Digital twin automatic docking method based on visual detection device

Abstract

This application discloses a digital twin automatic docking method based on a visual inspection device. The method includes: constructing a digital twin model based on an automatic docking unit and three-dimensional drawings of a cylinder segment; constructing a twin data model based on data measured by the visual inspection device, the docking device's operating path data, and docking force feedback data; the driving mechanism of the twin model is built according to the cylinder segment docking process, and receives visual inspection data through a specific data interface to drive the twin model's movement; the digital twin model generates a docking path based on the cylinder segment's spatial coordinate position and relative attitude deviation, and performs automatic docking pre-rehearsal in the digital twin docking system to automatically optimize the docking path; the docking path includes six degrees of freedom displacement settings, namely feed displacement, lateral displacement, heave displacement, roll angle, pitch displacement, and yaw displacement; the digital twin docking system optimizes the path based on historical docking data.

Description

Technical Field

[0001] This invention relates to the field of digital twin docking for shell-type products, and more particularly to an automatic digital twin docking method based on a visual inspection device. Background Technology

[0002] In the traditional assembly process of complex electromechanical products in aerospace, the docking accuracy is low, the success rate of automatic docking is not high, and docking data is difficult to collect. Furthermore, the automatic docking process often relies on manual operation at the end, resulting in frequent quality problems and making it difficult to guarantee docking success.

[0003] To ensure the success rate and quality of product docking and reduce docking errors, it is necessary to build a digital twin docking system. By simulating the product docking process through a twin model of the docking device, the docking path can be optimized, and the relatively optimal displacement and angular offset of the X-axis, Y-axis and Z-axis can be output. Furthermore, the docking forces in each direction can be simulated and analyzed to prevent product quality problems caused by uneven force.

[0004] Existing typical product docking processes rely on assembly workers to fix the cylinder segment to the docking equipment using tooling. The docking is then completed by visual observation, verbal commands, and jogging control of the docking device to adjust the cylinder segment's posture. This process suffers from low docking accuracy, low efficiency, and a low success rate. Therefore, researching and designing an efficient and reliable digital twin automated docking method is crucial for improving the quality and accuracy of cylinder segment docking. Summary of the Invention

[0005] To address the shortcomings of existing methods, this invention provides a digital twin automatic docking method based on a visual inspection device to achieve high-efficiency and high-precision docking of cylinder segments. By deploying a digital twin docking system to receive and process measurement data from the visual inspection device, a simulated and optimized docking path is obtained, which can improve the success rate and accuracy of cylinder segment docking, and achieve the effects of real-time display of docking data and autonomous optimization of docking paths based on historical docking data.

[0006] To achieve the above objectives, the present invention provides a method for automatic digital twin docking based on a visual inspection device, the method comprising:

[0007] A digital twin model is constructed based on the automatic docking unit and the 3D drawings of the cylinder section;

[0008] A twin data model is constructed based on data measured by the visual inspection device, data on the docking device's operating path, and data on docking force feedback.

[0009] The driving mechanism of the digital twin model is built based on the tube segment docking process, and receives visual inspection data through a specific data interface to drive the movement of the digital twin model; the digital twin model generates a docking path based on the spatial coordinate position and relative attitude deviation of the tube segment, and performs automatic docking pre-rehearsal in the digital twin docking system to automatically optimize the docking path;

[0010] The docking path includes six degrees of freedom of displacement settings, namely feed displacement, lateral displacement, heave displacement, roll angle, pitch displacement, and yaw displacement; the digital twin docking system optimizes the path based on historical docking data.

[0011] Furthermore, the digital twin model includes a docking device twin model, a cylindrical section twin model, and an auxiliary facility twin model. The docking device twin model sets up model kinematic pairs according to the motion mechanism of the real device, and drives the model kinematic pairs to move according to real-time data.

[0012] Furthermore, the visual inspection device is used to detect the spatial coordinate positions of the A-type cylindrical target and the B-type cylindrical target, calculate the attitude deviation of the B-type cylindrical target relative to the A-type cylindrical target based on the relative coordinate relationship, and send the spatial coordinate position data and attitude deviation data to the digital twin docking system.

