GIL through pipe auxiliary docking method and system based on machine vision

Abstract

The invention discloses a GIL through pipe auxiliary docking method and system based on machine vision. The method comprises the following steps: collecting a panoramic image of a construction site, synchronously obtaining spatial attitude data of image collecting equipment, extracting pipeline key position coordinates, and carrying out position calibration by combining equipment calibration parameters; a bird's-eye view is generated based on the calibrated image, a three-dimensional environment model is established, and the pose difference between the to-be-hoisted GIL pipeline and the fixed pipeline is calculated; the hoisting device moves the pipeline to the initial position according to the pose difference, and the mechanical arm executes butt joint operation based on the compensation instruction; in the butt joint process, the positions of a pipeline flange plate and a bolt hole are tracked in real time, the pose difference compensation value is dynamically adjusted, and accurate butt joint is ensured. By setting the operation boundary constraint of the hoisting device and the mechanical arm, the operation state is monitored in real time, the operation instruction is automatically updated when the boundary is exceeded, the butt joint precision is improved, and the risk of equipment damage or insulating medium leakage is avoided.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of power system equipment installation and intelligent control, and particularly relates to a GIL pipe auxiliary butt joint method and system based on machine vision. BACKGROUND

[0002] Gas-insulated Metal-enclosed Transmission Line (GIL pipe) is widely used in large-scale power transmission systems due to its transmission capacity and operational stability. With the development of power construction towards high voltage, large capacity and compactness, the installation and butt joint operation of GIL pipes presents the characteristics of limited space, complex line and high-altitude operation environment.

[0003] The patent application with publication number CN118635798A designs a mechanical butt joint scheme, which realizes precise positioning and reliable butt joint of pipe butt joint through the cooperative action of fixed upper end hoop and movable lower end hoop, and forced constraint of coaxiality with a guide mechanism. The patent application with publication number CN114512930A realizes fine adjustment of the height, inclination and circumference of the pipe to be butt joint by setting three independent control connecting ropes with three-degree-of-freedom attitude adjustment function and cooperating with optional moving devices, so that the flange face and bolt space can be automatically and accurately butt jointed.

[0004] However, in the context of large-scale power construction, the existing technology still has deficiencies. First, the construction space is limited; the GIL pipe covers a large area, and the on-site hoisting needs to realize long-distance high-precision pipe butt joint in a limited operation area, and the relative position accuracy of the flange hole position is extremely high. Second, the GIL pipe needs to be hoisted and spliced at a relatively high altitude, which is easily affected by wind and hoisting equipment. The pipe drifts and shakes during the butt joint process, reducing the accuracy and efficiency of positioning and installation. Finally, if the insulation medium in the GIL pipe is damaged and leaked due to misoperation, it will not only affect the safety of the operating personnel, but also affect the insulation performance of the equipment and bring environmental risks. SUMMARY

[0005] To solve the deficiencies in the existing technology, the present application provides a GIL pipe auxiliary butt joint method and system based on machine vision, which is based on a power operation robot and assisted by panoramic vision technology (AVM) to assist the hoisting equipment to complete the precise butt joint of the GIL pipe. In the context of large-scale power construction, the industrial hoisting equipment and the robot work cooperatively.

[0006] The first aspect of the present application discloses a GIL pipe auxiliary butt joint method based on machine vision, which adopts the following technical solution.

[0007] Collect panoramic images of GIL pipes in the construction site, and synchronously acquire spatial posture data of the image acquisition device; extract the pixel coordinates of the pipe outer contour, flange edge and bolt hole from the panoramic images, and perform spatial position calibration by combining the calibration parameters and spatial posture data of the image acquisition device; Generate an aerial view of the construction site based on the panoramic images after spatial position calibration, and establish a three-dimensional environment model; calculate the six-degree-of-freedom pose difference between the GIL pipe to be hoisted and the fixed GIL pipe according to the three-dimensional environment model; move the GIL pipe to be hoisted to the initial position according to the six-degree-of-freedom pose difference by the hoisting device; and calculate the six-degree-of-freedom pose difference again at the initial position as the initial compensation pose difference and send it to the mechanical arm; Based on the initial compensation pose difference, control the mechanical arm to perform the pipe butt joint operation; during the execution of the pipe butt joint operation, continuously track the position of the flange edge and bolt hole of the GIL pipe to be hoisted relative to the fixed GIL pipe, and predict the pose drift based on the multi-step historical pose difference data at the previous time step, dynamically update the compensation pose difference, and until the pipe butt joint operation is completed; Set the running boundary constraints of the hoisting device and the mechanical arm; continuously monitor the running state of the mechanical arm and the hoisting device during the execution of the butt joint operation, and update the butt joint operation instruction when the running state exceeds the running boundary constraint.

