Inductive assembly collaborative control method and system for W-pins of equipotential insulator strings

Abstract

The present invention discloses a method and system for collaborative control of the non-inductive assembly of W-pins in equipotential insulator strings, relating to the field of intelligent control technology. The method comprises: a straddle-type intelligent work vehicle equipped with a three-dimensional laser radar and a visual camera scans the insulator strings of a transmission line, constructs a three-dimensional surface mesh model and spatial posture information of the W-pin slots, identifies the insertion path, and determines a set of candidate paths; utilizes a wind disturbance sensing module to identify interference in real time, screens the insertion path, controls a robotic arm and a thin electric gripper to perform non-inductive assembly, and dynamically adjusts the insertion status using a force sensor. This method solves the technical problem in the prior art that traditional W-pin assembly methods are unable to cope with dynamic environmental changes, resulting in poor assembly success rate and stability. This method achieves the technical effect of dynamically optimizing the assembly path through real-time wind disturbance sensing and an intelligent path feedback adjustment mechanism, thereby improving the success rate and stability of non-inductive W-pin assembly in complex environments.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to a method and system for cooperatively controlling the non-inductive assembly of W pins in an equipotential insulator string. Background Art

[0002] In the maintenance and assembly of high-voltage transmission lines, traditional manual or semi-automated equipment often faces problems such as low efficiency, poor accuracy, and insufficient adaptability to external environmental changes. This is especially true in complex environmental conditions, where factors such as wind and temperature often lead to assembly failures. Existing technologies primarily rely on manual or fixed path planning methods, lacking real-time perception and intelligent adjustment of the operating environment. Although some automated equipment has been used in transmission line maintenance, it is often unable to cope with dynamic environmental changes, making it difficult to ensure operational stability and accuracy. Summary of the Invention

[0003] The present application provides a method and system for cooperative control of the non-inductive assembly of W pins of an equipotential insulator string, which is used to solve the technical problem that the traditional W pin assembly method in the prior art cannot cope with dynamic environmental changes, resulting in poor assembly success rate and stability.

[0004] The first aspect of the present application provides a method for cooperative control of the non-sensing assembly of W pins of equipotential insulator strings, the method comprising: using a straddle-type intelligent operation vehicle equipped with a three-dimensional laser radar and a visual camera to perform a spatial structure scan of the insulator string of the target transmission line, constructing a three-dimensional surface mesh model of the insulator string and spatial posture information of the W pin slot; combining the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W pin slot to perform insertion path trajectory recognition and determine a set of candidate W pin insertion paths; calling a wind disturbance perception module deployed in a robotic arm controller to perceive and identify the assembly operation area and determine real-time a W-pin insertion interference vector; based on the real-time W-pin insertion interference vector, the set of W-pin insertion candidate paths is screened to obtain a screened W-pin insertion path; the robotic arm controller receives the screened W-pin insertion path, controls the robotic arm and the thin electric claw to perform W-pin sensorless assembly according to the screened W-pin insertion path, and uses the force sensor arranged at the end of the thin electric claw to capture the insertion state perception to determine whether the W-pin is assembled. If the judgment fails, the screened W-pin insertion path is feedback-adjusted according to the insertion state perception result, and the feedback-adjusted W-pin insertion path is obtained to perform W-pin sensorless assembly again.

[0005] The second aspect of the present application provides a non-sensing assembly collaborative control system for W-pins of equipotential insulator strings, the system comprising: a spatial structure scanning module, the spatial structure scanning module is used to use a straddle-type intelligent working vehicle equipped with a three-dimensional laser radar and a visual camera to perform spatial structure scanning on the insulator string of the target transmission line, and construct a three-dimensional surface mesh model of the insulator string and spatial posture information of the W-pin slot; a path trajectory recognition module, the path trajectory recognition module is used to combine the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot to perform insertion path trajectory recognition and determine a set of candidate W-pin insertion paths; an environmental perception recognition module, the environmental perception recognition module is used to call the wind disturbance perception module deployed in the robotic arm controller to perform assembly The business area is perceived and identified to determine the real-time W-pin insertion interference vector; a path screening module is used to screen the W-pin insertion candidate path set based on the real-time W-pin insertion interference vector to obtain a screened W-pin insertion path; a senseless assembly module is used to receive the screened W-pin insertion path through the robotic arm controller, control the robotic arm and the thin electric claw to perform W-pin senseless assembly according to the screened W-pin insertion path, and use the force sensor arranged at the end of the thin electric claw to capture the insertion state perception to determine whether the W-pin is assembled. If the judgment fails, the screened W-pin insertion path is feedback-adjusted according to the insertion state perception result, and the feedback-adjusted W-pin insertion path is obtained to perform W-pin senseless assembly again.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The present application provides a method and system for collaborative control of the inductive assembly of W-pins in equipotential insulator strings, which relate to the field of intelligent control technology. By utilizing a three-dimensional laser radar, a visual camera, and a wind disturbance perception module, the insulator strings of the target transmission line are automatically identified and a three-dimensional model is constructed. By combining insertion path recognition and wind disturbance perception, the insertion path is intelligently screened, and the inductive assembly of W-pins is achieved through a robotic arm and a thin electric claw. This solves the technical problem in the prior art that the traditional W-pin assembly method cannot cope with dynamic environmental changes, resulting in poor assembly success rate and stability. The system achieves the technical effect of dynamically optimizing the assembly path through real-time wind disturbance perception and an intelligent path feedback adjustment mechanism, thereby improving the success rate and stability of W-pin inductive assembly in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1 A schematic flow chart of a method for cooperatively controlling the non-inductive assembly of W pins of an equipotential insulator string according to an embodiment of the present application;

[0010] Figure 2 Schematic diagram of the structure of the non-inductive assembly collaborative control system for the W pins of the equipotential insulator string provided in an embodiment of the present application.