[0013] Furthermore, the digital twin docking system includes a data access module, a twin model construction module, a 3D scene module, and a data processing module. The data access module communicates with the vision alignment measurement system and the docking equipment control system via a network, and obtains device motion information and measurement point position data after parsing the network data according to the protocol. The twin model construction module completes the modeling of the twin model of the inner cylinder section of the automatic docking unit and the twin model of the docking device, and establishes the final assembly docking model. On the one hand, the final assembly docking model can be serialized and used in the 3D scene module. On the other hand, the completed final assembly docking model uploads the static data of the model to the data management module through the database interface. The 3D scene module reconstructs the entire final assembly docking model through the file interface. After the data access module calls the model interface, it completes the simulation motion and simultaneously sends the real-time dynamic data asynchronously to the data processing module. The data processing module describes the physical world of the docking process based on a relational model, stores the modeled static data and dynamic driving data in a structured manner, and optimizes the docking path by querying and analyzing the historical data of final assembly docking.

[0014] Furthermore, the docking history data in the method includes: the docking path before the current moment, composed of feed displacement value X, lateral displacement value Y, heave displacement value Z, roll angle value a, pitch displacement value b, and yaw displacement value c; and the docking time T and X-axis docking stress F recorded after docking. x Y-axis butt joint stress Fy Z-axis butt joint stress F z The i-th docking path is recorded as P. i ={X i Y i Z i a i b i c i The result of the i-th docking is recorded as R. i ={T i F xi F yi F zi}, continuously accumulating to form a historical docking dataset P, and recording R from docking results according to Dijkstra's algorithm. i The optimal docking path P is selected by iterative selection. i and P i The initial docking path is used for iterative optimization, thereby reducing docking time, improving docking accuracy, and increasing docking success rate.

[0015] Furthermore, the visual inspection device in the method detects the outline data of the docking surfaces of the A-type and B-type cylindrical segments and sends it to the digital twin docking system. The outline data includes the major and minor radii of the end face circles.

[0016] Furthermore, the digital twin system in the method determines whether the set minimum docking accuracy requirements are met based on the outline data of the A-type and B-type cylindrical sections. If the set minimum docking accuracy requirements are met, the digital twin docking system generates the docking device operation path based on the spatial coordinate position data and attitude deviation data of the A-type and B-type cylindrical sections. The digital twin model can perform docking pre-rehearsal based on the generated operation path and automatically optimize the docking path.

[0017] Furthermore, the docking path optimization process in the method is as follows:

[0018] Step 1: Generate the initial docking path P based on the spatial coordinate position data and attitude deviation data of the vision inspection device. i The corresponding initial docking time is T. i and the value of the contact force F i F i The numerical values ​​of the joint force F in the X, Y, and Z axis directions. xi F yi and F zi The sum;

[0019] Step 2: Set the initial docking time T i and the value of the contact force F i Time T from the last docking i-1 Compare with the set docking force; if the docking time T is met...i ≤T i-1 , docking force F i If the value is not greater than the set value, the optimization process ends and the docking path is output; otherwise, the docking path is regenerated, and the X-axis docking stress F is compared. x Y-axis butt joint stress F y Z-axis butt joint stress F z The magnitude of the stress difference F between the maximum and minimum values ​​is obtained. d If F d If the value is not 0, the displacement value in the direction corresponding to the maximum butt stress will be finely adjusted by an adjustment amount of [value missing]. If F d If the value is 0, then the displacement values ​​in the X, Y, and Z axes are adjusted by an amount of [value missing].

[0020] Step 3: The regenerated docking path includes the adjusted feed displacement value X. i+1 Horizontal displacement value Y i+1 Lifting displacement value Z i+1 And the previous roll angle value a i Pitch displacement value b i and yaw displacement value c i Based on these displacement and angle values, the twin model of the docking device is driven to perform simulation, and the docking time and docking stress on the X-axis, Y-axis and Z-axis are calculated.