[0008] Further, the step of spatial position calibration comprises: According to the calibration information of the image acquisition device, convert the pixel coordinates in the two-dimensional image into the position coordinates in the three-dimensional space to obtain the spatial coordinates of the pipe outer contour, flange edge and bolt hole; According to the spatial posture data of the image acquisition device, convert the spatial coordinates of the pipe outer contour, flange edge and bolt hole into the global coordinate system of the construction site.

[0009] Further, the six-degree-of-freedom pose difference includes a translational position difference and a rotational attitude difference; Extract the key feature points of the GIL pipe to be hoisted and the fixed GIL pipe from the three-dimensional environment model, including the pipe outer contour, flange edge and bolt hole; each key feature point contains coordinates and rotational attitude in the global coordinate system; According to the key feature point clouds of the GIL pipe to be hoisted and the fixed GIL pipe, calculate the rotation matrix and translation vector therebetween by point cloud registration.

[0010] Further, the execution of the pipe butt joint operation comprises: Send the six-degree-of-freedom pose difference to the hoisting device in a digital signal instruction, and the hoisting device performs the first step of butt joint operation to move the pipe to be hoisted to the initial position according to the six-degree-of-freedom pose difference; In the initial position, the pose difference of the GIL pipeline to be hoisted relative to the fixed pipeline is calculated again as an initial compensation pose difference; based on the initial compensation pose difference, the mechanical arm is controlled to perform the second-step docking operation; During the second-step docking operation, the compensation pose difference is updated in real time to correct the operation of the mechanical arm; Further, the updating method of the compensation pose difference comprises: In the second-step docking operation, the real-time value of the pose difference at the current time step is calculated; based on the historical pose difference data of the previous time step The predicted pose difference at the future time step is output by linear modeling; The compensation pose difference at the current time step is updated in real time by the difference between the predicted pose difference and the real-time pose difference.

[0011] Further, the running boundary constraint comprises a spatial position boundary constraint, a contact boundary constraint, and a moving speed boundary constraint; In the spatial position boundary constraint, the maximum extension range and the maximum rotation angle of the hoisting device and the mechanical arm are set; In the contact boundary constraint, the distance between the clamping point of the mechanical arm and the docking surface of the GIL pipeline to be hoisted is greater than or equal to the clamping distance threshold; and the contact pressure applied by the clamping point of the mechanical arm to the GIL pipeline to be hoisted is less than or equal to the contact pressure threshold; In the moving speed boundary constraint, the maximum moving speed of the hoisting device is calculated according to the weight of the GIL pipeline to be hoisted, the carrying capacity of the hoisting equipment, and the length of the cable, as the moving speed boundary of the hoisting equipment.

[0012] Further, during the pipeline docking operation, the running state of the hoisting device and the mechanical arm is collected in real time, including the moving speed, the attitude change, and the contact pressure; whether the running state exceeds the running boundary constraint is analyzed, and if so, the pipeline docking operation is terminated; According to the spatial attitude data at the termination time, the six-degree-of-freedom pose difference of the GIL pipeline to be hoisted relative to the fixed GIL pipeline is recalculated.

