Wire repair robot control method and device, electronic equipment and storage medium

Abstract

The application belongs to the technical field of robot control, and discloses a control method and device of a wire repair robot, an electronic device and a storage medium, the method comprising: acquiring dynamic data, environment point cloud data, visual image data and surrounding wind field information of the wire repair robot, performing map construction to obtain dynamic environment map information, performing path planning according to the dynamic environment map information in combination with the dynamic data to obtain target planning path trajectory information of the wire repair robot, controlling the flying robot to fly to a target hovering position according to the target planning path trajectory information, and then putting a crawling repair robot onto a wire to be repaired, repairing the wire to be repaired based on an actuator control instruction signal of the crawling repair robot, actual motion state data and real-time environment wind speed information to obtain a repair result corresponding to the wire to be repaired; and through the above method, the repair efficiency of the wire repair robot is improved.

Description

Technical Field

[0001] This application pertains to the technology of the robotics industry within emerging industries, specifically relating to a control method, device, electronic equipment, and storage medium for a wire repair robot. Background Technology

[0002] Overhead power distribution lines are exposed to the elements for extended periods, making them highly susceptible to damage such as strand breakage and breakage due to factors like lightning strikes, corrosion, strong winds, and internal stress. If not addressed promptly, this can significantly reduce the tensile strength of the conductors, potentially leading to major safety accidents like line breakage or tower collapse. Currently, repairing these defects relies primarily on two traditional methods: first, manual tower climbing and line relocation after a power outage. This method is labor-intensive, inefficient, and carries serious safety risks such as falls from heights and electric shocks; second, the use of mechanical methods such as insulated bucket trucks. However, in complex terrain conditions such as mountainous areas or long spans, these methods are often limited by working space and safety distances, making them difficult to implement effectively.

[0003] In recent years, some studies have attempted to use drones equipped with repair robots for conductor repair operations, but existing technologies still have many shortcomings. Current robot designs are generally bulky, with mechanical structures and control systems separated, failing to adequately consider electromechanical coupling effects. This results in poor robot stability during actual operations, making it difficult to effectively adapt to conductor sag changes and complex wind field disturbances.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] The purpose of this application is to provide a control method, device, electronic device, and storage medium for a wire repair robot, in order to solve the problem that existing wire repair methods cannot accurately control the wire repair robot and thus affect the repair progress due to neglecting dynamic environmental factors and sudden disturbance factors. It can take into account dynamic environmental factors and sudden disturbance factors, ensure the feasibility of the planned path in the actual dynamic environment, effectively cope with sudden disturbance factors such as strong winds at high altitudes and wire galloping, and improve the repair efficiency of the wire repair robot.

[0006] In a first aspect, this application provides a control method for a wire repair robot, comprising: Acquire multimodal perception data of the wire repair robot; the multimodal perception data includes the robot's dynamic data, environmental point cloud data, visual image data, and surrounding wind field information; Based on the environmental point cloud data, the visual image data, and the surrounding wind field information, a map is constructed to obtain the dynamic environmental map information of the conductor repair robot regarding the conductor to be repaired. Based on the dynamic environment map information and the dynamic data, the path planning of the wire repair robot is performed to obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired; According to the target planned path trajectory information, after controlling the flying robot in the wire repair robot to fly to the target hovering position corresponding to the wire to be repaired, the crawling repair robot in the wire repair robot is deployed onto the wire to be repaired; Based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and real-time environmental wind speed information, the crawling repair robot is controlled to repair the wire to be repaired, and the repair result corresponding to the wire to be repaired is obtained.

[0007] The control method for the conductor repair robot provided in this application enables control of the conductor repair robot. By constructing a dynamic environmental map based on environmental point cloud data, visual image data, and surrounding wind field information, path planning is performed to obtain the target planned path trajectory information. After the flying robot in the conductor repair robot flies to the target hovering position, the crawling repair robot is deployed onto the conductor to be repaired. The crawling repair robot then repairs the conductor, yielding the corresponding repair result. This method solves the problem of existing conductor repair methods failing to accurately control the conductor repair robot and affecting repair progress due to neglecting dynamic environmental factors and sudden disturbances. It considers dynamic environmental factors and sudden disturbances, ensuring the feasibility of the planned path in a real dynamic environment. It effectively copes with sudden disturbances such as strong high-altitude winds and conductor galloping, improving the repair efficiency of the conductor repair robot.

[0008] Optionally, based on the environmental point cloud data, the visual image data, and the surrounding wind field information, a map is constructed to obtain dynamic environmental map information of the conductor repair robot regarding the conductor to be repaired, including: Based on the environmental point cloud data and the environmental image information in the visual image data, a map is constructed to obtain the static geometric map information of the wire repair robot; Based on the static geometric map information, the surrounding wind field information, the conductor image information in the visual image data, and the conductor feature point cloud data in the environmental point cloud data are fused to obtain the dynamic environmental map information of the conductor repair robot regarding the conductor to be repaired.

[0009] The control method for the conductor repair robot provided in this application can control the conductor repair robot. By effectively integrating static geometric information with dynamic environmental information (such as wind field and conductor images), a more comprehensive and accurate dynamic environmental map is constructed, providing a reliable environmental perception basis for subsequent path planning and repair operations, and overcoming the limitation of traditional static maps that cannot reflect dynamic changes.

[0010] Optionally, based on the dynamic environment map information and the dynamic data, path planning is performed on the wire repair robot to obtain the target planned path trajectory information of the wire repair robot flying from its current position to the wire to be repaired, including: Obtain the pre-built route planning comfort function; Based on the path planning comfort function, the dynamic environment map information and the dynamic data are used to perform path planning, and the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired is obtained.

[0011] Optionally, based on the path planning comfort function, the dynamic environment map information and the dynamic data are used to perform path planning to obtain the target planned path trajectory information of the conductor repair robot flying from its current position to the conductor to be repaired, including: According to the preset hovering criteria, the target hovering position corresponding to the conductor to be repaired is determined in the dynamic environment map information; Based on the target hovering position and combined with the dynamic environment map information, the path planning comfort function is used to perform path planning, thereby obtaining the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired.

[0012] Optionally, based on the target hovering position and combined with the dynamic environment map information, path planning is performed using the path planning comfort function to obtain the target planned path trajectory information of the wire repair robot flying from its current position to the wire to be repaired, including: Based on preset dynamic constraints, flow field constraints, and geometric constraints, and combined with the dynamic environment map information and the target hovering position, path planning is performed to obtain the path trajectory information of the wire repair robot flying from the current position to the target hovering position and the path parameters corresponding to the path trajectory information. The path parameters corresponding to each of the path trajectory information are input into the path planning comfort function for calculation to obtain the path planning comfort value corresponding to each of the path trajectory information; The minimum value among the path planning comfort values ​​is determined, and the path trajectory information corresponding to the minimum value is determined as the target planned path trajectory information for the conductor repair robot to fly from the current position to the conductor to be repaired.

[0013] Optionally, based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and real-time environmental wind speed information, the crawling repair robot is controlled to repair the conductor to be repaired, and the repair result corresponding to the conductor to be repaired is obtained, including: Acquire the actuator control command signals, actual motion state data, and real-time environmental wind speed information of the crawling repair robot; Based on the actuator control command signal of the crawling repair robot, prediction is performed to obtain the target motion state data corresponding to the actuator control command signal; Based on the difference between the target motion state data and the actual motion state data, and combined with the real-time environmental wind speed information, the crawling repair robot is controlled to repair the conductor to be repaired using a dynamic switching control method, thereby obtaining the repair result corresponding to the conductor to be repaired.

[0014] Optionally, based on the difference between the target motion state data and the actual motion state data, and combined with the real-time environmental wind speed information, a dynamic switching control method is used to control the crawling repair robot to repair the conductor to be repaired, thereby obtaining the repair result corresponding to the conductor to be repaired, including: Calculate the difference between the target motion state data and the actual motion state data; When the difference is less than a preset error threshold and the real-time wind speed value in the real-time environmental wind speed information is less than a preset disturbance wind speed threshold, a preset model-free adaptive control algorithm is used to control the crawling repair robot to repair the conductor to be repaired. When the difference is greater than or equal to a preset error threshold, or when the real-time wind speed value in the real-time environmental wind speed information is greater than or equal to a preset disturbance wind speed threshold, a preset active disturbance rejection control algorithm is used to control the crawling repair robot to repair the conductor to be repaired. When the sum of the number of times the model-free adaptive control algorithm and the active disturbance rejection control algorithm are used within a preset period is greater than or equal to a preset switching threshold, the algorithm usage weight coefficient is determined based on the historical usage number of the model-free adaptive control algorithm and the active disturbance rejection control algorithm. Based on the algorithm usage weight coefficient, and combined with the model-free adaptive control algorithm and the active disturbance rejection control algorithm, a compliant control quantity is generated to control the crawling repair robot to repair the wire to be repaired according to the compliant control quantity. After confirming that the wire to be repaired has been repaired, the repair result corresponding to the wire to be repaired is obtained.

[0015] The control method for the wire repair robot provided in this application enables control of the wire repair robot. By dynamically switching between model-free adaptive control and active disturbance rejection control algorithms, and introducing compliant control variables, it achieves precise and robust control of the crawling repair robot under different disturbance conditions, effectively improving the accuracy and reliability of the repair operation.

[0016] Secondly, this application provides a control device for a wire repair robot, comprising: The acquisition module is used to acquire multimodal perception data of the wire repair robot; the multimodal perception data includes the dynamic data of the wire repair robot, environmental point cloud data, visual image data and surrounding wind field information; The construction module is used to construct a map based on the environmental point cloud data, the visual image data and the surrounding wind field information, so as to obtain the dynamic environmental map information of the conductor repair robot about the conductor to be repaired. The planning module is used to perform path planning for the wire repair robot based on the dynamic environment map information and the dynamic data, so as to obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired; The control module is used to control the flying robot in the wire repair robot to fly to the target hovering position corresponding to the wire to be repaired according to the target planned path trajectory information, and then deploy the crawling repair robot in the wire repair robot onto the wire to be repaired. The repair module is used to control the crawling repair robot to repair the wire to be repaired based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and the real-time environmental wind speed information, so as to obtain the repair result corresponding to the wire to be repaired.

