline axis vector direction, the motion vector is projected onto the pipeline axis direction through vector dot product. The result of this projection is directly quantified into a scalar with direction and magnitude, namely "axial displacement data." For example, if it is calculated that the UAV has a displacement of +2 meters in the axial direction within 0.5 seconds, this data is encapsulated into a motion control command. This precise decoupling and extraction of the axial progress, which is crucial to the inspection task, from the UAV's motion in complex 3D space provides a clear control input for the ground robot's following.

[0063] In another implementation, when the UAV is flying axially, it experiences slight lateral drift due to crosswinds, resulting in a component perpendicular to the pipe axis in its IMU-output motion vector. However, the projection algorithm of this invention automatically ignores this vertical component and extracts only the axial component. As a result, the motion control commands received by the ground robot will only include displacement along the pipe direction, without incorrectly following the UAV's lateral drift. This enhances the anti-interference capability and path-following robustness of the cooperative system, ensuring the ground robot consistently and stably moves along the predetermined inspection path. It achieves efficient following control, resolving the safety and stability issues caused by excessively long cables resulting from the robot's inability to dynamically coordinate with the UAV in existing technologies.

[0064] More specifically, when generating motion control commands for the ground inspection robot based on the axial displacement data, the process further includes the following steps:

[0065] Before the tethered drone moves, its position data relative to the ground inspection robot is acquired; the position data includes its orientation, angle and distance relative to the ground inspection robot.

[0066] The relative positions of the ground inspection robot and the tethered drone are obtained based on location data; the relative position relationship is: the coordinates of the tethered drone in a three-dimensional coordinate system with the ground inspection robot as the origin;

[0067] After the tethered drone position is displaced, a navigation path is generated using a preset inertial navigation system based on the axial displacement data, with the aim of enabling the ground inspection robot to reproduce the displacement of the tethered drone in the axial direction of the horizontal pipeline.

[0068] The navigation path is used as the motion control command for the ground inspection robot.

[0069] like Figure 3 As shown, in one embodiment, the tethered UAV measures the azimuth α, pitch angle β, and distance d of the UAV relative to the ground inspection robot using a laser ranging system and an optical tracking system. The position of the UAV is then calculated using the position (x0, y0, z0) of the ground inspection robot. The specific calculation formula is shown below:

[0070] (x0+d*cos(β)*cos(α),y0+ d*cos(β)*sin(α),z0+d*cos(β))

[0071] For example, the server, using its onboard camera and LiDAR, determines that the tethered drone is currently facing the robot, with a pitch angle of 30 degrees and a distance of 10 meters. It immediately converts this spherical coordinate system into a Cartesian coordinate point P(0, 8.66, 5) in the local coordinate system with the robot as the origin. This coordinate point P is then locked as the target's relative position vector during this synchronized movement. This step quantifies a vague target maintaining a relative position into a precise mathematical expectation value that can be used for closed-loop control. Subsequently, as the drone moves along the pipe axis, its axial displacement data is acquired by the server, for example, +15 meters. Upon receiving the +15-meter displacement data, the server's target calculation module immediately executes a control law. The goal of this control law is to drive the robot to move so that its new relative position vector with the drone converges back to the locked target vector P(0, 8.66, 5).

[0072] To address this, the server issues a navigation task to its own inertial navigation system: calculate and plan the optimal path from the current position to a point 15 meters above the current position along the axial displacement, and decompose this path into a series of wheel speed and steering commands as the final motion control command output. This simplifies the complex collaborative task into a single-point navigation problem for the robot itself, ensuring the accuracy and reliability of the movement. It also virtually connects two originally independent platforms into a rigid whole with a constant geometric configuration, fundamentally solving the technical problem of loose coordination between ground and air platforms and the inability to maintain optimal cable orientation in existing technologies.

[0073] More specifically, the calculation of the lower limit constraint of the composite cable includes the following steps:

[0074] Based on the radius of the horizontal pipe and the preset safety distance, the distance of half a circumference of the tethered drone around the surface of the horizontal pipe is used as the first release amount;

[0075] Based on the initial position of the ground inspection robot and the pre-acquired three-dimensional geometric model of the inspection object, the second expansion is obtained;

[0076] The lower limit constraint of the composite cable is obtained by summing the lengths of the first expansion and the second expansion.