[0011] Explanation of the accompanying reference numerals: spatial structure scanning module 11, path trajectory recognition module 12, environment perception recognition module 13, path screening module 14, sensorless assembly module 15. DETAILED DESCRIPTION

[0012] The present application provides a method and system for cooperative control of the non-inductive assembly of W pins of an equipotential insulator string, which is used to solve the technical problem that the traditional W pin assembly method in the prior art cannot cope with dynamic environmental changes, resulting in poor assembly success rate and stability.

[0013] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only some of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0014] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices.

[0015] Example 1, as Figure 1 As shown, the present application provides a method for cooperative control of non-inductive assembly of W pins of an equipotential insulator string, the method comprising:

[0016] P10: Using a straddle-type intelligent work vehicle equipped with a 3D laser radar and a visual camera, the insulator string of the target transmission line is spatially scanned to construct a 3D surface mesh model of the insulator string and the spatial posture information of the W-pin slot.

[0017] Furthermore, step P10 in the embodiment of the present application further includes:

[0018] P11: Activate the IMU module of the straddle-type intelligent working vehicle for spatial alignment and complete the coordinate system initialization of the three-dimensional laser radar and the visual camera; P12: Start the three-dimensional laser radar to perform high-density point cloud acquisition on the insulator string to obtain point cloud data; P13: Synchronously start the visual camera to collect image information of the insulator string to obtain an insulator string image; P14: Fusion and recognition are performed on the point cloud data and the insulator string image to determine the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot.

[0019] It should be understood that the insulator string of the target transmission line is accurately scanned in spatial structure by a straddle-type intelligent operation vehicle equipped with advanced sensors, thereby constructing a three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot.

[0020] First, the straddle-type intelligent work vehicle's IMU (Inertial Measurement Unit) module is activated for spatial registration. The IMU module is a sensor component that measures and reports information such as the vehicle's acceleration and angular velocity in three-dimensional space. By sensing acceleration and angular velocity, the IMU module ensures that the LiDAR and camera measurement data are integrated within the same coordinate system. Its purpose is to achieve high-precision spatial registration, ensuring that subsequent sensor data is processed within a unified reference frame, avoiding precision errors caused by coordinate offsets and ensuring operational accuracy.

[0021] Next, a 3D LiDAR system was activated to collect a high-density point cloud of the insulator string. During this process, the LiDAR measures distance by emitting a laser beam and receiving the reflected signal, thereby obtaining spatial position data on the surface of the insulator string. The density and quality of the point cloud data directly impact the accuracy of the model. Higher-density point clouds more accurately reflect surface details. By collecting a large number of point clouds, the 3D LiDAR system can provide the overall structure and geometry of the insulator string, laying the foundation for subsequent 3D modeling. This high-density point cloud data not only captures the surface contours of the insulator string but also provides precise information support for subsequent path planning and assembly tasks.

[0022] At the same time, the visual camera and 3D LiDAR are activated synchronously to begin collecting image information of the insulator strings. The visual camera captures visual features such as the color and texture of the insulator strings through high-resolution images. These images contain these visual features. The combination of image information and point cloud data can overcome the shortcomings of single data types. For example, while point cloud data can provide precise geometric information, it may not be able to distinguish different materials or textures in some cases. Image information can provide this supplementary information.

[0023] After acquiring the point cloud data and image data, data fusion and recognition are performed. The point cloud data provides accurate 3D geometric information for the insulator string, while the image information complements the detailed surface features of the object. During the data fusion process, the fusion algorithm combines the geometric information in the point cloud data with the visual features in the image information according to preset rules to determine the 3D surface mesh model of the insulator string and the spatial pose information of the W-pin slot. The 3D surface mesh model of the insulator string is a digital representation of the surface geometry of the insulator string, accurately describing its shape and dimensions. The spatial pose information of the W-pin slot includes key parameters such as the slot's position and orientation, which can be used for subsequent insertion path planning and assembly operations.

[0024] By fusing the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot, accurate positioning of the target object can be achieved, providing basic data support for the robot arm's insertion path planning, interference avoidance and task execution.

[0025] Furthermore, step P14 of the embodiment of the present application also includes:

[0026] P14-1: Spatially align the point cloud data and the insulator string image, and reconstruct the three-dimensional surface mesh model using the ICP iterative nearest point algorithm and the sparse grid fitting algorithm based on the edge contour of the insulator string image and the density change of the point cloud data to obtain the three-dimensional surface mesh model of the insulator string; P14-2: Perform image segmentation based on the insulator string image to determine the W-pin slot spatial area; P14-3: Identify the slot posture parameters of the point cloud data based on the W-pin slot spatial area, and use the acquired slot center point position, insertion direction vector and slot depth as the W-pin slot spatial posture information.

[0027] Optionally, continue to perform deep data fusion and spatial analysis to further refine and optimize the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot.