[0017] The control device of this conductor repair robot performs path planning based on a dynamic environmental map constructed from environmental point cloud data, visual image data, and surrounding wind field information. The resulting target planned path trajectory information is used to control the flying robot within the conductor repair robot to fly to the target hovering position. Then, the crawling repair robot is deployed onto the conductor to be repaired, allowing the crawling repair robot to perform the repair and obtain the corresponding repair result. This addresses the problem in existing conductor repair methods where neglecting dynamic environmental factors and sudden disturbances leads to inaccurate control of the conductor repair robot, thus affecting repair progress. It considers dynamic environmental factors and sudden disturbances, ensuring the feasibility of the planned path in actual dynamic environments. It effectively copes with sudden disturbances such as strong high-altitude winds and conductor galloping, improving the repair efficiency of the conductor repair robot.

[0018] Thirdly, this application provides an electronic device including a processor and a memory, the memory storing a computer program executable by the processor, wherein when the processor executes the computer program, it performs the steps in the control method for the wire repair robot described above.

[0019] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the steps in the control method for the wire repair robot described above.

[0020] Beneficial Effects: The control method, device, electronic equipment, and storage medium for the conductor repair robot provided in this application utilize dynamic environmental map information constructed based on environmental point cloud data, visual image data, and surrounding wind field information to perform path planning. The resulting target planned path trajectory information is used to control the flying robot within the conductor repair robot to fly to the target hovering position, and then deploys the crawling repair robot onto the conductor to be repaired. This allows the crawling repair robot to repair the conductor, yielding the corresponding repair result. This addresses the problem in existing conductor repair methods where neglecting dynamic environmental factors and sudden disturbances leads to inaccurate control of the conductor repair robot, thus affecting repair progress. It considers dynamic environmental factors and sudden disturbances, ensuring the feasibility of the planned path in actual dynamic environments. It effectively copes with sudden disturbances such as strong high-altitude winds and conductor galloping, improving the repair efficiency of the conductor repair robot. Attached Figure Description

[0021] Figure 1 A flowchart illustrating the control method for the wire repair robot provided in this application embodiment.

[0022] Figure 2 This is a schematic diagram of the control device for the wire repair robot provided in an embodiment of this application.

[0023] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0024] Labeling Explanation: 1. Acquisition Module; 2. Construction Module; 3. Planning Module; 4. Control Module; 5. Repair Module; 301. Processor; 302. Memory; 303. Communication Bus. Detailed Implementation

[0025] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0026] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0027] In recent years, some studies have attempted to use drones equipped with repair robots for conductor repair operations, but existing technologies still have many shortcomings. Specifically, current repair robots mainly achieve environmental perception through visual inspection. However, traditional visual inspection is mainly based on strand recognition of RGB images, which is susceptible to changes in lighting conditions and electromagnetic interference, resulting in poor robustness. Furthermore, the path planning of existing repair robots is mostly based on static geometric maps, failing to consider the two-way fluid-structure interaction between dynamic environmental factors such as conductor flexibility vibration and wind load and the robot's own dynamics. This makes the planned path infeasible in actual dynamic environments and results in large tracking errors.

[0028] Furthermore, existing repair robots employ a single control mode, resulting in response lag. Specifically, the control strategies of existing repair robots are mainly based on traditional PID or model predictive control (MPC). However, when faced with sudden disturbances such as strong winds at high altitudes or conductor galloping, the model mismatch becomes severe, and control parameters are difficult to adjust adaptively, leading to response lag and significantly impacting repair accuracy. These shortcomings severely restrict the level of intelligence and engineering practicality of conductor repair operations.

[0029] To address the problem in existing technologies where neglecting dynamic environmental factors and sudden disturbances leads to the inaccurate control of wire repair robots and affects repair progress, this application provides a control method, device, electronic device, and storage medium for a wire repair robot. This effectively addresses sudden disturbances such as strong winds at high altitudes and wire galloping, enabling high-precision autonomous operation of the wire repair robot in complex environments and improving its repair efficiency.

[0030] Please refer to Figure 1 , Figure 1 This application discloses a control method for a wire repair robot in some embodiments, used to control the wire repair robot, including the following steps: Step S101: Acquire multimodal perception data of the wire repair robot; the multimodal perception data includes the dynamic data of the wire repair robot, environmental point cloud data, visual image data, and surrounding wind field information; Step S102: Based on environmental point cloud data, visual image data and surrounding wind field information, a map is constructed to obtain dynamic environmental map information of the wire repair robot about the wire to be repaired. Step S103: Based on the dynamic environment map information and combined with the dynamic data, perform path planning for the wire repair robot to obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired. Step S104: According to the target planned path trajectory information, control the flying robot in the wire repair robot to fly to the target hovering position corresponding to the wire to be repaired, and then deploy the crawling repair robot in the wire repair robot onto the wire to be repaired. Step S105: Based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and the real-time environmental wind speed information, control the crawling repair robot to repair the wire to be repaired, and obtain the repair result corresponding to the wire to be repaired.

[0031] This control method for the conductor repair robot utilizes a dynamic environmental map constructed based on environmental point cloud data, visual image data, and surrounding wind field information. Path planning is performed to obtain the target planned path trajectory information. The flying robot within the conductor repair robot flies to the target hovering position, and then the crawling repair robot is deployed onto the conductor to be repaired. The crawling repair robot then repairs the conductor, yielding the corresponding repair result. This method addresses the problem in existing conductor repair methods where neglecting dynamic environmental factors and sudden disturbances leads to inaccurate control of the repair robot, thus affecting repair progress. It considers dynamic environmental factors and sudden disturbances, ensuring the feasibility of the planned path in actual dynamic environments. It effectively copes with sudden disturbances such as strong high-altitude winds and conductor galloping, improving the repair efficiency of the conductor repair robot.

[0032] Specifically, in step S101, multimodal perception data of the conductor repair robot is acquired. This multimodal perception data includes the robot's dynamics data, environmental point cloud data, visual image data, and surrounding wind field information. The conductor repair robot is an integrated system comprising a flying robot and a crawling repair robot. The flying robot transports the crawling repair robot to the vicinity of the conductor to be repaired and deploys it onto the conductor. The crawling repair robot then moves and performs repair work on the conductor. Multimodal perception data refers to various types of data acquired from different sensors, including dynamics data, environmental point cloud data, visual image data, and surrounding wind field information. Dynamics data can include information such as the robot's attitude, velocity, and acceleration, and can be acquired through an inertial measurement unit (MIU). Environmental point cloud data describes the three-dimensional geometric information of the surrounding environment and can be acquired through lidar. Visual image data provides two-dimensional image information of the environment and can be acquired through a camera. The surrounding wind field information describes meteorological conditions such as wind speed and direction within the robot's operating area and can be acquired through an ultrasonic anemometer mounted on the conductor repair robot or through a third-party meteorological service platform.

[0033] Specifically, in step S102, based on environmental point cloud data, visual image data, and surrounding wind field information, a map is constructed to obtain dynamic environmental map information of the conductor repair robot regarding the conductor to be repaired, including: Based on the environmental image information in the environmental point cloud data and visual image data, a map is constructed to obtain the static geometric map information of the wire repair robot; Based on static geometric map information, and by integrating surrounding wind field information, conductor image information from visual image data, and conductor feature point cloud data from environmental point cloud data, dynamic environmental map information about the conductor to be repaired is obtained for the conductor repair robot.

[0034] In step S102, static geometric map information refers to a fixed and unchanging geometric structure representation of the environment constructed by processing and analyzing environmental image information from environmental point cloud data and visual image data. This static geometric map information is mainly used to describe the spatial distribution and geometric characteristics of static environmental elements such as terrain, buildings, fixed obstacles, and (relatively static) guide wires to be repaired, providing a stable and fundamental environmental reference for the guide wire repair robot. The static geometric map information can be represented in the form of a raster map, point cloud map, or 3D model.

[0035] By overlaying real-time dynamic information onto a static geometric map, a comprehensive dynamic environment map is formed, providing the wire guide repair robot with a dynamic environment map of the wire to be repaired. Specifically, surrounding wind field information is used to capture airflow disturbances in the environment, while wire guide image information and wire guide feature point cloud data focus on the precise location, orientation, and potential minor deformations of the wire to be repaired. This fusion of multi-source information ensures that the constructed dynamic environment map not only includes the static geometric structure of the environment but also reflects real-time dynamic changes within the wire guide repair robot's operating area, such as the impact of wind on the wire and the wire's real-time state. The wire guide image information and wire guide feature point cloud data can be obtained through existing technologies, such as the YOLOv8 semantic segmentation network, by performing wire recognition on the visual image data and environmental point cloud data, respectively.

[0036] The dynamic environment map information can be calculated using the following formula: ; in, For dynamic environment map information; The instantaneous wind speed at location x at time t can be calculated using the extended Kalman filter algorithm based on the surrounding wind field information. The pose information of the location point x of the conductor to be repaired at time t.

[0037] Specifically, in step S103, based on the dynamic environment map information and combined with dynamic data, path planning is performed on the wire repair robot to obtain the target planned path trajectory information of the wire repair robot flying from its current position to the wire to be repaired, including: Obtain the pre-built route planning comfort function; Based on the path planning comfort function, dynamic environment map information and dynamic data are used to perform path planning, and obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired.

[0038] In step S103, the path planning comfort function is a mathematical model that comprehensively evaluates path quality, aiming to quantify the path's performance across multiple dimensions such as safety, efficiency, energy consumption, and stability. This function is typically pre-designed and constructed based on the characteristics of the wire repair robot, the environmental features of the wire to be repaired, and the task requirements. For example, the path planning comfort function may include multiple sub-items such as flight time and energy consumption, which are then weighted and summed or combined in other ways to form a single comfort index. Pre-construction means that the function has been defined, parameterized, and stored in the control system or related computing unit of the wire repair robot before the path planning process begins, so that it can be directly invoked during path planning.

[0039] The path planning comfort function is specifically as follows: ; in, This represents the comfort value for route planning. For flight time; Energy consumption; The safety penalty coefficient is inversely proportional to the distance from the wire repair robot to the obstacle (including the wire to be repaired); This is a weighted value based on flight time. This is the weighted value for energy consumption; This is the weight value for security penalties.