[0021] Step 4: Output the docking time and docking force values. If the docking time T... i ≤T i-1 , docking force F i If the value is not greater than the set value, the optimization process ends and the docking operation path is output; otherwise, the docking path is regenerated and the optimization is iterated again.

[0022] Compared with existing technologies, the significant advantages of this invention are as follows: 1. The digital twin automatic docking method based on a visual inspection device provided by this invention realizes the key data acquisition and visual monitoring of the final assembly docking process, solving the problems of low transparency and difficulty in data collection during the final assembly docking process; 2. The digital twin automatic docking method based on a visual inspection device provided by this invention realizes the simulation optimization of the final assembly docking process, and can output docking paths with short docking time and high docking accuracy, solving the problems of low efficiency and low docking quality in the final assembly docking process; 3. The digital twin automatic docking method based on a visual inspection device provided by this invention realizes high-frequency model updates driven by real-time data, constructing a virtual docking scenario that is synchronized with the on-site docking process in real time; 4. The digital twin automatic docking method based on a visual inspection device provided by this invention can effectively collect and process data from the final assembly docking process, and provide support for optimizing the docking process by constructing a docking historical data model, effectively integrating and utilizing docking process data. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the physical device composition structure in the first application scenario of the present invention;

[0024] Figure 2 This is a flowchart of the automatic docking path optimization process for digital twins based on a visual inspection device, according to the present invention.

[0025] Figure 3 This is a framework diagram of the digital twin automatic docking unit of the present invention;

[0026] Figure 4 This is a schematic diagram of the physical device composition structure for the second application scenario of the present invention;

[0027] Figure 1 As shown: 1. Type A docking device, 2. Type A cylindrical section, 3. Visual inspection device, 4. Type B cylindrical section, 5. Type B docking device;

[0028] Figure 4 As shown: 11. Digital twin docking system operation terminal, 12. Type A cylinder section, 13. Type B cylinder section, 14. Finished product temporary storage area, 15. Vision inspection device, 16. Docking robot, 17. Pre-adjustment table, 18. Handling robot, 19. Incoming material buffer area. Detailed Implementation

[0029] The above content is only an overview of the technical solution of the present invention. In order to better demonstrate the design principle, working characteristics and advantages of the present invention, the following will describe it clearly, in detail and completely with reference to the accompanying drawings.

[0030] Example 1

[0031] like Figure 1As shown, the physical equipment components of the digital twin automatic docking method based on a visual inspection device disclosed in this invention include: A-type docking device 1, A-type cylindrical section 2, visual inspection device 3, B-type cylindrical section 4, and B-type docking device 5. Type A docking device 1 includes two support and attitude adjustment RGVs. Each support and attitude adjustment RGV has four degrees of freedom adjustment functions, including walking, lifting, lateral movement, and rolling. The two support and attitude adjustment RGVs work together to achieve pitch and yaw, thus realizing the six-degree-of-freedom adjustment function of the automatic docking and transfer device. Type B docking device 5 adopts a double-support structure, consisting of a walking mechanism, a feeding mechanism, two sets of support and attitude adjustment RGVs (RGV1 and RGV2) and a bracket. The walking mechanism enables the product to move forward along the track direction. The feeding mechanism is used for fine-tuning the attitude along the track direction during docking. Each support and attitude adjustment RGV has four degrees of freedom adjustment functions, including lifting, lateral movement, and rolling. The two support and attitude adjustment RGVs work together to achieve pitch and yaw. The vision inspection device 3 captures the position information of the product through the cooperation of a camera and auxiliary tooling. It mainly realizes camera control, lower-level machine control and data exchange, including modules such as industrial control computer, camera controller and binocular camera. The modules are connected by data lines or wireless communication, and communicate based on the TCP / IP protocol. The operation process is as follows: Before the automatic docking process begins, the A-type docking device 1 and the A-type cylinder segment 2 are in a fixed state during the final assembly docking process. Operators need to attach the targets to the corresponding positions. The spatial coordinate positions of the targets of the A-type cylinder segment 2 and the B-type cylinder segment 4 can be detected by the vision inspection device 3. The vision alignment measurement system software calculates the attitude deviation of the B-type cylinder segment 4 relative to the A-type cylinder segment 2 based on the relative coordinate relationship, and sends the spatial coordinate position data and attitude deviation to the digital twin docking system. The system generates an initial docking operation path based on these data, performs autonomous optimization simulation of the docking path according to the optimization process, until a docking operation path that meets the docking requirements is output and sent to the docking device. The docking device automatically docks the cylinder segments according to the optimized docking operation command.