[0013] The second aspect of the present application provides a GIL pipeline auxiliary docking system based on machine vision, which executes the technical solution of the auxiliary docking method provided by the first aspect of the present application. The system comprises: A spatial calibration module is used to collect panoramic images of GIL pipelines in the construction site, synchronously acquire spatial attitude data of the imaging device, extract pipeline outer contour, flange edge, and bolt hole pixel coordinates from the panoramic images, and perform spatial position calibration in combination with the calibration parameters and spatial attitude data of the imaging device; ​The docking instruction generation module generates a bird's eye view of the construction site based on the completion of the space position calibration panoramic image mapping, and establishes a three-dimensional environment model; according to the three-dimensional environment model, the six-degree-of-freedom pose difference of the GIL pipeline to be hoisted relative to the fixed GIL pipeline is calculated; the hoisting device moves the GIL pipeline to be hoisted to the initial position according to the six-degree-of-freedom pose difference; the six-degree-of-freedom pose difference is calculated again at the initial position as the initial compensation pose difference, and sent to the mechanical arm; The docking instruction execution module controls the mechanical arm to perform the pipeline docking operation based on the initial compensation pose difference; during the execution of the pipeline docking operation, the flange edge and bolt hole position of the GIL pipeline to be hoisted relative to the fixed GIL pipeline are continuously tracked, the pose drift is predicted with the multi-step historical pose difference data before the current time step, the compensation pose difference is dynamically updated, and the pipeline docking operation is completed; The docking operation safety constraint module is used to set the running boundary constraint of the hoisting device and the mechanical arm, and continuously monitor the contact pressure of the mechanical arm end and the hoisting device clamping point on the GIL pipeline to be hoisted during the execution of the docking operation, and update the pose adjustment instruction when the contact pressure exceeds the running boundary constraint.

[0014] The beneficial effects of the present application are that, compared with the prior art, 1. The present application ensures higher precision and reliability of the GIL pipeline docking operation through space position calibration based on machine vision, three-dimensional environment model construction and real-time dynamic updating of pose difference compensation value. During the docking process, the flange edge and bolt hole position are continuously tracked, and the pose drift is predicted combined with historical data, so as to realize accurate docking of the pipeline, reduce the possibility of human intervention, and reduce the error rate and accident risk in the operation process. The mechanical arm adjusts the motion trajectory in real time by receiving the updated pose difference compensation value, ensures the accurate alignment of the clamping point and the pipeline docking surface, and avoids pipeline damage caused by attitude error.

[0015] 2. The present application sets the running boundary constraint and monitors the running state of the hoisting device and the mechanical arm in real time, which can effectively prevent the mechanical arm from exerting excessive contact pressure, misoperation or inertia problem during the pipeline docking process, causing damage and leakage of the insulating medium in the GIL pipeline. Especially in the docking operation, the system automatically updates the docking instruction to ensure that the actions of the mechanical arm and the hoisting device are always within the safe range, avoid excessive clamping or unstable operation on the pipeline, and maximize the reduction of risks caused by excessive pressure or inaccurate pipeline attitude. By adjusting the pose difference compensation value in real time and monitoring the contact pressure, the system ensures the integrity of the insulating medium during the pipeline docking process, ensures the safety of the operating personnel and the insulation performance of the equipment, and reduces the environmental risk. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1The execution flow diagram of the GIL pipe-through auxiliary butt joint method provided for the embodiment is shown. DETAILED DESCRIPTION

[0017] To make the objectives, technical solutions, and advantages of the present application clearer, the technical solutions of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. The embodiments described in the present application are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art without creative work on the basis of the spirit of the present application shall fall within the protection scope of the present application.

[0018] As an embodiment of the present application, a visual acquisition device deployment mode for assisting GIL pipe-through butt joint is described. A plurality of fisheye cameras or 360° panorama cameras are configured on an industrial hoist above (may be near an electric hoist), a hoisting gantry, or a robot mechanical arm. It is ensured that the GIL pipe and the surrounding environment are covered, and a 360° panoramic view is acquired.

[0019] As an optional implementation of the embodiment, the mechanical arm can be installed on the hoisting gantry in cooperation, moves with the hoisting device, and the installation height can be adjusted according to actual conditions.

[0020] An inertial measurement unit (IMU) or a high-precision laser radar is configured on the electric hoist of the industrial hoist, and is kept synchronous with the acquisition of the visual image, so as to acquire the spatial pose data of the visual acquisition device in real time, including position, direction, inclination, and the like. As an optional implementation, if other hoisting devices are used, the inertial measurement unit is hoisted on the corresponding movable part.