[0040] Optionally, in step S103, based on the path planning comfort function, path planning is performed using dynamic environment map information and dynamic data to obtain the target planned path trajectory information of the wire repair robot flying from its current position to the wire to be repaired, including: According to the preset hovering criteria, the target hovering position corresponding to the guide wire to be repaired is determined in the dynamic environment map information; Based on the target hovering position and combined with dynamic environment map information, the path planning comfort function is used to plan the path, and obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired.

[0041] Specifically, in step S103, before path planning, the "target hovering position" that the flying robot needs to reach is accurately determined in the dynamic environment map information based on preset hovering criteria. This target hovering position is determined according to the preset hovering criteria and the dynamic environment map information. For example, an optimal hovering point can be determined based on the geometry of the guide wire, the characteristics of the repair area, and the characteristics of the flying robot's delivery mechanism, ensuring the smoothest and safest delivery process. The determination of this target hovering position provides a clear and optimized endpoint for subsequent path planning.

[0042] The preset hovering criteria consist of a series of rules or conditions used to evaluate and select the optimal hovering point. These criteria may include, but are not limited to: the relative distance and angle to the guide wire to be repaired, the safe distance from surrounding obstacles, the attitude and dynamic limitations of the flying robot itself, and the feasibility of deploying the crawling repair robot. The preset hovering criteria ensure that the flying robot can hover stably in a position that is both safe and conducive to subsequent repair work.

[0043] Optionally, in step S103, based on the target hovering position and combined with dynamic environment map information, a path planning comfort function is used to perform path planning, obtaining the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired, including: Based on preset dynamic constraints, flow field constraints, and geometric constraints, combined with dynamic environment map information and target hovering position, path planning is performed to obtain the path trajectory information and path parameters corresponding to the path trajectory information of the wire repair robot flying from the current position to the target hovering position. The path parameters corresponding to each path trajectory information are input into the path planning comfort function for calculation, and the path planning comfort value corresponding to each path trajectory information is obtained. Determine the minimum value among the comfort values ​​of each path planning, and use the path trajectory information corresponding to the minimum value as the target planned path trajectory information for the wire repair robot to fly from the current position to the wire to be repaired.

[0044] Specifically, in step S103, the dynamic constraints refer to the limitations on the flight performance of the wire repair robot. For example, the aerial robot (generally a multi-rotor flight platform) and the robotic arm (generally a multi-joint robotic arm) of the flying robot in the wire repair robot are regarded as a whole to limit the motion data of the flying robot, thereby constructing a dynamic equation representing the overall flying robot. The dynamic equation of the flying robot is as follows: ; Where q represents the pose information of the flying robot, namely the position and attitude of the aerial robot and the rotation angle of each joint of the robotic arm. The first derivative of the pose information of the flying robot with respect to time t is the velocity of the flying robot (specifically including angular velocity and linear velocity). The second derivative of the pose information of the flying robot with respect to time t is the acceleration of the flying robot (specifically including angular acceleration and linear acceleration). The inertia matrix; For Coriolis force and centripetal force terms; This is the term related to gravity. The control torque of the aerial robot (including the control torque of the aerial robot during flight and the control torque of the robotic arm when deploying the crawling repair robot); This refers to the generalized torque caused by external wind field disturbances. The feedback control torque is the torque when the robotic arm deploys the crawling repair robot onto the wire to be repaired (i.e., when the crawling repair robot comes into contact with the wire to be repaired). It is affected by the flexibility and sag of the wire to be repaired, and is also related to the position of the aerial robot and the robotic arm.

[0045] Flow field constraints refer to the restrictions on the influence of external wind fields. Flow field constraints help prevent robots from flying under harsh wind conditions, thereby ensuring flight stability.

[0046] For example, the specific flow field constraints are: ; in, Let t be the first derivative of the overall position (i.e., the position of each part) of the flying robot with respect to time t, which is the overall velocity of the flying robot; Instantaneous wind speed; This represents the maximum speed of the flying robot.

[0047] Geometric constraints refer to restrictions on the geometric characteristics of the flight space, such as minimum safe distance and flight area boundaries. These constraints ensure that the robot will not collide with obstacles in the environment during flight.

[0048] For example, the geometric constraint body is: ; ; in, The overall position of the flying robot; This is static geometric map information; This refers to the location of each obstacle in the dynamic environment map information (including the location of the guide wire to be repaired). This is the preset minimum safe distance.

[0049] By introducing dynamic constraints, flow field constraints, and geometric constraints, the path planning stage fully considers the physical limitations of the wire repair robot itself and the dynamic characteristics of the external environment, generating multiple candidate path trajectory information that meet these conditions, and determining the path parameters corresponding to each candidate path trajectory information. The path parameters are specific values ​​that describe these path trajectory information, such as the coordinates, velocity, acceleration, attitude angle, and arrival time of each path point.

[0050] By inputting the path parameters corresponding to these constrained path trajectory information into the path planning comfort function for calculation, the quantitative comfort value of each path can be obtained.

[0051] Furthermore, by determining the minimum value among these comfort values ​​and selecting the corresponding path trajectory information as the final target planning path trajectory information, it is ensured that the selected path is the optimal path in terms of comfort while satisfying all constraints.

[0052] For example, an improved quantum particle swarm optimization (QPSO) algorithm that describes particles using a quantum potential well model can be used for path generation and optimal solution to generate multiple candidate path trajectory information and corresponding path parameters that satisfy the constraints (preset dynamic constraints, flow field constraints, and geometric constraints). The path planning comfort function can then be used to calculate and obtain the target planned path trajectory information.

[0053] The improved quantum particle swarm optimization algorithm, which describes particles using the quantum potential well model, is specifically as follows: ; ; ; in, The j-th dimension component of the best individual among the i-th individuals generated at time t+1 (i.e., the best path trajectory information among the path trajectory information generated at time t+1, which means that the path planning comfort function value corresponding to this path trajectory information is the smallest among the path trajectory information generated at this time); The acceleration constant is a random number between 0 and 1; Let j be the j-th dimension component of the best individual among the i-th individuals generated at time t; It is the j-th dimension component of the globally optimal individual (i.e., the optimal path trajectory information among all path trajectory information, and the path planning comfort function value corresponding to this path trajectory information is the smallest among all path trajectory information); This represents the probability distribution in the quantum potential well model. This represents a random number that is uniformly distributed between (0,1); is the j-th dimension component of the average individual among all optimal individuals; N is the total number of individuals generated at the corresponding time, i≤N; Let j be the j-th dimension component of the i-th individual generated at time t+1; The coefficient of contraction and expansion; Let j be the j-th dimension component of the i-th individual generated at time t; It is a random number.

[0054] Therefore, by iteratively calculating using the above algorithm, the target planning path trajectory information that satisfies the constraints can be obtained.

[0055] Specifically, in step S104, after obtaining the target planned path trajectory information, the flying robot in the wire repair robot is controlled to fly to the target hovering position corresponding to the wire to be repaired according to the trajectory information. Once the flying robot successfully hovers at the target position, the crawling repair robot in the wire repair robot is deployed onto the wire to be repaired. The flight control of the flying robot can employ methods such as PID control, LQR control, or model predictive control, adjusting its flight attitude and speed in real time based on the target planned path trajectory information. Furthermore, it can adjust the dynamic environment map information based on the real-time dynamic position of the wire to be repaired, thereby adjusting the target hovering position and the target planned path trajectory information in real time. The deployment process can utilize a gripping mechanism at the end of the robotic arm or an electromagnetic adsorption mechanism to ensure that the crawling repair robot can stably and safely attach to the wire to be repaired.

[0056] Specifically, in step S105, based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and real-time environmental wind speed information, the crawling repair robot is controlled to repair the wire to be repaired, and the repair result corresponding to the wire to be repaired is obtained, including: Acquire actuator control command signals, actual motion status data, and real-time environmental wind speed information of the crawling repair robot; Based on the actuator control command signal of the crawling repair robot, prediction is performed to obtain the target motion state data corresponding to the actuator control command signal; Based on the difference between the target motion state data and the actual motion state data, combined with real-time environmental wind speed information, a dynamic switching control method is used to control the crawling repair robot to repair the wire to be repaired, and obtain the corresponding repair result of the wire to be repaired.

[0057] In step S105, the actuator control command signals, actual motion state data, and real-time environmental wind speed information of the crawling repair robot are acquired. The actuator control command signals refer to the commands sent by the control system to each actuator (e.g., motors, servos) of the crawling repair robot, instructing them to perform specific actions or posture adjustments. The actual motion state data refers to the real-time motion state data of the crawling repair robot collected by sensors (e.g., inertial measurement units, encoders, etc.), including position, speed, and posture data. The real-time environmental wind speed information refers to the current wind speed and direction data of the work area obtained through devices such as wind speed sensors.

[0058] When the crawling repair robot walks on the conductor to be repaired, the interaction force between its wheels and the conductor can be simplified into a nonlinear friction and variable stiffness contact model. The nonlinear friction and variable stiffness contact model is as follows: ; in, The tangential force exerted by the wire to be repaired on the crawling repair robot; The normal force between the traveling wheel and the wire to be repaired; The friction coefficient is non-linear and depends on the speed of the crawling repair robot. With positive pressure The function; The equivalent lateral stiffness of the conductor to be repaired is the robot position. The function is affected by sag and tension, and is time-varying and uncertain.

[0059] Although the friction coefficient in the above model Equivalent lateral stiffness of conductor While these two key observable data points cannot be precisely obtained, the model framework can clearly define them, creating the preconditions for subsequent model-free control.

[0060] Based on the actuator control command signals of the crawling repair robot, prediction is performed to obtain the target motion state data corresponding to the actuator control command signals. For example, based on the dynamic model, kinematic model, and response characteristics of the crawling repair robot, the motion state that may occur at a certain time or over a period of time in the future for a given actuator control command signal can be simulated or calculated. This yields the expected motion trajectory and state of the crawling repair robot under this command, establishing an expected benchmark for subsequent comparison with the actual motion state. The target motion state data refers to the ideal motion state predicted based on the actuator control command signals.

[0061] For example, the preset model-free adaptive control algorithm of this application can be used to calculate and predict the target motion state data.