[0028] First, the collected point cloud data and the insulator string image are spatially aligned. The goal of spatial alignment is to ensure that the image data and point cloud data precisely correspond in the same coordinate system, thereby achieving accurate 3D modeling. This process utilizes the ICP (Iterative Closest Point) algorithm, which iteratively calculates the minimum distance between the two point sets and continuously optimizes the registration results, ensuring precise alignment of the point cloud data with the image's edge contours and feature points. The ICP algorithm effectively handles data noise and gradually optimizes the registration results, ultimately achieving precise alignment of the image and point cloud data. Next, a sparse mesh fitting algorithm is used to reconstruct a 3D surface mesh model, leveraging density variations in the point cloud data and image edge contour information. This fitting algorithm interpolates points based on the distribution characteristics of the point cloud data to better fit the 3D surface morphology of the insulator string. This step results in a more accurate 3D surface mesh model of the insulator string. This model not only contains geometric information but also accurately describes surface details, providing precise data support for subsequent operations.

[0029] Next, the insulator string image is segmented to further determine the spatial region of the W-pin slot. Image segmentation technology clearly distinguishes the slot area within the insulator string by identifying and separating different regions within the image. This process uses an algorithm to extract specific areas within the image and filter out irrelevant background information, thereby accurately identifying the W-pin slot location. Image segmentation not only helps identify the slot location but also provides a basis for calculating the slot's spatial posture based on its shape and edge features.

[0030] Finally, based on the previously determined spatial region of the W-pin slot, the point cloud data is used to identify the slot's pose parameters. This process begins with an algorithm analyzing the corresponding slot region in the point cloud data to identify key pose parameters such as the slot's center point position, insertion direction vector, and slot depth. These parameters constitute the W-pin slot's spatial pose information, which is crucial for subsequent robotic arm control, insertion path planning, and interference detection. Accurately identifying the slot's spatial pose ensures that the W-pin assembly process is free from external interference during subsequent operations, while ensuring precise control of the insertion position and angle.

[0031] P20: Combining the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot, perform insertion path trajectory recognition to determine a set of candidate W-pin insertion paths.

[0032] Furthermore, step P20 in this embodiment of the present application further includes:

[0033] P21: Generate multiple candidate solutions for the insertion path based on the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot; P22: Obtain the basic information of the W-pin, and perform collision identification with the three-dimensional surface mesh model of the insulator string based on the multiple candidate solutions for the insertion path. If no collision exists, add them to the set of candidate W-pin insertion paths.

[0034] Specifically, accurate insertion path trajectory recognition is performed through the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot, thereby determining a set of candidate paths for W-pin insertion.

[0035] First, based on the three-dimensional surface mesh model of the insulator string and the spatial orientation information of the W-pin slot, multiple candidate assembly path solutions are generated. To further improve the accuracy and efficiency of path identification, a neural network model can be used to automatically generate these candidate paths. By learning from a large amount of historical data, the neural network can identify underlying patterns in the assembly paths, thereby providing multiple candidate path solutions for the assembly operation. This process, leveraging the deep learning capabilities of the neural network, can rapidly generate multiple paths for selection and automatically adjust the path generation strategy based on different job scenarios and requirements to accommodate diverse assembly task requirements.

[0036] Next, after obtaining the basic information of the W-pin, collision identification is performed. Specifically, collision detection is performed on the three-dimensional surface mesh model for the multiple previously generated candidate solutions for the insertion path. Collision identification is the process of using an algorithm to detect whether the insertion path collides with the surface of the insulator string in three-dimensional space. If a candidate solution for the insertion path collides with the surface of the insulator string during the simulated insertion process, the path will be excluded; conversely, if there is no collision, the path will be added to the set of candidate paths for W-pin insertion. The purpose of this step is to filter out unsuitable paths and ensure that the path finally selected can complete the insertion task safely and effectively.

[0037] Through the above steps, a feasible candidate path set can be effectively screened out from numerous possible insertion paths, providing a basis for selection for subsequent path optimization and actual insertion operations.

[0038] Furthermore, to generate multiple candidate solutions for the insertion path, step P21 of the embodiment of the present application further includes:

[0039] P21-1: Construct an insertion path trajectory identifier using a sample W-pin slot spatial posture information set as the insertion target, a sample insulator string three-dimensional surface mesh model set as the interference factor, and a plurality of sample insertion path candidate solution sets as the output; P21-2: Based on the insertion path trajectory identifier, the insulator string three-dimensional surface mesh model and the W-pin slot spatial posture information are identified to determine the plurality of insertion path candidate solutions.

[0040] Optionally, an instrumentation path trajectory identifier can be constructed to generate multiple candidate instrumentation path solutions, providing a basis for subsequent path selection and optimization.

[0041] To generate multiple candidate insertion path solutions, an insertion path trajectory identifier is first constructed, using a set of sample W-pin slot spatial pose information as the insertion target, a set of sample insulator string 3D surface mesh models as interference factors, and multiple sets of sample insertion path candidate solutions as outputs. Specifically, the set of sample W-pin slot spatial pose information contains multiple possible W-pin slot spatial pose information, which serves as the target state for insertion; the set of sample insulator string 3D surface mesh models contains multiple insulator string 3D surface mesh models, which serve as interference factors that may be encountered during the insertion process; and the multiple sets of sample insertion path candidate solutions are multiple possible insertion paths pre-generated based on these targets and interference factors. Using this sample data and combined with a neural network model for training, the constructed insertion path trajectory identifier can learn to generate appropriate insertion paths under different target and interference conditions.

[0042] Next, the constructed insertion path trajectory identifier is used to identify the actual insulator string 3D surface mesh model and W-pin slot spatial pose information. Using its internal learning algorithm and model structure, the identifier analyzes the input 3D surface mesh model and W-pin slot spatial pose information to determine multiple candidate insertion path solutions. These candidate solutions, generated based on the actual insulator string model and W-pin slot pose information, are adaptable to different insertion scenarios and conditions, providing diverse options for subsequent collision identification and path screening.