[0032] The aforementioned docking device has a built-in Beckhoff CX5130 embedded controller, which controls multiple motor drivers via EtherCAT bus, enabling control of the motors of each motion mechanism.

[0033] The aforementioned visual inspection device includes a high-precision real-time measuring camera and a camera controller. The former measures the positioning pins and pin holes on the end face of the cylinder section, while the latter controls the camera to perform synchronous data acquisition.

[0034] The aforementioned visual inspection device includes visual alignment and measurement system software, which can realize functions such as image processing, coordinate transformation, data communication, and real-time control. It mainly includes the following steps:

[0035] Step 1: The software processes the images captured by the camera and calculates the three-dimensional coordinates of the coded points in the field of view in the camera coordinate system based on the camera calibration data;

[0036] Step 2: The software calculates the six-degree-of-freedom coordinate information of each measured object based on the coordinates of the coded points. This mainly includes the coordinate system FrameA for fixing the A-type cylindrical section, the coordinate system FrameB for adjusting the B-type cylindrical section, and the coordinate system FrameTCP for the end TCP of the pose adjustment mechanism.

[0037] Step 3: Transform the coordinate system of all measurement points to the coordinate system of the adjustment structure, and calculate the coordinates of the coded points after the docking is completed.

[0038] The aforementioned docking device adopts a dual-support structure, consisting of a walking mechanism, a feeding mechanism, and two sets of support and attitude adjustment RGVs (RGV1 and RGV2). The walking mechanism enables the cylinder section to move forward along the track direction; the feeding mechanism is used for fine-tuning the attitude along the track direction during docking. Each support and attitude adjustment RGV has four degrees of freedom adjustment functions, including walking, lifting, lateral movement, and rolling.

[0039] Example 2

[0040] like Figure 4 As shown, this invention discloses an automatic digital twin docking method based on a visual inspection device, the method comprising:

[0041] A digital twin model is constructed based on the automatic docking unit and the 3D drawings of the cylinder section;

[0042] A twin data model is constructed based on data measured by the visual inspection device, data on the docking device's operating path, and data on docking force feedback.

[0043] The driving mechanism of the digital twin model is built based on the tube segment docking process, and receives visual inspection data through a specific data interface to drive the movement of the digital twin model; the digital twin model generates a docking path based on the spatial coordinate position and relative attitude deviation of the tube segment, and performs automatic docking pre-rehearsal in the digital twin docking system to automatically optimize the docking path;

[0044] The docking path includes six degrees of freedom of displacement settings, namely feed displacement, lateral displacement, heave displacement, roll angle, pitch displacement, and yaw displacement; the digital twin docking system optimizes the path based on historical docking data.

[0045] Furthermore, the digital twin model includes a docking device twin model, a cylindrical section twin model, and an auxiliary facility twin model. The docking device twin model sets up model kinematic pairs according to the motion mechanism of the real device, and drives the model kinematic pairs to move according to real-time data.

[0046] Furthermore, the visual inspection device is used to detect the spatial coordinate positions of the A-type cylindrical target and the B-type cylindrical target, calculate the attitude deviation of the B-type cylindrical target relative to the A-type cylindrical target based on the relative coordinate relationship, and send the spatial coordinate position data and attitude deviation data to the digital twin docking system.