[0062] Model-free adaptive control algorithms do not require a specific system model. Instead, at each operating point, they equate complex nonlinear control to a linear data model with time-varying parameters. Specifically, the linear data model is as follows: ; in, The system output parameters at time k+1 are the motion state data, including the position, velocity, and attitude angle of the crawling repair robot on the guide wire to be repaired. The system output parameters at time k; for System output parameters at time; The system input parameters at time k are the control parameters corresponding to the actuator control command signal, including parameters such as the motor speed and control torque of the crawling repair robot. This is a function that maps system input parameters to system output parameters.

[0063] In model-free adaptive control algorithms, the compact scheme dynamic linearization (CFDL) method is adopted, that is, when At any given time, there must exist a time-varying parameter called the pseudo-partial derivative (PPD). This makes the following equation true: ; in, This represents the change in the system output parameters from time k+1 to time k, i.e. ; The pseudo-partial derivative time-varying parameter at time k contains all the dynamic, nonlinear and coupling information of the system at time k; The change in the system input parameters at time k is... Near every k-time, this calculation formula is precisely equivalent to the nonlinear control at every k-time.

[0064] Due to the time-varying parameters of pseudopartial derivatives Since the information is unknown, it needs to be estimated online first. Therefore, the pseudo-partial derivative estimation algorithm model is shown below: ; in, The time-varying pseudo-partial derivative parameter at time k is predicted based on the data at time k-1, i.e., the time-varying pseudo-partial derivative parameter at time k is an estimated value. The pseudo-partial derivative time-varying parameter at time k-1 is predicted based on the data at time k-2; This is a pseudo-partial derivative step size factor used to control the estimation speed; This is a pseudo-partial derivative weighting factor used to prevent the denominator from being zero, thus ensuring... There will not be any major changes; Let be the change in the system input parameters at time k-1. and It can be configured according to actual needs.

[0065] Design a small positive number (Specific values ​​can be set according to actual needs), so that and .when or At that time, it is necessary to estimate the value. Reset the algorithm to enhance its ability to track time-varying parameters.

[0066] Therefore, based on the above pseudo-partial derivative estimation algorithm model, by minimizing the one-step forward prediction error, a weighted one-step optimal control law is designed, yielding the control input criterion function, which is specifically as follows: ; in, To control the input criterion function value; The system output parameters at time k+1 are predicted, i.e., the desired output trajectory; To control the input weighting factor.

[0067] Substituting the pseudo-partial derivative estimation algorithm model into the control input criterion function mentioned above, and then... Taking the derivative, we obtain the analytical model of the control law, which is as follows: ; in, To control the input step size factor.

[0068] In summary, the model-free adaptive control algorithm, using only the current system output parameters and historical system input parameters (historical records of system input parameters), (first predicts the system output parameters (target motion state data) at the next moment) and then updates the pseudo-partial derivative time-varying parameters and calculates the current system input parameters. This enables the crawling repair robot to walk smoothly and track its trajectory with high precision under normal disturbances.

[0069] Optionally, in step S105, based on the difference between the target motion state data and the actual motion state data, and combined with real-time environmental wind speed information, a dynamic switching control method is used to control the crawling repair robot to repair the wire to be repaired, thereby obtaining the repair result corresponding to the wire to be repaired, including: Calculate the difference between the target motion state data and the actual motion state data; When the difference is less than the preset error threshold and the real-time wind speed value in the real-time environmental wind speed information is less than the preset disturbance wind speed threshold, the preset model-free adaptive control algorithm is used to control the crawling repair robot to repair the wire to be repaired. When the difference is greater than or equal to the preset error threshold, or when the real-time wind speed value in the real-time environmental wind speed information is greater than or equal to the preset disturbance wind speed threshold, the preset active disturbance rejection control algorithm is used to control the crawling repair robot to repair the wire to be repaired. When the sum of the number of times the model-free adaptive control algorithm and the active disturbance rejection control algorithm are used within a preset period is greater than or equal to a preset switching threshold, the algorithm usage weight coefficient is determined based on the historical usage of the model-free adaptive control algorithm and the active disturbance rejection control algorithm. Based on the algorithm usage weight coefficient, the model-free adaptive control algorithm and the active disturbance rejection control algorithm are combined to generate a compliant control quantity, so as to control the crawling repair robot to repair the wire to be repaired according to the compliant control quantity. After confirming that the wire to be repaired has been repaired, the repair result corresponding to the wire to be repaired is obtained.

[0070] Specifically, in step S105, during the process of the crawling repair robot repairing the wire, it is first necessary to calculate the difference between the target motion state data and the actual motion state data. This difference reflects the deviation between the current motion state and the desired motion state of the crawling repair robot, and is an important basis for evaluating the control effect and adjusting the control strategy.

[0071] When the calculated difference is less than the preset error threshold, and the real-time wind speed value in the real-time environmental wind speed information is less than the preset disturbance wind speed threshold, it indicates that the crawling repair robot is in a relatively stable and less disturbed working condition. In this case, a preset model-free adaptive control algorithm can be used to control the crawling repair robot. The model-free adaptive control algorithm does not require a precise system model and can adapt to dynamic changes in the system by learning and adjusting control parameters online. It is suitable for scenarios with high accuracy requirements and low external disturbances. The specific control process of the model-free adaptive control algorithm is shown above.

[0072] When the calculated difference is greater than or equal to a preset error threshold, or when the real-time wind speed value in the real-time environmental wind speed information is greater than or equal to a preset disturbance wind speed threshold, it indicates that the crawling repair robot may face significant motion deviations or substantial environmental disturbances (such as strong winds). In this case, to enhance the system's anti-disturbance capability and robustness, a preset active disturbance rejection control algorithm can be switched on. The active disturbance rejection control algorithm can treat internal and external disturbances as a "total disturbance" for real-time estimation and compensation, thereby effectively suppressing the impact of disturbances on control performance and ensuring that the crawling repair robot can still perform repair operations stably and efficiently in harsh environments.

[0073] The core idea of ​​the preset active disturbance rejection control algorithm is to combine the system's internal uncertainties (such as modeling errors and parameter perturbations) and external disturbances (such as strong gusts and sudden vibrations of conductors) into a "total disturbance". At the same time, it is estimated in real time through an extended state observer (ESO) and actively compensated in the control law.

[0074] For example, the control process of the active disturbance rejection control algorithm is as follows: Taking one degree of freedom of motion of the crawling repair robot (such as position x along the guide wire) as an example, its second-order dynamics can be designed as follows: ; in, Let x be the second derivative of position x with respect to time t, which is the acceleration of position x. Let x be the first derivative of position x with respect to time t, which is the velocity of position x. External time-varying disturbances (external interferences) that vary with time t; It is an approximate estimate of the control gain and is an adjustable parameter; The desired control quantity (i.e., the desired system input parameter); The total disturbance function of the active disturbance rejection control algorithm is a relational expression that integrates all unknown nonlinearities, coupling effects, internal uncertainties, and external disturbances.

[0075] Will The calculated value is used as a new state variable. (i.e., "expanded state"), constructing a continuously expanding state space, we obtain: ; in, The actual position of the crawling repair robot (i.e., the actual position); The first derivative of the position of the crawling repair robot with respect to time t; This represents the actual speed of the crawling repair robot. The first derivative of the crawling repair robot's velocity with respect to time t; The total disturbance for the crawling repair robot; Let be the first derivative of the total perturbation of the crawling repair robot with respect to time t. The total perturbation rate of change for the crawling repair robot; Among them, due to the total disturbance change rate For unknown but bounded systems, a third-order linear extended state observer (LESO) is designed to represent the estimated values ​​of the system output parameters. Specifically, the third-order linear extended state observer is as follows: ; in, The position estimation error is the difference between the estimated position and the actual position of the crawling repair robot. The position estimate for the crawling repair robot; Let be the first derivative of the position estimate of the crawling repair robot with respect to time t; The speed estimate for the crawling repair robot; The first derivative of the speed estimate of the crawling repair robot with respect to time t; This is the estimated total disturbance for the crawling repair robot; The first derivative of the total perturbation estimate of the crawling repair robot with respect to time t; This is the position error feedback coefficient; This is the speed error feedback coefficient; This represents the disturbance error feedback coefficient. , and Value and observer bandwidth Related, such as setting , , It can be adjusted according to actual needs.

[0076] The design goal of the Active Disturbance Rejection (ADRR) algorithm is to eliminate the total disturbance and make the system behave as an ideal integral-series structure. Therefore, by using the total disturbance estimated in real time by the ESO and compensating for it in the control input, we obtain: ; in, The actual system input parameters are the actual control parameters of the actuator control command signals when using the active disturbance rejection control algorithm, which are used to drive the crawling repair robot to reach the desired state.

[0077] For the position tracking problem, the simplest PD control combination can be used to obtain: ; in, To control the position error gain coefficient; The expected position of the crawling repair robot; To control the speed error gain coefficient; Let be the expected speed of the crawling repair robot. and With controller bandwidth Related, such as setting , It can be adjusted according to actual needs.

[0078] In each control cycle, the Active Disturbance Rejection (ESO) algorithm estimates the actual system input parameters and total disturbance in real time using the aforementioned model, based on the actuator control command signal (desired system input parameters) and actual motion state data (actual system output data). Then, the actual control quantity is calculated using the model. Because ESO can actively estimate and compensate for disturbances, the ESO algorithm exhibits strong robustness to model errors and external disturbances, making it suitable for maintaining system stability in scenarios such as strong winds or violently swaying wires.

[0079] In an optional embodiment, when the total disturbance is greater than a preset total disturbance control threshold, or the total disturbance change rate is greater than a preset disturbance change rate control threshold, an active disturbance rejection control algorithm can be used to control the crawling repair robot to repair the wire to be repaired. The preset total disturbance control threshold and the preset disturbance change rate control threshold can be set according to actual needs.

[0080] Furthermore, to avoid system instability issues that may arise from frequent switching of control algorithms, a compliant control mechanism is introduced. Specifically, when the sum of the number of times the model-free adaptive control algorithm and the active disturbance rejection control algorithm are used within a preset period is greater than or equal to a preset switching threshold (the preset switching threshold can be set according to actual needs), algorithm usage weight coefficients are determined based on the historical usage counts of the two algorithms. These weight coefficients reflect the applicability and contribution of the two algorithms under the current operating conditions. Subsequently, based on these algorithm usage weight coefficients, a compliant control quantity is generated by combining the model-free adaptive control algorithm and the active disturbance rejection control algorithm. This compliant control quantity is a weighted combination of the outputs of the two algorithms, which enables a smooth transition of the control strategy, avoids the shock caused by hard switching, and thus controls the crawling repair robot to repair the guide wire to be repaired according to the compliant control quantity.