[0043] Through the above steps, this embodiment can efficiently generate multiple candidate insertion path solutions, providing reliable data support for subsequent path selection and optimization. The advantage of this process is that it uses machine learning models to automate path identification, allowing for flexible path adjustment in complex environments, ensuring the efficiency and accuracy of the entire W-pin sensorless assembly process.

[0044] P30: Call the wind disturbance perception module deployed in the robotic arm controller to perceive and identify the assembly operation area and determine the real-time W-pin insertion interference vector.

[0045] Furthermore, step P30 in the embodiment of the present application further includes:

[0046] P31: Use the wind disturbance perception module to continuously perceive the wind disturbance in the assembly operation area to obtain a wind force-wind direction interference perception vector sequence; P32: Extract the first wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector from the wind force-wind direction interference perception vector sequence; P33: Perform adjacency matrix analysis on the first wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector to determine the first adjacency matrix; P34: Extract the third wind force-wind direction interference perception vector again, and perform adjacency matrix analysis on the second wind force-wind direction interference perception vector, and so on to obtain an adjacency matrix set; P35: Calculate the mean of the adjacency matrix, enhance the last wind force-wind direction interference perception vector in the wind force-wind direction interference perception vector sequence, and obtain a real-time W-pin insertion interference vector.

[0047] It should be understood that by using the wind disturbance perception module to perform continuous wind disturbance perception analysis on the assembly operation area, the real-time W-pin insertion interference vector is ultimately determined, thereby helping the robot arm to perform precise insertion operations.

[0048] First, the wind disturbance perception module continuously monitors the wind speed and direction in the assembly area to obtain a sequence of wind force-direction interference perception vectors. This vector sequence records the dynamic changes in wind speed and direction in the assembly area, providing real-time insights into the impact of wind disturbances on the assembly process. Wind speed and direction are key factors affecting assembly operations. Therefore, continuously monitoring these interference factors throughout the assembly process can provide a basis for subsequent path adjustments and task optimization.

[0049] Next, the first and second wind force-direction interference perception vectors are extracted from the wind force-direction interference perception vector sequence. These vectors represent the wind disturbance state at different points in time and serve as the basis for subsequent analysis. By extracting these two vectors, we can identify the changing trends of wind force and direction, providing the necessary time series data for further calculations.

[0050] Subsequently, adjacency matrix analysis was performed on the extracted first and second wind force-direction interference perception vectors. The adjacency matrix is ​​a mathematical tool used to describe the relationship between data. By calculating the relationship between these two vectors, the first adjacency matrix was determined. This matrix reflects the relationship between the first and second wind force-direction interference perception vectors, providing a spatial and temporal understanding of the impact of wind disturbances and an effective reference for subsequent adjustments to the interference vectors.

[0051] Next, the third wind force-direction interference perception vector is extracted and subjected to adjacency matrix analysis with the second wind force-direction interference perception vector. This process is repeated to gradually obtain a set of adjacency matrices. This method allows for systematic analysis of the correlation between each vector in the wind disturbance perception vector sequence, thereby capturing the changing patterns of wind disturbance.

[0052] Finally, the mean of the adjacency matrix set is calculated, and the last wind force-direction interference perception vector in the wind force-direction interference perception vector sequence is enhanced to obtain a real-time W-pin insertion interference vector. This enhancement is achieved through convolution, a mathematical method that extracts signal features and enhances signal strength. In this way, the information in the wind disturbance perception vector sequence is comprehensively processed to obtain a real-time W-pin insertion interference vector that reflects the current wind disturbance status. This vector serves as an important basis for subsequent insertion path adjustments.

[0053] By implementing the above steps, changes in wind speed and direction can be continuously monitored. Adjacency matrix analysis and vector enhancement are then used to accurately calculate the real-time W-pin insertion interference vector. This process not only improves the adaptability of the insertion path but also ensures the safety and efficiency of assembly operations in complex environments.

[0054] Furthermore, step P33 of the embodiment of the present application also includes:

[0055] P33-1: Perform element mapping similarity calculation on the first wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector respectively to determine the element mapping similarity set; P33-2: Obtain an initial adjacency matrix that is initially empty, perform mean calculation on the element mapping similarity set, and add the calculation result to the initial adjacency matrix to determine the first adjacency matrix.

[0056] Specifically, the adjacency matrix analysis process can be further optimized by accurately calculating the similarity between wind force-direction interference perception vectors and integrating the results into the adjacency matrix through mean calculation, thereby enhancing the accuracy of association analysis of wind disturbance perception vectors and path optimization.

[0057] First, element-mapping similarity is calculated for the first wind force-direction interference perception vector and the second wind force-direction interference perception vector. This calculation is performed by comparing the similarities between corresponding elements in the two vectors. An appropriate similarity calculation method, such as Euclidean distance or cosine similarity, can be selected to obtain a set of multiple similarity values, namely the element-mapping similarity set. These similarity values ​​reflect the degree of similarity between the two wind force-direction interference perception vectors in various dimensions, providing basic data for the subsequent construction of the adjacency matrix.