[0047] Furthermore, the digital twin docking system includes a data access module, a twin model construction module, a 3D scene module, and a data processing module. The data access module communicates with the vision alignment measurement system and the docking equipment control system via a network, and obtains device motion information and measurement point position data after parsing the network data according to the protocol. The twin model construction module completes the modeling of the twin model of the inner cylinder section of the automatic docking unit and the twin model of the docking device, and establishes the final assembly docking model. On the one hand, the final assembly docking model can be serialized and used in the 3D scene module. On the other hand, the completed final assembly docking model uploads the static data of the model to the data management module through the database interface. The 3D scene module reconstructs the entire final assembly docking model through the file interface. After the data access module calls the model interface, it completes the simulation motion and simultaneously sends the real-time dynamic data asynchronously to the data processing module. The data processing module describes the physical world of the docking process based on a relational model, stores the modeled static data and dynamic driving data in a structured manner, and optimizes the docking path by querying and analyzing the historical data of final assembly docking.

[0048] Furthermore, the docking history data in the method includes: the docking path before the current moment, composed of feed displacement value X, lateral displacement value Y, heave displacement value Z, roll angle value a, pitch displacement value b, and yaw displacement value c; and the docking time T and X-axis docking stress F recorded after docking. x Y-axis butt joint stress F y Z-axis butt joint stress F z The i-th docking path is recorded as P. i ={X i Y i Z i a i b i c i The result of the i-th docking is recorded as R. i ={T i Fx i F yi F zi}, continuously accumulating to form a historical docking dataset P, and recording R from docking results according to Dijkstra's algorithm. i The optimal docking path P is selected by iterative selection. i and P iThe initial docking path is used for iterative optimization, thereby reducing docking time, improving docking accuracy, and increasing docking success rate.

[0049] Furthermore, the visual inspection device in the method detects the outline data of the docking surfaces of the A-type and B-type cylindrical segments and sends it to the digital twin docking system. The outline data includes the major and minor radii of the end face circles.

[0050] Furthermore, the digital twin system in the method determines whether the set minimum docking accuracy requirements are met based on the outline data of the A-type and B-type cylindrical sections. If the set minimum docking accuracy requirements are met, the digital twin docking system generates the docking device operation path based on the spatial coordinate position data and attitude deviation data of the A-type and B-type cylindrical sections. The digital twin model can perform docking pre-rehearsal based on the generated operation path and automatically optimize the docking path.

[0051] Furthermore, the docking path optimization process in the method is as follows:

[0052] Step 1: Generate the initial docking path P based on the spatial coordinate position data and attitude deviation data of the vision inspection device. i The corresponding initial docking time is T. i and the value of the contact force F i F i The numerical values ​​of the joint force F in the X, Y, and Z axis directions. xi F yi and F zi The sum;

[0053] Step 2: Set the initial docking time T i and the value of the contact force F i Time T from the last docking i-1 Compare with the set docking force; if the docking time T is met... i ≤T i-1 , docking force F i If the value is not greater than the set value, the optimization process ends and the docking path is output; otherwise, the docking path is regenerated, and the X-axis docking stress F is compared. x Y-axis butt joint stress F y Z-axis butt joint stress F z The magnitude of the stress difference F between the maximum and minimum values ​​is obtained. d If F d If the value is not 0, the displacement value in the direction corresponding to the maximum butt stress will be finely adjusted by an adjustment amount of [value missing]. If F d If the value is 0, then the displacement values ​​in the X, Y, and Z axes are adjusted by an amount of [value missing].

[0054] Step 3: The regenerated docking path includes the adjusted feed displacement value X. i+1 Horizontal displacement value Y i+1 Lifting displacement value Z i+1 And the previous roll angle value a i Pitch displacement value b i and yaw displacement value c i Based on these displacement and angle values, the twin model of the docking device is driven to perform simulation, and the docking time and docking stress on the X-axis, Y-axis and Z-axis are calculated.

[0055] Step 4: Output the docking time and docking force values. If the docking time T... i ≤T i-1 , docking force F i If the value is not greater than the set value, the optimization process ends and the docking operation path is output; otherwise, the docking path is regenerated and the optimization is iterated again.