[0081] The specific formula for calculating the compliance control amount is as follows: ; in, To control the amount of softness, Weighting coefficients are used in the model-free adaptive control algorithm; These are the system input parameters for the model-free adaptive control algorithm. These are the system input parameters for the active disturbance rejection control algorithm.

[0082] In step S105, after confirming that the wire to be repaired has been repaired, the repair result corresponding to the wire to be repaired is obtained.

[0083] As described above, the control method for this wire repair robot acquires multimodal perception data, including the robot's dynamics, environmental point cloud data, visual image data, and surrounding wind field information. Based on these data, a map is constructed to obtain a dynamic environmental map of the wire repair robot regarding the wire to be repaired. Following this dynamic map and dynamics data, path planning is performed on the robot to obtain its target trajectory from its current position to the wire to be repaired. Then, following this trajectory, the flying robot within the robot flies to the target hovering position corresponding to the wire to be repaired. Finally, the crawling repair robot is deployed onto the wire to be repaired, based on the actuator control command signals from the crawling repair robot and its actual motion state data. Based on real-time environmental wind speed information, the system controls a crawling repair robot to repair the conductor to be repaired, obtaining the corresponding repair result. Furthermore, by using a dynamic environmental map constructed based on environmental point cloud data, visual image data, and surrounding wind field information, path planning is performed. The resulting target planned path trajectory information is then used to control the flying robot within the conductor repair robot to fly to the target hovering position, after which the crawling repair robot is deployed onto the conductor to be repaired. This allows the crawling repair robot to repair the conductor, obtaining the corresponding repair result. This addresses the problem in existing conductor repair methods where neglecting dynamic environmental factors and sudden disturbances leads to inaccurate control of the conductor repair robot, affecting repair progress. It considers dynamic environmental factors and sudden disturbances, ensuring the feasibility of the planned path in actual dynamic environments, effectively coping with sudden disturbances such as strong high-altitude winds and conductor galloping, and improving the repair efficiency of the conductor repair robot.

[0084] refer to Figure 2 This application provides a control device for a wire repair robot, used to control the wire repair robot, including: The acquisition module 1 is used to acquire multimodal perception data of the wire repair robot; the multimodal perception data includes the dynamic data of the wire repair robot, environmental point cloud data, visual image data and surrounding wind field information; Module 2 is used to construct a map based on environmental point cloud data, visual image data and surrounding wind field information, so as to obtain dynamic environmental map information of the conductor repair robot about the conductor to be repaired. Planning module 3 is used to plan the path of the wire repair robot based on dynamic environment map information and dynamic data, and obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired. Control module 4 is used to control the flying robot in the wire repair robot to fly to the target hovering position corresponding to the wire to be repaired according to the target planned path trajectory information, and then deploy the crawling repair robot in the wire repair robot onto the wire to be repaired. Repair module 5 is used to control the crawling repair robot to repair the wire to be repaired based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and the real-time environmental wind speed information, so as to obtain the repair result corresponding to the wire to be repaired.

[0085] The control device of this conductor repair robot performs path planning based on a dynamic environmental map constructed from environmental point cloud data, visual image data, and surrounding wind field information. The resulting target planned path trajectory information is used to control the flying robot within the conductor repair robot to fly to the target hovering position. Then, the crawling repair robot is deployed onto the conductor to be repaired, allowing the crawling repair robot to perform the repair and obtain the corresponding repair result. This addresses the problem in existing conductor repair methods where neglecting dynamic environmental factors and sudden disturbances leads to inaccurate control of the conductor repair robot, thus affecting repair progress. It considers dynamic environmental factors and sudden disturbances, ensuring the feasibility of the planned path in actual dynamic environments. It effectively copes with sudden disturbances such as strong high-altitude winds and conductor galloping, improving the repair efficiency of the conductor repair robot.

[0086] Specifically, during execution, module 1 acquires multimodal perception data from the wire repair robot. This multimodal perception data includes the robot's dynamics data, environmental point cloud data, visual image data, and surrounding wind field information. The wire repair robot is an integrated system comprising a flying robot and a crawling repair robot. The flying robot transports the crawling repair robot to the vicinity of the wire to be repaired and deploys it onto the wire. The crawling repair robot then moves and performs repair work on the wire. Multimodal perception data refers to various types of data acquired from different sensors, including dynamics data, environmental point cloud data, visual image data, and surrounding wind field information. Dynamics data can include information such as the robot's attitude, velocity, and acceleration, and can be acquired through an inertial measurement unit (MIU). Environmental point cloud data describes the three-dimensional geometry of the surrounding environment and can be acquired through lidar. Visual image data provides two-dimensional image information of the environment and can be acquired through a camera. Surrounding wind field information describes meteorological conditions such as wind speed and direction within the robot's operating area and can be acquired through an ultrasonic anemometer mounted on the wire repair robot or through a third-party meteorological service platform.

[0087] Specifically, when module 2 constructs a map based on environmental point cloud data, visual image data, and surrounding wind field information to obtain dynamic environmental map information of the conductor repair robot regarding the conductor to be repaired, it executes the following: Based on the environmental image information in the environmental point cloud data and visual image data, a map is constructed to obtain the static geometric map information of the wire repair robot; Based on static geometric map information, and by integrating surrounding wind field information, conductor image information from visual image data, and conductor feature point cloud data from environmental point cloud data, dynamic environmental map information about the conductor to be repaired is obtained for the conductor repair robot.

[0088] When Module 2 is executed, static geometric map information refers to the fixed and unchanging geometric structure representation of the environment constructed by processing and analyzing environmental image information from environmental point cloud data and visual image data. This static geometric map information is mainly used to describe the spatial distribution and geometric characteristics of static environmental elements such as terrain, buildings, fixed obstacles, and (in a relatively static state) guide wires to be repaired, providing a stable and fundamental environmental reference for the guide wire repair robot. The static geometric map information can be represented in the form of a raster map, point cloud map, or 3D model.

[0089] By overlaying real-time dynamic information onto a static geometric map, a comprehensive dynamic environment map is formed, providing the wire guide repair robot with a dynamic environment map of the wire to be repaired. Specifically, surrounding wind field information is used to capture airflow disturbances in the environment, while wire guide image information and wire guide feature point cloud data focus on the precise location, orientation, and potential minor deformations of the wire to be repaired. This fusion of multi-source information ensures that the constructed dynamic environment map not only includes the static geometric structure of the environment but also reflects real-time dynamic changes within the wire guide repair robot's operating area, such as the impact of wind on the wire and the wire's real-time state. The wire guide image information and wire guide feature point cloud data can be obtained through existing technologies, such as the YOLOv8 semantic segmentation network, by performing wire recognition on the visual image data and environmental point cloud data, respectively.

[0090] The dynamic environment map information can be calculated using the following formula: ; in, For dynamic environment map information; The instantaneous wind speed at location x at time t can be calculated using the extended Kalman filter algorithm based on the surrounding wind field information. The pose information of the location point x of the conductor to be repaired at time t.

[0091] Specifically, when planning module 3 performs path planning for the wire repair robot based on dynamic environment map information and combined with dynamic data, and obtains the target planned path trajectory information of the wire repair robot flying from its current position to the wire to be repaired, it executes: Obtain the pre-built route planning comfort function; Based on the path planning comfort function, dynamic environment map information and dynamic data are used to perform path planning, and obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired.

[0092] During the execution of planning module 3, the path planning comfort function is a mathematical model that comprehensively evaluates path quality, aiming to quantify the path's performance across multiple dimensions such as safety, efficiency, energy consumption, and stability. This function is typically pre-designed and constructed based on the characteristics of the wire repair robot, the environmental features of the wire to be repaired, and the task requirements. For example, the path planning comfort function may include multiple sub-items such as flight time and energy consumption, which are then weighted and summed or combined in other ways to form a single comfort index. Pre-construction means that the function has been defined, parameterized, and stored in the control system or related computing unit of the wire repair robot before the path planning process begins, so that it can be directly invoked during path planning.

[0093] The path planning comfort function is specifically as follows: ; in, This represents the comfort value for route planning. For flight time; Energy consumption; The safety penalty coefficient is inversely proportional to the distance from the wire repair robot to the obstacle (including the wire to be repaired); This is a weighted value based on flight time. This is the weighted value for energy consumption; This is the weight value for security penalties.

[0094] Optionally, when planning module 3 performs path planning based on the path planning comfort function, using dynamic environment map information and dynamic data, and obtains the target planned path trajectory information for the wire repair robot to fly from its current position to the wire to be repaired, it executes: According to the preset hovering criteria, the target hovering position corresponding to the guide wire to be repaired is determined in the dynamic environment map information; Based on the target hovering position and combined with dynamic environment map information, the path planning comfort function is used to plan the path, and obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired.

[0095] Specifically, during execution, before path planning, planning module 3 first determines the "target hovering position" that the flying robot needs to reach in the dynamic environment map information based on preset hovering criteria. This target hovering position is determined according to preset hovering criteria combined with dynamic environment map information. For example, an optimal hovering point can be determined based on the geometry of the guide wire, the characteristics of the repair area, and the characteristics of the flying robot's delivery mechanism, making the delivery process as smooth and safe as possible. The determination of this target hovering position provides a clear and optimized endpoint for subsequent path planning.

[0096] The preset hovering criteria consist of a series of rules or conditions used to evaluate and select the optimal hovering point. These criteria may include, but are not limited to: the relative distance and angle to the guide wire to be repaired, the safe distance from surrounding obstacles, the attitude and dynamic limitations of the flying robot itself, and the feasibility of deploying the crawling repair robot. The preset hovering criteria ensure that the flying robot can hover stably in a position that is both safe and conducive to subsequent repair work.