[0058] Next, an initial adjacency matrix that is initially empty is obtained, and the mean of the element mapping similarity set is calculated. By averaging all the similarity values ​​in the set, a mean value is obtained that can comprehensively reflect the overall similarity of the two wind force-wind direction interference perception vectors. Then, this calculated mean value is added to the initial adjacency matrix to determine the first adjacency matrix. The adjacency matrix is ​​a matrix form used to describe the association relationship between data. Here, the first adjacency matrix can clearly reflect the association relationship between the first wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector. Through this matrix form description, complex vector similarity information can be simplified and structured, which is convenient for subsequent analysis and processing.

[0059] The combination of these two steps enables accurate real-time assessment of wind force and direction changes in wind-disturbed environments, providing detailed and precise interference data for subsequent real-time calculation of the W-pin insertion interference vector. This not only improves the safety and stability of the insertion path but also provides the necessary basis for path optimization during the robot arm's insertion task, ensuring precise and efficient operation.

[0060] P40: Filter the W-pin insertion candidate path set based on the real-time W-pin insertion interference vector to obtain a filtered W-pin insertion path.

[0061] Optionally, the W-pin insertion candidate path set is screened based on the real-time W-pin insertion interference vector to obtain the optimal insertion path, thereby ensuring the efficiency and accuracy of the insertion operation.

[0062] First, the real-time W-pin insertion interference vector, determined in the previous step, is obtained. This interference vector, derived through continuous monitoring and analysis by the wind disturbance perception module, accurately reflects the wind speed and direction changes within the current assembly area. This real-time W-pin insertion interference vector is crucial for subsequent path selection, as it directly correlates to the dynamic interference that may be encountered during the assembly process.

[0063] On this basis, the real-time W-pin insertion interference vector is compared and analyzed with each path in the set of W-pin insertion candidate paths. Specifically, for each candidate path, its insertion effect under the influence of the current interference vector is calculated. This analysis can be accomplished by simulating the insertion process, taking into account the impact of the interference vector on the insertion path, including factors such as path deviation and changes in insertion accuracy.

[0064] During the analysis, each candidate path is evaluated for feasibility and safety under interference conditions. If a path can still ensure accurate insertion of the W pin into the slot and the insertion process is safe and reliable despite the influence of real-time interference vectors, then the path is considered feasible. Conversely, if the path poses a risk of insertion failure under interference, such as excessive path deviation or collision with the insulator string, then the path is excluded.

[0065] Finally, a set of candidate W-pin insertion paths is selected from the set of candidate paths that remain feasible under real-time interference conditions, forming the selected W-pin insertion path. This selected path set serves as the basis for subsequent robotic arm insertion operations, ensuring that the robotic arm follows the optimized path during actual assembly, thereby improving assembly success rate and efficiency. This screening process is crucial for robotic arm operation, ensuring that insertion operations continue as expected in changing environments, thereby improving efficiency and reducing the risk of errors.

[0066] P50: The robotic arm controller receives the screened W-pin insertion path, controls the robotic arm and the thin electric claw to perform W-pin inductive assembly according to the screened W-pin insertion path, and uses the force sensor arranged at the end of the thin electric claw to capture the insertion status perception to determine whether the W-pin is assembled. If the judgment fails, the screened W-pin insertion path is feedback-adjusted according to the insertion status perception result, and the W-pin inductive assembly is performed again after obtaining the feedback-adjusted W-pin insertion path.

[0067] Furthermore, step P50 in the embodiment of the present application further includes:

[0068] P50a: If the determination is successful, the assembled W-pin is captured using a visual camera and the image is archived.

[0069] Specifically, the robotic arm controller receives a filtered W-pin insertion path and, based on this path, controls the robotic arm and thin electric gripper to perform sensorless W-pin assembly along a preset trajectory. During the assembly process, a force sensor positioned at the end of the thin electric gripper captures real-time information about the insertion status. The force sensor monitors key parameters during the insertion process, such as insertion force, displacement, and friction. These parameters are comprehensively analyzed using a state-based judgment model and compared with a preset insertion depth value to determine whether the W-pin is fully assembled.

[0070] If the force sensor feedback indicates that the W-pin assembly is successful, a visual camera captures an image of the assembled W-pin to obtain a complete assembly image. This image is then archived for subsequent documentation and tracing, ensuring the verifiability and traceability of the assembly process.

[0071] If the force sensor determines that the W-pin assembly has failed, feedback is then applied to the selected W-pin insertion path based on the insertion status sensing results. This adjustment is optimized based on the cause of the failure and may include fine-tuning parameters such as the insertion path direction, angle, or speed. The adjusted path is then re-executed to ensure the W-pin is successfully inserted into the target position and the insertion task is completed. This feedback mechanism enables real-time response to various situations that may arise during the operation, ensuring the accuracy and stability of the insertion process.

[0072] This series of steps enables high-precision W-pin assembly in complex environments, ensuring both success and reliability through real-time monitoring and feedback adjustments. Furthermore, vision cameras are used to confirm and archive the assembly process, providing data support for subsequent quality inspections and improvements.

[0073] Furthermore, if the determination fails, feedback adjustment is performed on the selected W-pin insertion path according to the insertion state perception result. In the embodiment of the present application, step P50 further includes:

[0074] P51: Extract the force-displacement-friction three-dimensional interlocking parameters from the insertion state perception results; P52: Perform deviation identification based on the force-displacement-friction three-dimensional interlocking parameters and the standard force-displacement-friction three-dimensional interlocking parameters, and determine the force-displacement-friction three-dimensional interlocking deviation parameters; P53: Perform feedback adjustment on the screened W-pin insertion path according to the force-displacement-friction three-dimensional interlocking deviation parameters to obtain the feedback adjusted W-pin insertion path.