[0021] As an optional implementation, the hoisting device and the mechanical arm can be scheduled and controlled by using a PLC control system or other types of unified control systems.

[0022] As an embodiment of the present application, a specific implementation of a GIL pipe-through auxiliary butt joint method based on machine vision is described. The execution flow diagram of the method is shown in Figure 1 .

[0023] S1: Acquire panoramic images around the GIL pipe at the construction site by using a plurality of angle acquisition devices, including panoramic images of the GIL pipe to be hoisted and the fixed pipe.

[0024] 1.1: The spatial pose data acquired synchronously with the panoramic images record the spatial position, angle, and rotation pose (such as the pitch angle) of the image acquisition device.

[0025] Distortion correction is performed on the panoramic image, especially the geometric distortion caused by the fisheye lens. The multi-angle images are stitched and registered to obtain a complete and distortion-free panoramic image. As an optional implementation, AVM technology can be called to obtain a complete and distortion-free GIL pipeline panoramic image.

[0026] 1.2: For the pre-processed panoramic image, an edge detection algorithm is used to identify the pipeline outer contour, flange edge and bolt hole in the image, and then the pixel coordinates of these key points in the two-dimensional panoramic image are converted into coordinates in the three-dimensional space using the calibration parameters of the imaging device.

[0027] The three-dimensional space coordinates at this time are in the coordinate system of the imaging device, which need to be converted into the global coordinate system of the construction site. The three-dimensional space coordinates of the pipeline outer contour, flange edge and bolt hole are converted to the global coordinate system of the construction site using the spatial pose data of the imaging device.

[0028] S2: Generate an aerial view of the construction site by mapping the panoramic image after completing spatial position calibration, establish a three-dimensional environment model and calculate the six-degree-of-freedom pose difference of the GIL pipeline to be hoisted relative to the fixed GIL pipeline.

[0029] 2.1: Generate an aerial view of the construction site using the panoramic image that has completed spatial calibration, which reflects the spatial layout of the entire construction area and provides an intuitive visual reference for subsequent calculations.

[0030] A three-dimensional environment model is established based on the aerial view and the extracted key feature point information (three-dimensional coordinates of the pipeline outer contour, flange edge and bolt hole in the global coordinate system) through computer vision and three-dimensional modeling technology. The three-dimensional environment model reflects the spatial distribution of the construction site, including the positions, poses and relative positional relationships of the GIL pipeline to be hoisted and the fixed GIL pipeline.

[0031] 2.2: Extract the coordinates of the key feature points of the GIL to be hoisted and the fixed pipeline from the three-dimensional environment model, represented as key feature point cloud and ; wherein , and represent the key feature point subsets of the pipeline outer contour, flange edge and bolt hole, respectively.

[0032] As an embodiment of the present embodiment, the SVD point cloud registration algorithm is used to calculate the optimal rigid transformation from to . The optimal rigid transformation is the pose difference between the GIL to be hoisted and the fixed pipeline, wherein is a rotation matrix, is a translation vector.

[0033] Using the SVD point cloud registration algorithm, the precise position and orientation difference between the to-be-lifted pipeline and the fixed pipeline can be accurately calculated. This precise docking difference ensures that the lifting device and the mechanical arm work more coordinately, reduces errors and problems in operation, and optimizes the docking process of the pipeline.

[0034] The control system sends the rotation matrix and the translation vector as control instructions to the lifting device to perform the coarse adjustment docking operation and drive the lifting device to move the to-be-lifted GIL pipeline to the initial position. The initial position here is a pre-set position, which can be understood as being close to a certain place of the fixed pipeline.

[0035] In the initial position, the pose difference between the to-be-lifted GIL and the fixed pipeline is calculated again according to the key feature point clouds of the two, and is sent to the mechanical arm as an initial pose difference compensation value for further docking operation.

[0036] S3: During the execution of the pipeline docking operation, the pose difference compensation value is dynamically updated until the pipeline docking operation is completed.