[0097] Optionally, when planning module 3 obtains the target planned path trajectory information of the wire repair robot flying from its current position to the wire to be repaired based on the target hovering position, combined with dynamic environment map information, and using the path planning comfort function, it executes: Based on preset dynamic constraints, flow field constraints, and geometric constraints, combined with dynamic environment map information and target hovering position, path planning is performed to obtain the path trajectory information and path parameters corresponding to the path trajectory information of the wire repair robot flying from the current position to the target hovering position. The path parameters corresponding to each path trajectory information are input into the path planning comfort function for calculation, and the path planning comfort value corresponding to each path trajectory information is obtained. Determine the minimum value among the comfort values ​​of each path planning, and use the path trajectory information corresponding to the minimum value as the target planned path trajectory information for the wire repair robot to fly from the current position to the wire to be repaired.

[0098] Specifically, when planning module 3 is executed, the dynamic constraints refer to the limitations on the flight performance of the wire repair robot. For example, the aerial robot (generally a multi-rotor flight platform) and the robotic arm (generally a multi-joint robotic arm) of the flying robot in the wire repair robot are regarded as a whole to limit the motion data of the flying robot, thereby constructing a dynamic equation representing the overall flying robot. The dynamic equation of the flying robot is as follows: ; Where q represents the pose information of the flying robot, namely the position and attitude of the aerial robot and the rotation angle of each joint of the robotic arm. The first derivative of the pose information of the flying robot with respect to time t is the velocity of the flying robot (specifically including angular velocity and linear velocity). The second derivative of the pose information of the flying robot with respect to time t is the acceleration of the flying robot (specifically including angular acceleration and linear acceleration). The inertia matrix; For Coriolis force and centripetal force terms; This is the term related to gravity. The control torque of the aerial robot (including the control torque of the aerial robot during flight and the control torque of the robotic arm when deploying the crawling repair robot); This refers to the generalized torque caused by external wind field disturbances. The feedback control torque is the torque when the robotic arm deploys the crawling repair robot onto the wire to be repaired (i.e., when the crawling repair robot comes into contact with the wire to be repaired). It is affected by the flexibility and sag of the wire to be repaired, and is also related to the position of the aerial robot and the robotic arm.

[0099] Flow field constraints refer to the restrictions on the influence of external wind fields. Flow field constraints help prevent robots from flying under harsh wind conditions, thereby ensuring flight stability.

[0100] For example, the specific flow field constraints are: ; in, Let t be the first derivative of the overall position (i.e., the position of each part) of the flying robot with respect to time t, which is the overall velocity of the flying robot; Instantaneous wind speed; This represents the maximum speed of the flying robot.

[0101] Geometric constraints refer to restrictions on the geometric characteristics of the flight space, such as minimum safe distance and flight area boundaries. These constraints ensure that the robot will not collide with obstacles in the environment during flight.

[0102] For example, the geometric constraint body is: ; ; in, The overall position of the flying robot; This is static geometric map information; This refers to the location of each obstacle in the dynamic environment map information (including the location of the guide wire to be repaired). This is the preset minimum safe distance.

[0103] By introducing dynamic constraints, flow field constraints, and geometric constraints, the path planning stage fully considers the physical limitations of the wire repair robot itself and the dynamic characteristics of the external environment, generating multiple candidate path trajectory information that meet these conditions, and determining the path parameters corresponding to each candidate path trajectory information. The path parameters are specific values ​​that describe these path trajectory information, such as the coordinates, velocity, acceleration, attitude angle, and arrival time of each path point.

[0104] By inputting the path parameters corresponding to these constrained path trajectory information into the path planning comfort function for calculation, the quantitative comfort value of each path can be obtained.

[0105] Furthermore, by determining the minimum value among these comfort values ​​and selecting the corresponding path trajectory information as the final target planning path trajectory information, it is ensured that the selected path is the optimal path in terms of comfort while satisfying all constraints.

[0106] For example, an improved quantum particle swarm optimization (QPSO) algorithm that describes particles using a quantum potential well model can be used for path generation and optimal solution to generate multiple candidate path trajectory information and corresponding path parameters that satisfy the constraints (preset dynamic constraints, flow field constraints, and geometric constraints). The path planning comfort function can then be used to calculate and obtain the target planned path trajectory information.

[0107] The improved quantum particle swarm optimization algorithm, which describes particles using the quantum potential well model, is specifically as follows: ; ; ; in, The j-th dimension component of the best individual among the i-th individuals generated at time t+1 (i.e., the best path trajectory information among the path trajectory information generated at time t+1, which means that the path planning comfort function value corresponding to this path trajectory information is the smallest among the path trajectory information generated at this time); The acceleration constant is a random number between 0 and 1; Let j be the j-th dimension component of the best individual among the i-th individuals generated at time t; It is the j-th dimension component of the globally optimal individual (i.e., the optimal path trajectory information among all path trajectory information, and the path planning comfort function value corresponding to this path trajectory information is the smallest among all path trajectory information); This represents the probability distribution in the quantum potential well model. This represents a random number that is uniformly distributed between (0,1); is the j-th dimension component of the average individual among all optimal individuals; N is the total number of individuals generated at the corresponding time, i≤N; Let j be the j-th dimension component of the i-th individual generated at time t+1; The coefficient of contraction and expansion; Let j be the j-th dimension component of the i-th individual generated at time t; It is a random number.

[0108] Therefore, by iteratively calculating using the above algorithm, the target planning path trajectory information that satisfies the constraints can be obtained.

[0109] Specifically, during execution, after obtaining the target planned path trajectory information, control module 4 controls the flying robot in the wire repair robot to fly to the target hovering position corresponding to the wire to be repaired according to the trajectory information. Once the flying robot successfully hovers at the target position, the crawling repair robot in the wire repair robot is deployed onto the wire to be repaired. The flight control of the flying robot can employ methods such as PID control, LQR control, or model predictive control, adjusting its flight attitude and speed in real time based on the target planned path trajectory information. Furthermore, it can adjust the dynamic environment map information based on the real-time dynamic position of the wire to be repaired, thereby adjusting the target hovering position and the target planned path trajectory information in real time. The deployment process can utilize a gripping mechanism at the end of the robotic arm or an electromagnetic adsorption mechanism to ensure that the crawling repair robot can stably and safely attach to the wire to be repaired.

[0110] Optionally, when the repair module 5 controls the crawling repair robot to repair the wire to be repaired based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and the real-time environmental wind speed information, and obtains the repair result corresponding to the wire to be repaired, it executes: Acquire actuator control command signals, actual motion status data, and real-time environmental wind speed information of the crawling repair robot; Based on the actuator control command signal of the crawling repair robot, prediction is performed to obtain the target motion state data corresponding to the actuator control command signal; Based on the difference between the target motion state data and the actual motion state data, combined with real-time environmental wind speed information, a dynamic switching control method is used to control the crawling repair robot to repair the wire to be repaired, and obtain the corresponding repair result of the wire to be repaired.

[0111] During execution, repair module 5 acquires actuator control command signals, actual motion state data, and real-time environmental wind speed information from the crawling repair robot. The actuator control command signals are commands sent by the control system to each actuator (e.g., motors, servos) of the crawling repair robot, instructing them to perform specific actions or posture adjustments. The actual motion state data refers to the real-time motion state data of the crawling repair robot collected by sensors (e.g., inertial measurement units, encoders, etc.), including position, speed, and posture data. The real-time environmental wind speed information refers to the current wind speed and direction data of the work area obtained through devices such as wind speed sensors.

[0112] When the crawling repair robot walks on the conductor to be repaired, the interaction force between its wheels and the conductor can be simplified into a nonlinear friction and variable stiffness contact model. The nonlinear friction and variable stiffness contact model is as follows: ; in, The tangential force exerted by the wire to be repaired on the crawling repair robot; The normal force between the traveling wheel and the wire to be repaired; The friction coefficient is non-linear and depends on the speed of the crawling repair robot. With positive pressure The function; The equivalent lateral stiffness of the conductor to be repaired is the robot position. The function is affected by sag and tension, and is time-varying and uncertain.

[0113] Although the friction coefficient in the above model Equivalent lateral stiffness of conductor While these two key observable data points cannot be precisely obtained, the model framework can clearly define them, creating the preconditions for subsequent model-free control.

[0114] Based on the actuator control command signals of the crawling repair robot, prediction is performed to obtain the target motion state data corresponding to the actuator control command signals. For example, based on the dynamic model, kinematic model, and response characteristics of the crawling repair robot, the motion state that may occur at a certain time or over a period of time in the future for a given actuator control command signal can be simulated or calculated. This yields the expected motion trajectory and state of the crawling repair robot under this command, establishing an expected benchmark for subsequent comparison with the actual motion state. The target motion state data refers to the ideal motion state predicted based on the actuator control command signals.

[0115] For example, the preset model-free adaptive control algorithm of this application can be used to calculate and predict the target motion state data.

[0116] Model-free adaptive control algorithms do not require a specific system model. Instead, at each operating point, they equate complex nonlinear control to a linear data model with time-varying parameters. Specifically, the linear data model is as follows: ; in, The system output parameters at time k+1 are the motion state data, including the position, velocity, and attitude angle of the crawling repair robot on the guide wire to be repaired. The system output parameters at time k; for System output parameters at time; The system input parameters at time k are the control parameters corresponding to the actuator control command signal, including parameters such as the motor speed and control torque of the crawling repair robot. This is a function that maps system input parameters to system output parameters.

[0117] In model-free adaptive control algorithms, the compact scheme dynamic linearization (CFDL) method is adopted, that is, when At any given time, there must exist a time-varying parameter called the pseudo-partial derivative (PPD). This makes the following equation true: ; in, This represents the change in the system output parameters from time k+1 to time k, i.e. ; The pseudo-partial derivative time-varying parameter at time k contains all the dynamic, nonlinear and coupling information of the system at time k; The change in the system input parameters at time k is... Near every k-time, this calculation formula is precisely equivalent to the nonlinear control at every k-time.

[0118] Due to the time-varying parameters of pseudopartial derivatives Since the information is unknown, it needs to be estimated online first. Therefore, the pseudo-partial derivative estimation algorithm model is shown below: ; in, The time-varying pseudo-partial derivative parameter at time k is predicted based on the data at time k-1, i.e., the time-varying pseudo-partial derivative parameter at time k is an estimated value. The pseudo-partial derivative time-varying parameter at time k-1 is predicted based on the data at time k-2; This is a pseudo-partial derivative step size factor used to control the estimation speed; This is a pseudo-partial derivative weighting factor used to prevent the denominator from being zero, thus ensuring... There will not be any major changes; Let be the change in the system input parameters at time k-1. and It can be configured according to actual needs.