[0075] In a possible embodiment of the present application, it is possible to further refine how to perform feedback adjustment on the screening W-pin insertion path according to the insertion state perception result when the insertion fails, so as to ensure that the W-pin can be successfully assembled.

[0076] When the insertion state sensing results indicate that the W-pin assembly was unsuccessful, the force-displacement-friction three-dimensional interlocking parameters are first extracted from the insertion state sensing results. These parameters are obtained in real time through force sensor monitoring and reflect the interaction force, displacement changes, and friction between the W-pin and the socket during the insertion process. The force-displacement-friction three-dimensional interlocking parameters are key indicators for determining the insertion state and can comprehensively reflect the mechanical characteristics of the insertion process.

[0077] Next, based on the extracted force-displacement-friction three-dimensional interlocking parameters, deviations are identified from the standard force-displacement-friction three-dimensional interlocking parameters. The standard force-displacement-friction three-dimensional interlocking parameters are ideal reference values, which represent the standard performance of the mechanical behavior between the W pin and the slot under normal insertion conditions. By comparing with the standard parameters, the system can identify deviations between force, displacement and friction. For example, if the increase in force during the insertion process does not match the change in displacement, or the friction force exceeds the expected range, the force-displacement-friction three-dimensional interlocking deviation parameters can be determined by comparing the difference between the actual perceived parameters and the standard parameters. These deviation data help to determine the possible causes of errors in the insertion process and then make path adjustments.

[0078] Finally, based on the identified deviation parameters for the three-dimensional interlocking force, displacement, and friction, feedback adjustments are made to the selected W-pin insertion path. Based on the magnitude and type of deviation, the insertion path parameters are optimized, such as fine-tuning the insertion angle, speed, and force level, to ensure path accuracy. In this way, the system can compensate for deviations caused by interference and equipment errors during the insertion process, ensuring a successful next insertion.

[0079] Ultimately, feedback adjustment yields a feedback-adjusted W-pin insertion path. This path, optimized based on the insertion status sensing results, effectively avoids previous insertion failures and ensures the W-pin is successfully inserted into the target position. This adjustment mechanism enables the robot arm to precisely adapt to complex and dynamic operating environments, completing high-precision insertion tasks and enhancing the stability and reliability of the insertion operation.

[0080] In summary, the embodiments of the present application have at least the following technical effects:

[0081] This application ensures high-precision identification and planning of the W-pin insertion path through precise data collection by three-dimensional laser radar and visual cameras, reducing the impact of external interference on assembly accuracy. The wind disturbance perception module is used to monitor environmental changes in real time and dynamically adjust the insertion path to ensure that the assembly task can still be carried out stably under different wind speeds and wind directions. At the same time, through intelligent path screening and feedback adjustment mechanism, the best selection of the insertion path is ensured to avoid insertion failures due to interference or inappropriate paths. Achieve the senseless assembly of W pins, reduce manual operation and intervention, improve work efficiency and safety, avoid personnel risks in high-altitude operations, and ensure the safety of operators.

[0082] The technical effect of dynamically optimizing the assembly path and improving the success rate and stability of W-pin sensorless assembly in complex environments has been achieved through real-time wind disturbance perception and intelligent path feedback adjustment mechanism.

[0083] Embodiment 2 is based on the same inventive concept as the inductive assembly coordinated control method of the W pin of the equipotential insulator string in the above embodiment. Figure 2 As shown, the present application provides a non-inductive assembly collaborative control system for the W pins of an equipotential insulator string. The system and method embodiments in the present application are based on the same inventive concept. The system includes:

[0084] The spatial structure scanning module 11 is used to use a straddle-type intelligent working vehicle equipped with a three-dimensional laser radar and a visual camera to perform spatial structure scanning on the insulator string of the target transmission line, and construct a three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot.

[0085] The path trajectory recognition module 12 is used to perform insertion path trajectory recognition based on the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot, and determine a set of candidate W-pin insertion paths.

[0086] The environment perception and recognition module 13 is used to call the wind disturbance perception module deployed in the robot arm controller, perceive and identify the assembly operation area, and determine the real-time W-pin insertion interference vector.

[0087] The path screening module 14 is configured to screen the W-pin insertion candidate path set based on the real-time W-pin insertion interference vector to obtain a screened W-pin insertion path.

[0088] The sensorless assembly module 15 is used to receive the screened W-pin insertion path through the robotic arm controller, control the robotic arm and the thin electric claw to perform W-pin sensorless assembly according to the screened W-pin insertion path, and use the force sensor arranged at the end of the thin electric claw to capture the insertion state perception to determine whether the W-pin is assembled. If the judgment fails, the screened W-pin insertion path is feedback-adjusted according to the insertion state perception result, and the W-pin sensorless assembly is performed again after obtaining the feedback-adjusted W-pin insertion path.

[0089] Furthermore, the spatial structure scanning module 11 is further configured to perform the following steps:

[0090] The IMU module of the straddle-type intelligent working vehicle is activated for spatial alignment to complete the coordinate system initialization of the three-dimensional laser radar and the visual camera; the three-dimensional laser radar is started to perform high-density point cloud acquisition on the insulator string to obtain point cloud data; the visual camera is simultaneously started to collect image information of the insulator string to obtain an insulator string image; the point cloud data and the insulator string image are fused and recognized to determine the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot.