[0037] As an embodiment of the application, in the further docking process, a time window step is set to continuously track the point clouds of the key point positions of the to-be-lifted GIL pipeline and the fixed GIL pipeline.

[0038] Suppose that at the initial moment of the further docking operation , the mechanical arm receives the initial compensation pose difference , which is the initial target for the mechanical arm to run; the mechanical arm moves from the initial position to the target position according to the compensation pose difference. In this process, the imaging device on the mechanical arm captures the images of the docking area of the two pipelines in real time, including the high-precision point clouds of the flange edges and bolt holes of the two docking surfaces. During the movement of the mechanical arm, the real-time pose difference of the to-be-lifted pipeline relative to the fixed pipeline is calculated in real time.

[0039] According to the real-time pose difference , the movement instruction of the mechanical arm is adjusted, and the six components (three components of the translation position difference and three components of the rotation attitude difference) of the six-degree-of-freedom pose difference are independently linearly regressed according to the real-time value sequence of the pose difference at the current time step , fit the trend of change, output the predicted pose difference at the future time, including the rotation matrix and the translation vector.

[0040] The prediction error between the predicted pose difference and the real-time pose difference is calculated , which is represented as: ; wherein, is the predicted pose difference, is the inverse matrix of the real-time pose difference.

[0041] According to the prediction error, dynamically update the compensation pose difference sent to the robot arm, denoted as: ; wherein, is the compensation pose difference at the current time step is the compensation pose difference before updating.

[0042] By updating the pose difference compensation of the pipeline in real time, the robot arm can adjust the action according to the real-time data, avoiding the misalignment of the pipeline or the damage of the equipment caused by error accumulation or inaccurate operation. This greatly improves the safety during the construction process and avoids unnecessary equipment damage and collision problems.

[0043] S4: As an embodiment of the present application, in order to protect the docking process from failure caused by inertia problems, misoperation, etc., and to prevent the insulation medium from being damaged and leaking, the running state of the robot arm and the hoisting device is continuously monitored during the execution of the docking operation, and when the running state exceeds the running boundary constraint, the docking operation instruction is updated. Specifically: 4.1: Set the running boundary constraint of the hoisting device and the robot arm.

[0044] (1) Spatial position boundary constraint; The maximum extension range of the hoisting device is set according to the structure, carrying capacity and construction environment of the hoisting device. For example, the maximum operable distance of the hoisting device in the horizontal and vertical directions is defined to ensure that it will not exceed the safe operation range.

[0045] Maximum rotation angle of the robot arm: according to the design of the robot arm, set the maximum rotation angle range of the robot arm to prevent exceeding the physical limit or causing dangerous operation when rotating.

[0046] (2) Contact boundary constraint; Clamping distance threshold: set the minimum distance between the clamping point of the robot arm and the docking surface of the pipeline to be hoisted to ensure that the clamping position is not lower than the safety threshold. Usually this distance is determined based on the accuracy requirements of the docking operation and the size of the pipeline.

[0047] Contact pressure threshold: to protect the GIL pipeline to be hoisted, set the maximum contact pressure applied by the clamping point of the robot arm to prevent excessive pressure on the pipeline and avoid damaging the pipeline or affecting the insulation medium.

[0048] (3) Moving speed boundary constraint; ​Maximum moving speed of the lifting equipment: Based on the weight of the GIL pipe to be lifted, the load-bearing capacity of the lifting equipment, and the length of the cables, the maximum movable speed of the lifting equipment is calculated. Calculations may include the effects of factors such as gravity, cable length, and the load-bearing capacity of the lifting equipment to ensure safety and stability during the lifting process. Maximum moving speed The calculation method is expressed as follows: ; in, The adjustment coefficient is usually set according to the material properties; The weight of the GIL pipe to be hoisted. This represents the maximum load-bearing capacity of the hoisting device; This refers to the hoisting length of the cable. This is the maximum working length of the cable.

[0049] 4.2: Continuously monitor the operating status of the hoisting device and the robotic arm.