[0119] Design a small positive number (Specific values ​​can be set according to actual needs), so that and .when or At that time, it is necessary to estimate the value. Reset the algorithm to enhance its ability to track time-varying parameters.

[0120] Therefore, based on the above pseudo-partial derivative estimation algorithm model, by minimizing the one-step forward prediction error, a weighted one-step optimal control law is designed, yielding the control input criterion function, which is specifically as follows: ; in, To control the input criterion function value; The system output parameters at time k+1 are predicted, i.e., the desired output trajectory; To control the input weighting factor.

[0121] Substituting the pseudo-partial derivative estimation algorithm model into the control input criterion function mentioned above, and then... Taking the derivative, we obtain the analytical model of the control law, which is as follows: ; in, To control the input step size factor.

[0122] In summary, the model-free adaptive control algorithm, using only the current system output parameters and historical system input parameters (historical records of system input parameters), (first predicts the system output parameters (target motion state data) at the next moment) and then updates the pseudo-partial derivative time-varying parameters and calculates the current system input parameters. This enables the crawling repair robot to walk smoothly and track its trajectory with high precision under normal disturbances.

[0123] Specifically, when repair module 5 controls the crawling repair robot to repair the guide wire based on the difference between the target motion state data and the actual motion state data, combined with real-time environmental wind speed information, and using a dynamic switching control method, and obtains the repair result for the guide wire, it executes: Calculate the difference between the target motion state data and the actual motion state data; When the difference is less than the preset error threshold and the real-time wind speed value in the real-time environmental wind speed information is less than the preset disturbance wind speed threshold, the preset model-free adaptive control algorithm is used to control the crawling repair robot to repair the wire to be repaired. When the difference is greater than or equal to the preset error threshold, or when the real-time wind speed value in the real-time environmental wind speed information is greater than or equal to the preset disturbance wind speed threshold, the preset active disturbance rejection control algorithm is used to control the crawling repair robot to repair the wire to be repaired. When the sum of the number of times the model-free adaptive control algorithm and the active disturbance rejection control algorithm are used within a preset period is greater than or equal to a preset switching threshold, the algorithm usage weight coefficient is determined based on the historical usage of the model-free adaptive control algorithm and the active disturbance rejection control algorithm. Based on the algorithm usage weight coefficient, the model-free adaptive control algorithm and the active disturbance rejection control algorithm are combined to generate a compliant control quantity, so as to control the crawling repair robot to repair the wire to be repaired according to the compliant control quantity. After confirming that the wire to be repaired has been repaired, the repair result corresponding to the wire to be repaired is obtained.

[0124] Specifically, when repair module 5 is executed, during the process of the crawling repair robot repairing the wire, it first needs to calculate the difference between the target motion state data and the actual motion state data. This difference reflects the deviation between the current motion state and the desired motion state of the crawling repair robot, and is an important basis for evaluating the control effect and adjusting the control strategy.

[0125] When the calculated difference is less than the preset error threshold, and the real-time wind speed value in the real-time environmental wind speed information is less than the preset disturbance wind speed threshold, it indicates that the crawling repair robot is in a relatively stable and less disturbed working condition. In this case, a preset model-free adaptive control algorithm can be used to control the crawling repair robot. The model-free adaptive control algorithm does not require a precise system model and can adapt to dynamic changes in the system by learning and adjusting control parameters online. It is suitable for scenarios with high accuracy requirements and low external disturbances. The specific control process of the model-free adaptive control algorithm is shown above.

[0126] When the calculated difference is greater than or equal to a preset error threshold, or when the real-time wind speed value in the real-time environmental wind speed information is greater than or equal to a preset disturbance wind speed threshold, it indicates that the crawling repair robot may face significant motion deviations or substantial environmental disturbances (such as strong winds). In this case, to enhance the system's anti-disturbance capability and robustness, a preset active disturbance rejection control algorithm can be switched on. The active disturbance rejection control algorithm can treat internal and external disturbances as a "total disturbance" for real-time estimation and compensation, thereby effectively suppressing the impact of disturbances on control performance and ensuring that the crawling repair robot can still perform repair operations stably and efficiently in harsh environments.

[0127] The core idea of ​​the preset active disturbance rejection control algorithm is to combine the system's internal uncertainties (such as modeling errors and parameter perturbations) and external disturbances (such as strong gusts and sudden vibrations of conductors) into a "total disturbance". At the same time, it is estimated in real time through an extended state observer (ESO) and actively compensated in the control law.

[0128] For example, the control process of the active disturbance rejection control algorithm is as follows: Taking one degree of freedom of motion of the crawling repair robot (such as position x along the guide wire) as an example, its second-order dynamics can be designed as follows: ; in, Let x be the second derivative of position x with respect to time t, which is the acceleration of position x. Let x be the first derivative of position x with respect to time t, which is the velocity of position x. External time-varying disturbances (external interferences) that vary with time t; It is an approximate estimate of the control gain and is an adjustable parameter; The desired control quantity (i.e., the desired system input parameter); The total disturbance function of the active disturbance rejection control algorithm is a relational expression that integrates all unknown nonlinearities, coupling effects, internal uncertainties, and external disturbances.

[0129] Will The calculated value is used as a new state variable. (i.e., "expanded state"), constructing a continuously expanding state space, we obtain: ; in, The actual position of the crawling repair robot (i.e., the actual position); The first derivative of the position of the crawling repair robot with respect to time t; This represents the actual speed of the crawling repair robot. The first derivative of the crawling repair robot's velocity with respect to time t; The total disturbance for the crawling repair robot; Let be the first derivative of the total perturbation of the crawling repair robot with respect to time t. The total perturbation rate of change for the crawling repair robot; Among them, due to the total disturbance change rate For unknown but bounded systems, a third-order linear extended state observer (LESO) is designed to represent the estimated values ​​of the system output parameters. Specifically, the third-order linear extended state observer is as follows: ; in, The position estimation error is the difference between the estimated position and the actual position of the crawling repair robot. The position estimate for the crawling repair robot; Let be the first derivative of the position estimate of the crawling repair robot with respect to time t; The speed estimate for the crawling repair robot; The first derivative of the speed estimate of the crawling repair robot with respect to time t; This is the estimated total disturbance for the crawling repair robot; The first derivative of the total perturbation estimate of the crawling repair robot with respect to time t; This is the position error feedback coefficient; This is the speed error feedback coefficient; represents the disturbance error feedback coefficient. Wherein, , and Value and observer bandwidth Related, such as setting , , It can be adjusted according to actual needs.

[0130] The design goal of the Active Disturbance Rejection (ADRR) algorithm is to eliminate the total disturbance and make the system behave as an ideal integral-series structure. Therefore, by using the total disturbance estimated in real time by the ESO and compensating for it in the control input, we obtain: ; in, The actual system input parameters are the actual control parameters of the actuator control command signals when using the active disturbance rejection control algorithm, which are used to drive the crawling repair robot to reach the desired state.

[0131] For the position tracking problem, the simplest PD control combination can be used to obtain: ; in, To control the position error gain coefficient; The expected position of the crawling repair robot; To control the speed error gain coefficient; Let be the expected speed of the crawling repair robot. and With controller bandwidth Related, such as setting , It can be adjusted according to actual needs.

[0132] In each control cycle, the Active Disturbance Rejection (ESO) algorithm estimates the actual system input parameters and total disturbance in real time using the aforementioned model, based on the actuator control command signal (desired system input parameters) and actual motion state data (actual system output data). Then, the actual control quantity is calculated using the model. Because ESO can actively estimate and compensate for disturbances, the ESO algorithm exhibits strong robustness to model errors and external disturbances, making it suitable for maintaining system stability in scenarios such as strong winds or violently swaying wires.

[0133] In an optional embodiment, when the total disturbance is greater than a preset total disturbance control threshold, or the total disturbance change rate is greater than a preset disturbance change rate control threshold, an active disturbance rejection control algorithm can be used to control the crawling repair robot to repair the wire to be repaired. The preset total disturbance control threshold and the preset disturbance change rate control threshold can be set according to actual needs.

[0134] Furthermore, to avoid system instability issues that may arise from frequent switching of control algorithms, a compliant control mechanism is introduced. Specifically, when the sum of the number of times the model-free adaptive control algorithm and the active disturbance rejection control algorithm are used within a preset period is greater than or equal to a preset switching threshold (the preset switching threshold can be set according to actual needs), algorithm usage weight coefficients are determined based on the historical usage counts of the two algorithms. These weight coefficients reflect the applicability and contribution of the two algorithms under the current operating conditions. Subsequently, based on these algorithm usage weight coefficients, a compliant control quantity is generated by combining the model-free adaptive control algorithm and the active disturbance rejection control algorithm. This compliant control quantity is a weighted combination of the outputs of the two algorithms, which enables a smooth transition of the control strategy, avoids the shock caused by hard switching, and thus controls the crawling repair robot to repair the guide wire to be repaired according to the compliant control quantity.

[0135] The specific formula for calculating the compliance control amount is as follows: ; in, To control the amount of softness, Weighting coefficients are used in the model-free adaptive control algorithm. These are the system input parameters for the model-free adaptive control algorithm. These are the system input parameters for the active disturbance rejection control algorithm.

[0136] When the repair module 5 is executed, it obtains the repair result corresponding to the wire to be repaired after confirming that the repair has been completed.