[0056] The aforementioned digital twin docking system includes a data access module, a twin model construction module, a 3D scene module, and a data processing module. The data access module communicates with the vision alignment measurement system and the docking equipment control system via a network. After parsing the network data according to the protocol, it obtains the device motion information and measurement point position data. The twin model construction module can complete the modeling of the twin models of the inner cylinder section of the automatic docking unit and the docking device, establishing a final assembly docking model. On the one hand, the final assembly docking model can be serialized and used in the 3D scene module; on the other hand, the completed final assembly docking model uploads its static data to the data management module through a database interface. The 3D scene module reconstructs the entire final assembly docking model through a file interface. After the data access module calls the model interface, it completes the simulation motion, updates the graphical dashboard, the visualized model within the 3D scene, and data labels in real time, and asynchronously sends the real-time dynamic data to the data processing module. The data processing module describes the physical world of the docking process based on a relational model, stores the modeled static data and dynamic driving data in a structured manner, and performs query analysis through historical final assembly docking data to optimize the docking path.

[0057] The physical equipment components of the above-mentioned docking method application scenario include: a digital twin docking system operation terminal 11, a type A cylindrical section 12, a type B cylindrical section 13, a finished product temporary storage area 14, a vision inspection device 15, a docking robot 16, a pre-adjustment table 17, a handling robot 18, and an incoming material buffer area 19. The vision inspection device 15 consists of a 3D vision camera, a horizontal motion platform, and a camera movement platform, used for measuring the axis and pin hole positions of the cylindrical section products, and has three functions: positioning and photography, data extraction, and data processing. The docking robot 16 consists of a KUKA robot, a seventh-axis guide rail, a six-dimensional force sensor, a clamping mechanism, a vision camera, and a control system, responsible for the flexible docking and docking force detection of the cylindrical section products. The handling robot 18 consists of a KUKA robot, a clamping mechanism, a seventh-axis guide rail, and a control system, responsible for the handling of cylindrical section products on and off the line, as well as the material transfer between various subsystems within the docking unit. The operation process is as follows: The cylindrical segment product is placed in the incoming material buffer area 19. The handling robot 18 transports the cylindrical segment to the pre-adjustment table 17 to complete the end face deformation and pin hole position measurement and adjustment. The docking robot 16 transports the cylindrical segment to the docking table. The vision inspection device 15 calculates the spatial coordinate positions of the A-type cylindrical segment 12 and the B-type cylindrical segment 13 by taking pictures. The vision alignment measurement system software calculates the attitude deviation of the B-type cylindrical segment 13 relative to the A-type cylindrical segment 12 according to the relative coordinate relationship, and sends the spatial coordinate position data and attitude deviation to the digital twin docking system. The system generates an initial docking operation path based on these data, and performs autonomous optimization simulation of the docking path according to the optimization process until a docking operation path that meets the docking requirements is output and sent to the docking robot 16. The docking robot 16 automatically docks the cylindrical segments according to the optimized docking instructions, and measures the docking force through the six-dimensional force sensor at the end of the robot when the docking is completed, and sends the docking force data to the digital twin docking system. The system binds this robot operation path and docking force data and stores it in the system database as historical data.