[0077] More specifically, when the ground inspection robot is located within the orthographic projection of the horizontal pipeline axis onto the ground below it,

[0078] The process of obtaining the second volume specifically includes the following steps:

[0079] The distance between the bottom of the horizontal pipe and the ground inspection robot in the vertical direction is used as the second measurement.

[0080] The process begins with obtaining the first allowance. In one implementation, for a horizontal pipe with a radius of 0.4 meters, and setting a safe distance of 0.5 meters between the drone and the pipe wall during inspection, the server calculates the path length of the drone circling half a lap, i.e., π*(0.4m + 0.5m) ≈ 2.83m. This is the first allowance, representing the minimum cable length necessary to complete one core inspection action. This step precisely quantifies the geometric requirements of the inspection operation, forming the foundation for cable length calculation. Next, once the server confirms it is within the projected area, it obtains the Z-axis coordinate of the pipe bottom from the 3D model (e.g., 5.0m) and subtracts the Z-axis coordinate of the robot top (e.g., 1.2m), calculating a vertical distance of 3.8m. This distance is determined as the second allowance, representing the shortest straight cable length connecting the ground base station and the aerial operation starting point. Finally, the server sums the first and second allowances to obtain 6.63 meters, which is then used as the final lower limit constraint for the composite cable and sent to the drone's cable deployment and retrieval system. This approach dynamically and accurately calculates the cable length required to connect the robot to the work area based on the actual deployment location of the ground robot. By breaking down the cable requirements into workload and connection quantity and then summing them up after accurate calculation, it eliminates the high snagging risk and low flight stability problems caused by relying on manual estimation or using fixed ultra-long redundancy in existing technologies, and achieves automated, optimized, and scenario-adaptive setting of cable length.

[0081] More specifically, when the ground inspection robot deviates laterally from the horizontal pipeline axis by its orthogonal projection onto the ground below,

[0082] The process of obtaining the second volume specifically includes the following steps:

[0083] In a three-dimensional coordinate system, with the position of the ground inspection robot as the origin and the cable straightening as a constraint, the geometric tangent point from the origin to the surface of the horizontal pipe is calculated.

[0084] Calculate the distance between the origin and the geometric tangent point, and use this distance as the second scaling factor.

[0085] When the ground inspection robot deviates laterally from the projection area directly beneath the pipe, the server first establishes an origin in its internal three-dimensional coordinate system based on its own GPS / Lidar positioning data, and loads a mathematical model of an infinitely long cylinder defined by the pipe's axis vector and radius parameters. In one implementation, when the robot is positioned to the side and in front of the pipe, the server executes an analytical geometry algorithm. The goal of this algorithm is to solve for the equation of a straight line tangent to the surface of the cylinder, originating from the coordinate origin (the robot's position). Essentially, the problem is to find a point of tangency in three-dimensional space such that the plane formed by the robot, the point of tangency, and the pipe's axis is perpendicular to the plane containing the pipe's radius passing through the point of tangency.

[0086] By constructing and solving this set of constraint equations, the three-dimensional coordinates of the geometric tangency point can be uniquely determined. The beneficial effect of this step is that it abstracts a complex physical problem and transforms it into a mathematical problem with a definite solution, laying the foundation for precise calculation. Once the server calculates the coordinates of the tangency point, for example (10, 5, 8), it immediately calculates the Euclidean distance between that point and the origin, i.e., √(10²+5²+8²)≈13.75 meters. This calculated distance is defined as the absolute shortest length for the cable to connect to the pipe surface in a straight state at the current robot position, and is determined as the second allowance. Compared to existing technologies that cannot handle deviations from the scene or can only use excessive redundancy for fuzzy coverage, this invention achieves precise and automated calculation of the required cable length for any deviation position through rigorous geometric calculation. This not only ensures the optimization of cable length and avoids the risk of entanglement caused by unnecessary redundancy, but also greatly expands the effective operating range of the ground robot, freeing it from being limited to a narrow passage directly under the pipe, thereby significantly improving the deployment flexibility and environmental adaptability of the entire system.

[0087] More specifically, when there is an obstruction between the ground inspection robot and the horizontal pipeline,

[0088] The process of obtaining the second volume specifically includes the following steps:

[0089] In a three-dimensional coordinate system, with the position of the ground inspection robot as the origin, obtain the three-dimensional geometric model of the obstruction.