[0091] Furthermore, the spatial structure scanning module 11 is further configured to perform the following steps:

[0092] The point cloud data and the insulator string image are spatially aligned, and based on the edge contour of the insulator string image and the density change of the point cloud data, an ICP iterative closest point algorithm and a sparse grid fitting algorithm are used to reconstruct a three-dimensional surface mesh model to obtain a three-dimensional surface mesh model of the insulator string; image segmentation is performed based on the insulator string image to determine the W-pin slot spatial area; slot posture parameter recognition is performed on the point cloud data based on the W-pin slot spatial area, and the obtained slot center point position, insertion direction vector and notch depth are used as the W-pin slot spatial posture information.

[0093] Furthermore, the path trajectory recognition module 12 is further configured to perform the following steps:

[0094] Based on the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot, multiple candidate insertion path solutions are generated; basic information of the W-pin is obtained, and collision identification is performed based on the multiple candidate insertion path solutions and the three-dimensional surface mesh model of the insulator string. If no collision exists, the solution is added to the W-pin insertion candidate path set.

[0095] Furthermore, the path trajectory recognition module 12 is further configured to perform the following steps:

[0096] An insertion path trajectory identifier is constructed with a set of sample W-pin slot spatial posture information as the insertion target, a set of sample insulator string three-dimensional surface mesh models as the interference factor, and a set of multiple sample insertion path candidate solutions as the output. Based on the insertion path trajectory identifier, the three-dimensional surface mesh model of the insulator string and the W-pin slot spatial posture information are identified to determine the multiple insertion path candidate solutions.

[0097] Furthermore, the environment perception and recognition module 13 is further configured to perform the following steps:

[0098] A wind disturbance perception module is used to continuously perceive wind disturbance in the assembly operation area to obtain a wind force-wind direction interference perception vector sequence; a first wind force-wind direction interference perception vector and a second wind force-wind direction interference perception vector are extracted from the wind force-wind direction interference perception vector sequence; an adjacency matrix analysis is performed on the first wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector to determine a first adjacency matrix; a third wind force-wind direction interference perception vector is extracted again, and an adjacency matrix analysis is performed on the third wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector, and so on to obtain an adjacency matrix set; a mean value of the adjacency matrix is ​​calculated, and the last wind force-wind direction interference perception vector in the wind force-wind direction interference perception vector sequence is enhanced to obtain a real-time W-pin insertion interference vector.

[0099] Furthermore, the environment perception and recognition module 13 is further configured to perform the following steps:

[0100] Element mapping similarity calculation is performed on the first wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector respectively to determine an element mapping similarity set; an initial adjacency matrix that is initially empty is obtained, and a mean calculation is performed on the element mapping similarity set, and the calculation result is added to the initial adjacency matrix to determine a first adjacency matrix.

[0101] Furthermore, the sensorless assembly module 15 is further configured to perform the following steps:

[0102] If the judgment is successful, the assembled W-pin is captured using a visual camera and the image is archived.

[0103] Furthermore, the sensorless assembly module 15 is further configured to perform the following steps:

[0104] Extract the force-displacement-friction three-dimensional interlocking parameters from the insertion state perception result; perform deviation identification based on the force-displacement-friction three-dimensional interlocking parameters and standard force-displacement-friction three-dimensional interlocking parameters to determine the force-displacement-friction three-dimensional interlocking deviation parameters; and perform feedback adjustment on the screened W-pin insertion path according to the force-displacement-friction three-dimensional interlocking deviation parameters to obtain the feedback-adjusted W-pin insertion path.

[0105] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0106] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0107] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A non-inductive assembly coordinated control method for W pins of an equipotential insulator string, characterized in that: The method comprises: Using a straddle-type intelligent operation vehicle equipped with a 3D laser radar and a visual camera, the insulator strings of the target transmission line are spatially scanned to construct a 3D surface mesh model of the insulator string and the spatial posture information of the W-pin slot. Combining the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot to perform insertion path trajectory recognition, and determine a set of candidate W-pin insertion paths; The wind disturbance perception module deployed in the robotic arm controller is called up to perceive and identify the assembly operation area and determine the real-time W-pin insertion interference vector. Filtering the W-pin insertion candidate path set based on the real-time W-pin insertion interference vector to obtain a filtered W-pin insertion path; The robotic arm controller receives the screened W-pin insertion path, controls the robotic arm and the thin electric claw to perform W-pin inductive assembly according to the screened W-pin insertion path, and uses the force sensor arranged at the end of the thin electric claw to capture the insertion state perception to determine whether the W-pin is assembled. If the judgment fails, the screened W-pin insertion path is feedback-adjusted according to the insertion state perception result, and the W-pin inductive assembly is performed again after obtaining the feedback-adjusted W-pin insertion path.

2. The non-inductive assembly coordinated control method of the W pins of the equipotential insulator string according to claim 1, characterized in that: Using a straddle-type intelligent operation vehicle equipped with a 3D laser radar and a visual camera, the insulator strings of the target transmission line are spatially scanned to construct a 3D surface mesh model of the insulator string and spatial posture information of the W-pin slot, including: Activate the IMU module of the straddle-type intelligent working vehicle for spatial registration and complete the coordinate system initialization of the three-dimensional laser radar and the visual camera; Starting the three-dimensional laser radar to perform high-density point cloud acquisition on the insulator string to obtain point cloud data; Synchronously starting the visual camera to collect image information of the insulator string to obtain an insulator string image; The point cloud data and the insulator string image are fused and recognized to determine the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot.