[0050] The system monitors the movement speed, position, and attitude changes of the hoisting device in real time, including horizontal and vertical movements. It also collects motion data of the robotic arm in real time, including rotation angle, clamping force, position, speed, and contact pressure applied to the pipe to be hoisted. The collected data is then constrained and compared through the control system.

[0051] When any operating parameter exceeds the preset boundary constraints, the system triggers an alarm and initiates an emergency shutdown or operation abort mechanism. If certain operating parameters are close to the boundary, the system will issue a warning signal in advance, prompting the operator to adjust the operating method.

[0052] As one implementation of this embodiment, if the operating state exceeds the set operating boundary constraints (e.g., the contact pressure applied by the robotic arm is too large, or the lifting device exceeds its maximum extension range), the current pipeline docking operation will be automatically terminated. A termination command will be immediately sent to the lifting device and the robotic arm to stop all docking operations.

[0053] The spatial attitude data, robotic arm status, and hoisting device status at the time of termination are saved to the operation log for subsequent analysis and debugging.

[0054] In a further implementation, based on the spatial attitude data at the time of abort, the positional difference between the GIL pipe to be hoisted and the fixed GIL pipe is recalculated, and the initial compensation value of the docking operation is adjusted.

[0055] The hoisting device and robotic arm continue to perform subsequent docking operations based on the new positional difference compensation value to ensure that the next operation does not exceed the boundary.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A machine vision-based GIL tube-assisted docking method, characterized in that, The method comprises the following steps: Collect panoramic images of GIL pipes in the construction site, and synchronously acquire spatial posture data of the image acquisition device; Extract the pixel coordinates of the pipe outer contour, flange edge and bolt hole from the panoramic images, and perform spatial position calibration by combining the calibration parameters and spatial posture data of the image acquisition device; Generate an aerial view of the construction site based on the panoramic images after spatial position calibration, and establish a three-dimensional environment model; calculate the six-degree-of-freedom pose difference between the GIL pipe to be hoisted and the fixed GIL pipe according to the three-dimensional environment model; move the GIL pipe to be hoisted to the initial position according to the six-degree-of-freedom pose difference by the hoisting device; calculate the six-degree-of-freedom pose difference again at the initial position as the initial compensation pose difference, and send it to the mechanical arm; Control the mechanical arm to perform pipe docking operation based on the initial compensation pose difference; continuously track the position of the flange edge and bolt hole of the GIL pipe to be hoisted relative to the fixed GIL pipe during the execution of the pipe docking operation, predict the pose drift by using the multi-step historical pose difference data at the previous time step, dynamically update the compensation pose difference, and stop until the pipe docking operation is completed; Set the running boundary constraints of the hoisting device and the mechanical arm; continuously monitor the running state of the mechanical arm and the hoisting device during the execution of the docking operation, and update the docking operation instruction when the running state exceeds the running boundary constraints.

2. The GIL tube-passing assisted docking method based on machine vision according to claim 1, characterized in that, The spatial position calibration step comprises: Convert the pixel coordinates in the two-dimensional image into the position coordinates in the three-dimensional space according to the calibration information of the image acquisition device, and acquire the spatial coordinates of the pipe outer contour, flange edge and bolt hole; Convert the spatial coordinates of the pipe outer contour, flange edge and bolt hole to the global coordinate system of the construction site according to the spatial posture data of the image acquisition device.

3. The GIL tube-traversing assisted docking method based on machine vision according to claim 1, characterized in that, The six-degree-of-freedom pose difference comprises a translation position difference and a rotation attitude difference; Extract the key feature points of the GIL pipe to be hoisted and the fixed GIL pipe from the three-dimensional environment model, including the pipe outer contour, flange edge and bolt hole; each key feature point contains the coordinates and rotation attitude in the global coordinate system; Calculate the rotation matrix and translation vector between the key feature point clouds of the GIL pipe to be hoisted and the fixed GIL pipe by point cloud registration.