[0090] And extract the coordinates of all the vertices that make up the vertex;

[0091] For any vertex, the sum of the shortest path lengths to the origin and the surface of the horizontal pipe is calculated to obtain the candidate detour path length.

[0092] Among all the calculated candidate detour path lengths, the one with the smallest value is selected as the second release.

[0093] When an obstacle exists between the ground inspection robot and a horizontal pipe, the server first loads a pre-acquired 3D geometric model of the horizontal pipe and the obstacle, such as a cubic equipment box, into a 3D coordinate system with its own position as the origin. In a specific implementation, the server automatically extracts the 3D coordinates of all vertices from the equipment box model. Then, the server performs calculations for one vertex V: first, it calculates the straight-line distance L1 from the origin to vertex V; then, it calculates the shortest distance L2 from vertex V to the surface of the horizontal pipe, calculated as the perpendicular distance from vertex V to the central axis of the pipe minus the pipe radius. The server sums L1 and L2 to obtain the length of a candidate bypass path passing through vertex V. This creatively simplifies a complex nonlinear problem of a rope bypassing an obstacle in the physical world into a discrete optimization problem that iterates through the key feature points of the obstacle. The server repeats the aforementioned path length calculation for all vertices of the equipment box, thus obtaining multiple different candidate bypass path length values. Subsequently, the server selects the smallest value from a set containing these multiple values, for example, 11.2 meters. This minimum value is ultimately determined as the second release in the current obstructed scenario. The direct benefit of this traversal optimization method is that it automatically finds the shortest path connecting the robot and the pipeline without colliding with obstacles in a computationally efficient and logically complete manner, thus optimizing the cable length. In summary, compared to the limitations of existing technologies that are completely unable to operate when facing obstacles or can only rely on manual judgment to release excessively long cables, this invention provides an automated detour path calculation scheme by traversing and optimizing the obstacle vertices. This solves the problem of intelligent inspection deployment in non-line-of-sight scenarios, greatly improving the system's environmental adaptability and operational autonomy.

[0094] More specifically, the inspection task includes at least one of using a camera to read equipment meter readings and using a gas sensor to detect gas leak concentration.

[0095] When performing automatic readings of equipment meters: the smallest value is selected as the second measurement. When the inspection target is a pressure gauge located above a horizontal pipeline at a high altitude, the ground server will pre-calculate the optimal observation pose point based on the pre-loaded 3D model of the pipeline and meter. This pose point ensures that the high-resolution camera on the UAV can take pictures at an angle perpendicular to the dial, thereby eliminating the effects of parallax and reflection. After the task starts, the system switches to "static lock state". The ground robot remains stationary, and only through fine-tuning of the cable reel mechanism, the UAV is precisely controlled to reach the preset observation point and hover stably. Subsequently, the camera acquires high-definition images, which are processed by the optical character recognition algorithm in the server, and structured data such as 1.52MPa is automatically output.