3. The non-inductive assembly coordinated control method of the W pins of the equipotential insulator string according to claim 2, characterized in that: The point cloud data and the insulator string image are fused and recognized to determine the insulator string three-dimensional surface mesh model and W-pin slot spatial posture information, including: spatially aligning the point cloud data and the insulator string image, and reconstructing a three-dimensional surface mesh model using an ICP iterative closest point algorithm and a sparse grid fitting algorithm based on the edge contour of the insulator string image and the density change of the point cloud data, to obtain a three-dimensional surface mesh model of the insulator string; Performing image segmentation based on the insulator string image to determine a W-pin slot spatial area; The slot posture parameters of the point cloud data are identified based on the W-pin slot spatial area, and the obtained slot center point position, insertion direction vector and slot depth are used as the W-pin slot spatial posture information.

4. The method for cooperative control of the non-inductive assembly of the W pins of the equipotential insulator string according to claim 1, characterized in that: Combining the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot to perform insertion path trajectory recognition and determine a set of candidate W-pin insertion paths, including: generating a plurality of candidate solutions for insertion paths based on the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot; The basic information of the W pin is obtained, and collision identification is performed on the insulator string three-dimensional surface mesh model according to the multiple insertion path candidate solutions. If no collision exists, the insulator string is added to the W pin insertion candidate path set.

5. The method for cooperative control of the non-inductive assembly of the W pins of the equipotential insulator string according to claim 4, characterized in that: include: An insertion path trajectory identifier is constructed with a set of sample W-pin slot spatial posture information as the insertion target, a set of sample insulator string 3D surface mesh models as interference factors, and a set of multiple sample insertion path candidate solutions as output. The insertion path trajectory identifier identifies the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot, and determines the multiple candidate solutions of the insertion path.

6. The method for cooperative control of the non-inductive assembly of the W pins of the equipotential insulator string according to claim 1, characterized in that: The wind disturbance perception module deployed in the robotic arm controller is called to perceive and identify the assembly operation area and determine the real-time W-pin insertion interference vector, including: Utilizing a wind disturbance perception module to continuously perceive wind disturbances in the assembly operation area, and obtaining a wind force-direction interference perception vector sequence; Extracting a first wind force-wind direction interference perception vector and a second wind force-wind direction interference perception vector from the wind force-wind direction interference perception vector sequence; performing an adjacency matrix analysis on the first wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector to determine a first adjacency matrix; Extracting the third wind force-wind direction interference perception vector again, performing adjacency matrix analysis on the third wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector, and so on, to obtain an adjacency matrix set; The mean value of the adjacency matrix is ​​calculated, and the last wind force-wind direction interference perception vector in the wind force-wind direction interference perception vector sequence is enhanced to obtain a real-time W-pin insertion interference vector.

7. The non-inductive assembly coordinated control method for W pins of an equipotential insulator string according to claim 6, characterized in that: Performing adjacency matrix analysis on the first wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector to determine a first adjacency matrix includes: Performing element mapping similarity calculation on the first wind force-wind direction interference perception vector and the second wind force-wind direction interference perception vector respectively to determine an element mapping similarity set; An initially empty initial adjacency matrix is ​​obtained, and a mean value of the element mapping similarity set is calculated. The calculation result is added to the initial adjacency matrix to determine a first adjacency matrix.

8. The non-inductive assembly coordinated control method for W pins of an equipotential insulator string according to claim 1, characterized in that: If the judgment is successful, the assembled W-pin is captured using a visual camera and the image is archived.

9. The method for cooperative control of the non-inductive assembly of the W pins of the equipotential insulator string according to claim 1, characterized in that: If the determination fails, feedback adjustment is performed on the screened W-pin insertion path according to the insertion state sensing result, and the feedback adjustment W-pin insertion path is obtained to perform W-pin sensorless assembly again, including: Extracting force-displacement-friction three-dimensional interlocking parameters from the insertion state sensing result; performing deviation identification based on the force-displacement-friction three-dimensional interlocking parameter and the standard force-displacement-friction three-dimensional interlocking parameter to determine the force-displacement-friction three-dimensional interlocking deviation parameter; Feedback adjustment is performed on the screened W-pin insertion path according to the force-displacement-friction three-dimensional interlocking deviation parameters to obtain the feedback-adjusted W-pin insertion path.

10. The non-inductive assembly coordinated control system of the W pin of the equipotential insulator string is characterized by: The system comprises: A spatial structure scanning module, which uses a straddle-type intelligent operation vehicle equipped with a 3D laser radar and a visual camera to perform spatial structural scanning on the insulator string of the target transmission line, constructing a 3D surface mesh model of the insulator string and spatial posture information of the W-pin slot; A path trajectory recognition module, the path trajectory recognition module is used to perform insertion path trajectory recognition based on the three-dimensional surface mesh model of the insulator string and the spatial posture information of the W-pin slot, and determine a set of candidate W-pin insertion paths; An environmental perception and recognition module, which is used to call the wind disturbance perception module deployed in the robotic arm controller to perceive and identify the assembly operation area and determine the real-time W-pin insertion interference vector; A path screening module, configured to screen the W-pin insertion candidate path set based on the real-time W-pin insertion interference vector to obtain a screened W-pin insertion path; The sensorless assembly module is used to receive the screened W-pin insertion path through the robotic arm controller, control the robotic arm and the thin electric claw to perform W-pin sensorless assembly according to the screened W-pin insertion path, and use the force sensor arranged at the end of the thin electric claw to capture the insertion state perception to determine whether the W-pin is assembled. If the judgment fails, the screened W-pin insertion path is feedback-adjusted according to the insertion state perception result, and the W-pin sensorless assembly is performed again after obtaining the feedback-adjusted W-pin insertion path.

Images