4. The GIL tube-traversing assisted docking method based on machine vision of claim 1, wherein, Perform the pipe docking operation, comprising: Send the six-degree-of-freedom pose difference to the hoisting device in digital signal instruction, and move the pipe to be hoisted to the initial position according to the six-degree-of-freedom pose difference by the hoisting device in the first step of docking operation; Calculate the pose difference between the GIL pipe to be hoisted and the fixed pipe again at the initial position as the initial compensation pose difference; the mechanical arm performs the second step of docking operation based on the initial compensation pose difference; Update the compensation pose difference in real time to correct the operation of the mechanical arm during the second step of docking operation.

5. The GIL through-tubing assisted docking method based on machine vision according to any one of claims 1 or 4, characterized in that, The updating method of the compensation pose difference comprises: In the second step of the docking operation, the real-time value of the pose difference at the current time step is calculated; the predicted pose difference at the future time step is output by using linear modeling with the historical pose difference data at the current time step Pre vious steps Update the compensation pose difference of the current time step in real time by the difference between the predicted pose difference and the real-time pose difference.

6. The method of claim 1, wherein, The running boundary constraints comprise spatial position boundary constraints, contact boundary constraints and moving speed boundary constraints; In the spatial position boundary constraints, the maximum extension range and the maximum rotation angle of the hoisting device and the mechanical arm are set; In the contact boundary constraint, the distance between the mechanical arm clamping point and the butt joint surface of the GIL pipeline to be hoisted is greater than or equal to the clamping distance threshold; and the contact pressure applied by the mechanical arm clamping point to the GIL pipeline to be hoisted is less than or equal to the contact pressure threshold. In the moving speed boundary constraint, the maximum moving speed of the hoisting device is calculated according to the weight of the GIL pipeline to be hoisted, the carrying capacity of the hoisting device and the length of the cable, and is taken as the moving speed boundary of the hoisting device.

7. The GIL pipeline butt joint method based on machine vision according to any one of claims 1 or 6, characterized in that, During the pipeline butt joint operation, the running state of the hoisting device and the mechanical arm is collected in real time, including the moving speed, the attitude change and the contact pressure; whether the running state exceeds the running boundary constraint is analyzed, and if so, the pipeline butt joint operation is stopped; According to the spatial attitude data at the stopping moment, the six-degree-of-freedom pose difference of the GIL pipeline to be hoisted relative to the fixed GIL pipeline is recalculated.

8. A machine vision based GIL tube walking aid docking system, performing the method of aiding docking according to any one of claims 1 to 7, characterized in that, The system comprises: a spatial calibration module, which is used to collect panoramic images of the GIL pipelines in the construction site, synchronously acquire spatial attitude data of the imaging device, extract the pixel coordinates of the pipeline outer contour, flange edge and bolt hole from the panoramic images, and perform spatial position calibration in combination with the calibration parameters of the imaging device and the spatial attitude data; a butt joint instruction generation module, which is used to generate a bird's eye view of the construction site based on the panoramic images after spatial position calibration, establish a three-dimensional environment model, calculate the six-degree-of-freedom pose difference of the GIL pipeline to be hoisted relative to the fixed GIL pipeline according to the three-dimensional environment model, move the GIL pipeline to be hoisted to an initial position according to the six-degree-of-freedom pose difference, and calculate the six-degree-of-freedom pose difference again at the initial position as an initial compensation pose difference and send it to the mechanical arm; a butt joint instruction execution module, which is used to control the mechanical arm to perform the pipeline butt joint operation based on the initial compensation pose difference, continuously track the positions of the flange edge and the bolt hole of the GIL pipeline to be hoisted relative to the fixed GIL pipeline during the execution of the pipeline butt joint operation, predict the pose drift with the multi-step historical pose difference data before the current time step, dynamically update the compensation pose difference, and stop until the pipeline butt joint operation is completed; a butt joint operation safety constraint module, which is used to set the running boundary constraint of the hoisting device and the mechanical arm, continuously monitor the contact pressure of the mechanical arm end and the hoisting device clamping point on the GIL pipeline to be hoisted during the execution of the butt joint operation, and update the pose adjustment instruction when the contact pressure exceeds the running boundary constraint.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The computer program is loaded into the processor to realize the butt joint assisting method according to any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program is executed by the processor to realize the butt joint assisting method according to any one of claims 1-7.

Images