[0096] When performing gas leak concentration detection tasks: When the inspection target is the hundreds of flange connection points along the pipeline, the drone will be equipped with a laser methane telemetry sensor and will continuously scan the pipeline surface along a looping path. During this process, the system moves synchronously throughout, and the ground robot steadily follows the pipeline's projection path on the ground, ensuring that the drone can complete a full-coverage scan of the entire pipeline without being limited by cable length. The concentration data collected by the sensor is linked to the real-time 3D coordinates provided by the drone, ultimately generating a visualized gas concentration cloud map overlaid on the 3D model of the pipeline on the server, thereby accurately locating the leak source.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention 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 or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. An intelligent inspection method based on tethered unmanned aerial vehicles (UAVs) and ground inspection robots, characterized in that, Includes the following steps: Based on the initial position of the ground inspection robot and the pre-acquired three-dimensional geometric model of the inspection object, the lower limit constraint of the composite cable is calculated, and the working length of the composite cable for deploying the tethered UAV to perform the inspection task is set according to the lower limit constraint; the working length is greater than or equal to the lower limit constraint. During the inspection mission, collaborative control is activated. The collaborative control is used to switch the ground inspection robot between a stationary locked state and a synchronous moving state based on the motion vector of the tethered UAV. After the ground inspection robot is displaced due to performing the synchronous movement state, the three-dimensional geometric model of the inspection object is updated using the three-dimensional data collected by the tethered UAV in the inspection task. Based on the updated three-dimensional geometric model and the current position of the ground inspection robot, the lower limit constraint of the composite cable is recalculated. The calculation of the lower limit constraint of the composite cable specifically includes the following steps: Based on the radius of the horizontal pipe and the preset safety distance, the distance of half a circumference of the tethered drone around the surface of the horizontal pipe is used as the first release amount; Based on the initial position of the ground inspection robot and the pre-acquired three-dimensional geometric model of the inspection object, the second expansion is obtained; The lower limit constraint of the composite cable is obtained by summing the lengths of the first expansion and the second expansion. When the ground inspection robot is located within the orthographic projection of the horizontal pipeline axis onto the ground below it. The process of obtaining the second volume specifically includes the following steps: The distance between the bottom of the horizontal pipe and the ground inspection robot in the vertical direction is used as the second measurement. When the ground inspection robot deviates laterally from the horizontal pipeline axis, its projection onto the ground below it is... The process of obtaining the second volume specifically includes the following steps: In a three-dimensional coordinate system, with the position of the ground inspection robot as the origin and the cable straightening as a constraint, the geometric tangent point from the origin to the surface of the horizontal pipe is calculated. Calculate the distance between the origin and the geometric tangent point, and use this distance as the second scaling factor; When there is an obstruction between the ground inspection robot and the horizontal pipeline. The process of obtaining the second volume specifically includes the following steps: In a three-dimensional coordinate system, with the position of the ground inspection robot as the origin, the three-dimensional geometric model of the obstruction is obtained, and the coordinates of all its constituent vertices are extracted. For any vertex, the sum of the shortest path lengths to the origin and the surface of the horizontal pipe is calculated to obtain the candidate detour path length. Among all the calculated candidate detour path lengths, the one with the smallest value is selected as the second release.

2. The intelligent inspection method based on tethered unmanned aerial vehicles and ground inspection robots according to claim 1, characterized in that, When the target object is a horizontal pipe, the specific switching method of the coordinated control is as follows: The movement of the tethered drone around the horizontal pipe is identified as a movement type that triggers a stationary locking state. The movement of the tethered drone along the extension of the horizontal pipe is identified as a movement type that triggers a synchronized movement state.

3. The intelligent inspection method based on tethered unmanned aerial vehicles and ground inspection robots according to claim 2, characterized in that, The movement of the tethered drone around the horizontal pipe is a zigzag circling path, in which the center point of the tethered drone and the side wall surface of the horizontal pipe are always kept at a preset safe distance. The path specifically includes: alternating between clockwise and counterclockwise directions to circle the horizontal pipe, and moving along the axial direction of the horizontal pipe after completing one circle.

4. The intelligent inspection method based on tethered unmanned aerial vehicles and ground inspection robots according to claim 3, characterized in that, The synchronized movement of the ground inspection robot is achieved in the following way: The projection of the motion vector of the tethered drone onto the axial direction of the horizontal pipe is analyzed to determine the axial displacement data; The motion control commands for the ground inspection robot are generated based on the axial displacement data; the axial displacement data includes the displacement direction and displacement distance.

5. The intelligent inspection method based on tethered unmanned aerial vehicles and ground inspection robots according to claim 4, characterized in that, When generating motion control commands for the ground inspection robot based on the axial displacement data, the process further includes the following steps: Before the tethered drone moves, its position data relative to the ground inspection robot is acquired; the position data includes its orientation, angle and distance relative to the ground inspection robot. The relative positions of the ground inspection robot and the tethered drone are obtained based on location data; the relative position relationship is: the coordinates of the tethered drone in a three-dimensional coordinate system with the ground inspection robot as the origin; After the tethered drone position is displaced, a navigation path is generated using a preset inertial navigation system based on the axial displacement data, with the aim of enabling the ground inspection robot to reproduce the displacement of the tethered drone in the axial direction of the horizontal pipeline. The navigation path is used as the motion control command for the ground inspection robot.

6. The intelligent inspection method based on tethered unmanned aerial vehicles and ground inspection robots according to claim 1, characterized in that, The inspection task involves at least one of the following: using a camera to read equipment meter readings and using a gas sensor to detect gas leak concentration.

Images