[0137] As described above, the control device of this wire repair robot acquires multimodal perception data of the robot, including dynamic data, environmental point cloud data, visual image data, and surrounding wind field information. Based on the environmental point cloud data, visual image data, and surrounding wind field information, a map is constructed to obtain dynamic environmental map information of the wire repair robot regarding the wire to be repaired. According to the dynamic environmental map information and combined with the dynamic data, path planning is performed on the wire repair robot to obtain the target planned path trajectory information of the robot flying from its current position to the wire to be repaired. Following the target planned path trajectory information, the flying robot within the wire repair robot flies to the target hovering position corresponding to the wire to be repaired, and then the crawling repair robot is deployed onto the wire to be repaired. Based on the actuator control command signals of the crawling repair robot, combined with the actual motion state data of the crawling repair robot... Based on real-time environmental wind speed information, the system controls a crawling repair robot to repair the conductor to be repaired, obtaining the corresponding repair result. Furthermore, by using a dynamic environmental map constructed based on environmental point cloud data, visual image data, and surrounding wind field information, path planning is performed. The resulting target planned path trajectory information is then used to control the flying robot within the conductor repair robot to fly to the target hovering position, after which the crawling repair robot is deployed onto the conductor to be repaired. This allows the crawling repair robot to repair the conductor, obtaining the corresponding repair result. This addresses the problem in existing conductor repair methods where neglecting dynamic environmental factors and sudden disturbances leads to inaccurate control of the conductor repair robot, affecting repair progress. It considers dynamic environmental factors and sudden disturbances, ensuring the feasibility of the planned path in actual dynamic environments, effectively coping with sudden disturbances such as strong high-altitude winds and conductor galloping, and improving the repair efficiency of the conductor repair robot.

[0138] Please refer to Figure 3 , Figure 3This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device includes a processor 301 and a memory 302. The processor 301 and the memory 302 are interconnected and communicate with each other via a communication bus 303 and / or other forms of connection mechanisms (not shown). The memory 302 stores a computer program executable by the processor 301. When the electronic device is running, the processor 301 executes the computer program to perform the control method for the wire repair robot in any optional implementation of the above embodiments, to achieve the following functions: acquiring multimodal perception data of the wire repair robot, including dynamic data of the wire repair robot, environmental point cloud data, visual image data, and surrounding wind field information; and based on the environmental point cloud data and visual image data... By combining the surrounding wind field information, a map is constructed to obtain the dynamic environment map information of the conductor repair robot regarding the conductor to be repaired. Based on the dynamic environment map information and combined with dynamic data, path planning is performed on the conductor repair robot to obtain the target planned path trajectory information of the conductor repair robot flying from the current position to the conductor to be repaired. According to the target planned path trajectory information, the flying robot in the conductor repair robot is controlled to fly to the target hovering position corresponding to the conductor to be repaired. Then, the crawling repair robot in the conductor repair robot is deployed onto the conductor to be repaired. Based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and the real-time environmental wind speed information, the crawling repair robot is controlled to repair the conductor to be repaired, and the corresponding repair result of the conductor to be repaired is obtained.

[0139] This application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it executes the control method for a wire repair robot in any optional implementation of the above embodiments to achieve the following functions: acquiring multimodal perception data of the wire repair robot, including dynamic data of the wire repair robot, environmental point cloud data, visual image data, and surrounding wind field information; constructing a map based on the environmental point cloud data, visual image data, and surrounding wind field information to obtain dynamic environmental map information of the wire repair robot regarding the wire to be repaired; performing path planning for the wire repair robot based on the dynamic environmental map information and combined with the dynamic data to obtain the target planned path trajectory information of the wire repair robot flying from its current position to the wire to be repaired; controlling the flying robot in the wire repair robot to fly to the target hovering position corresponding to the wire to be repaired according to the target planned path trajectory information; deploying the crawling repair robot in the wire repair robot onto the wire to be repaired; and controlling the crawling repair robot to repair the wire to be repaired based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and real-time environmental wind speed information to obtain the repair result corresponding to the wire to be repaired. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0140] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0141] Furthermore, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0142] Furthermore, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0143] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.

[0144] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A control method for a wire repair robot, characterized in that, include: Acquire multimodal perception data from the wire repair robot; The multimodal perception data includes the dynamics data of the wire repair robot, environmental point cloud data, visual image data, and surrounding wind field information; Based on the environmental point cloud data, the visual image data, and the surrounding wind field information, a map is constructed to obtain the dynamic environmental map information of the conductor repair robot regarding the conductor to be repaired. Based on the dynamic environment map information and the dynamic data, the path planning of the wire repair robot is performed to obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired; According to the target planned path trajectory information, after controlling the flying robot in the wire repair robot to fly to the target hovering position corresponding to the wire to be repaired, the crawling repair robot in the wire repair robot is deployed onto the wire to be repaired; Based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and real-time environmental wind speed information, the crawling repair robot is controlled to repair the wire to be repaired, and the repair result corresponding to the wire to be repaired is obtained.

2. The control method for the wire repair robot according to claim 1, characterized in that, Based on the environmental point cloud data, the visual image data, and the surrounding wind field information, a map is constructed to obtain dynamic environmental map information of the conductor repair robot regarding the conductor to be repaired, including: Based on the environmental point cloud data and the environmental image information in the visual image data, a map is constructed to obtain the static geometric map information of the wire repair robot; Based on the static geometric map information, the surrounding wind field information, the conductor image information in the visual image data, and the conductor feature point cloud data in the environmental point cloud data are fused to obtain the dynamic environmental map information of the conductor repair robot regarding the conductor to be repaired.

3. The control method for the wire repair robot according to claim 1, characterized in that, Based on the dynamic environment map information and the dynamic data, path planning is performed on the wire repair robot to obtain the target planned path trajectory information of the wire repair robot flying from its current position to the wire to be repaired, including: Obtain the pre-built route planning comfort function; Based on the path planning comfort function, the dynamic environment map information and the dynamic data are used to perform path planning, and the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired is obtained.

4. The control method for the wire repair robot according to claim 3, characterized in that, Based on the path planning comfort function, using the dynamic environment map information and the dynamic data, path planning is performed to obtain the target planned path trajectory information of the conductor repair robot flying from its current position to the conductor to be repaired, including: According to the preset hovering criteria, the target hovering position corresponding to the conductor to be repaired is determined in the dynamic environment map information; Based on the target hovering position and combined with the dynamic environment map information, the path planning comfort function is used to perform path planning, thereby obtaining the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired.

5. The control method for the wire repair robot according to claim 4, characterized in that, Based on the target hovering position and the dynamic environment map information, the path planning comfort function is used to perform path planning, resulting in the target planned path trajectory information of the wire repair robot flying from its current position to the wire to be repaired, including: Based on preset dynamic constraints, flow field constraints, and geometric constraints, and combined with the dynamic environment map information and the target hovering position, path planning is performed to obtain the path trajectory information of the wire repair robot flying from the current position to the target hovering position and the path parameters corresponding to the path trajectory information. The path parameters corresponding to each of the path trajectory information are input into the path planning comfort function for calculation to obtain the path planning comfort value corresponding to each of the path trajectory information; The minimum value among the path planning comfort values ​​is determined, and the path trajectory information corresponding to the minimum value is determined as the target planned path trajectory information for the conductor repair robot to fly from the current position to the conductor to be repaired.

6. The control method for the wire repair robot according to claim 1, characterized in that, Based on the actuator control command signals of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and real-time environmental wind speed information, the crawling repair robot is controlled to repair the conductor to be repaired, and the repair result corresponding to the conductor to be repaired is obtained, including: Acquire the actuator control command signals, actual motion state data, and real-time environmental wind speed information of the crawling repair robot; Based on the actuator control command signal of the crawling repair robot, prediction is performed to obtain the target motion state data corresponding to the actuator control command signal; Based on the difference between the target motion state data and the actual motion state data, and combined with the real-time environmental wind speed information, the crawling repair robot is controlled to repair the conductor to be repaired using a dynamic switching control method, thereby obtaining the repair result corresponding to the conductor to be repaired.

7. The control method for the wire repair robot according to claim 6, characterized in that, Based on the difference between the target motion state data and the actual motion state data, and combined with the real-time environmental wind speed information, a dynamic switching control method is used to control the crawling repair robot to repair the conductor to be repaired, obtaining the repair result corresponding to the conductor to be repaired, including: Calculate the difference between the target motion state data and the actual motion state data; When the difference is less than a preset error threshold and the real-time wind speed value in the real-time environmental wind speed information is less than a preset disturbance wind speed threshold, a preset model-free adaptive control algorithm is used to control the crawling repair robot to repair the conductor to be repaired. When the difference is greater than or equal to a preset error threshold, or when the real-time wind speed value in the real-time environmental wind speed information is greater than or equal to a preset disturbance wind speed threshold, a preset active disturbance rejection control algorithm is used to control the crawling repair robot to repair the conductor to be repaired. When the sum of the number of times the model-free adaptive control algorithm and the active disturbance rejection control algorithm are used within a preset period is greater than or equal to a preset switching threshold, the algorithm usage weight coefficient is determined based on the historical usage number of the model-free adaptive control algorithm and the active disturbance rejection control algorithm. Based on the algorithm usage weight coefficient, and in combination with the model-free adaptive control algorithm and the active disturbance rejection control algorithm, a compliant control quantity is generated to control the crawling repair robot to repair the wire to be repaired according to the compliant control quantity. After confirming that the wire to be repaired has been repaired, the repair result corresponding to the wire to be repaired is obtained.

8. A control device for a wire repair robot, characterized in that, include: The acquisition module is used to acquire multimodal perception data from the wire repair robot. The multimodal perception data includes the dynamics data of the wire repair robot, environmental point cloud data, visual image data, and surrounding wind field information; The construction module is used to construct a map based on the environmental point cloud data, the visual image data and the surrounding wind field information, so as to obtain the dynamic environmental map information of the conductor repair robot about the conductor to be repaired. The planning module is used to perform path planning for the wire repair robot based on the dynamic environment map information and the dynamic data, so as to obtain the target planned path trajectory information of the wire repair robot flying from the current position to the wire to be repaired; The control module is used to control the flying robot in the wire repair robot to fly to the target hovering position corresponding to the wire to be repaired according to the target planned path trajectory information, and then deploy the crawling repair robot in the wire repair robot onto the wire to be repaired. The repair module is used to control the crawling repair robot to repair the wire to be repaired based on the actuator control command signal of the crawling repair robot, combined with the actual motion state data of the crawling repair robot and the real-time environmental wind speed information, so as to obtain the repair result corresponding to the wire to be repaired.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing a computer program executable by the processor, wherein when the processor executes the computer program, it performs the steps of the control method for the wire repair robot as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it performs the steps in the control method of the wire repair robot as described in any one of claims 1-7.

Images