[0058] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A digital twin automatic docking method based on a visual inspection device, characterized in that, The method includes: A digital twin model is constructed based on the automatic docking unit and the 3D drawings of the cylinder section; A twin data model is constructed based on data measured by the visual inspection device, data on the docking device's operating path, and data on docking force feedback. The driving mechanism of the digital twin model is built based on the tube segment docking process, and receives visual inspection data through a specific data interface to drive the movement of the digital twin model; the digital twin model generates a docking path based on the spatial coordinate position and relative attitude deviation of the tube segment, and performs automatic docking pre-rehearsal in the digital twin docking system to automatically optimize the docking path; The docking path includes six degrees of freedom of displacement settings: feed displacement, lateral displacement, heave displacement, roll angle, pitch displacement, and yaw displacement; the digital twin docking system optimizes the path based on historical docking data. The process for optimizing the docking path is as follows: Step 1: Generate the initial docking path based on the spatial coordinate position data and attitude deviation data of the vision inspection device. The corresponding initial docking time is and the value of the connection force , The values ​​of the joint force in the X, Y, and Z axes. , as well as The sum; Step 2: Set the initial docking time and the value of the connection force Compared to the last docking time Compare with the set docking force; if the docking time is met... , to connect If the value is not greater than the set value, the optimization process ends and the docking operation path is output; otherwise, the docking path is regenerated, and the X-axis docking stress is compared. Y-axis butt joint stress Z-axis butt joint stress The magnitude of the stress is determined, and the stress difference between the maximum and minimum values ​​is obtained. ,like If the value is not 0, the displacement value in the direction corresponding to the maximum butt stress will be finely adjusted by - mm; if If the value is 0, then the displacement values ​​in the X, Y, and Z axes will be adjusted by - mm; Step 3: The regenerated docking path includes the adjusted feed displacement values. Lateral displacement value Lifting and lowering displacement values and the previous roll angle value Pitch displacement value and yaw displacement value Based on these displacement and angle values, the twin model of the docking device is driven to perform simulation, and the docking time and docking stress on the X-axis, Y-axis and Z-axis are calculated. Step 4: Output the docking time and docking force values. (If the docking time...) , to connect If the value is not greater than the set value, the optimization process ends and the docking operation path is output; otherwise, the docking path is regenerated and the optimization is iterated again.

2. The automatic digital twin docking method based on a visual inspection device according to claim 1, characterized in that, The digital twin model includes a docking device twin model, a cylindrical section twin model, and an auxiliary facility twin model. The docking device twin model sets up model kinematic pairs based on the motion mechanism of the real device, and drives the model kinematic pairs to move according to real-time data.

3. The automatic digital twin docking method based on a visual inspection device according to claim 1, characterized in that, The visual inspection device is used to detect the spatial coordinate positions of the A-type cylindrical target and the B-type cylindrical target, calculate the attitude deviation of the B-type cylindrical target relative to the A-type cylindrical target based on the relative coordinate relationship, and send the spatial coordinate position data and attitude deviation data to the digital twin docking system.

4. The automatic digital twin docking method based on a visual inspection device according to claim 1, characterized in that, The digital twin docking system includes a data access module, a twin model construction module, a 3D scene module, and a data processing module. The data access module communicates with the vision alignment measurement system and the docking equipment control system via a network, parsing network data according to a protocol to obtain device motion information and measurement point location data. The twin model construction module completes the modeling of the twin models of the inner cylinder section of the automatic docking unit and the docking device, establishing a final assembly docking model. This final assembly docking model can be serialized and used in the 3D scene module. Furthermore, the completed final assembly docking model uploads its static data to the data management module via a database interface. The 3D scene module reconstructs the entire final assembly docking model via a file interface. After the data access module calls the model interface, it completes the simulation motion and asynchronously sends real-time dynamic data to the data processing module. The data processing module describes the physical world of the docking process based on a relational model, structurally storing the modeled static data and dynamic driving data. It can also optimize the docking path through querying and analysis of historical final assembly docking data.

5. The automatic digital twin docking method based on a visual inspection device according to claim 1, characterized in that, The docking history data in the method includes the docking path up to the current moment, based on the feed displacement value. Lateral displacement value Lifting and lowering displacement values Roll angle value Pitch displacement value and yaw displacement value Composition, and the docking time recorded after docking. X-axis butt joint stress Y-axis butt joint stress Z-axis butt joint stress , No. The secondary docking path record is as follows , No. The docking result was recorded as follows: Continuously accumulate and form a docking historical dataset According to Dijkstra's algorithm, the docking results are recorded. Iterate through the network to select the optimal docking path. and will Iterative optimization is performed using this as the initial docking path.

6. The automatic digital twin docking method based on a visual inspection device according to claim 1, characterized in that, The visual inspection device in the method detects the outline data of the docking surfaces of the A-type and B-type cylindrical sections and sends it to the digital twin docking system. The outline data includes the major and minor radii of the end face circles.

7. The automatic digital twin docking method based on a visual inspection device according to claim 1, characterized in that, The digital twin system in the method determines whether the set minimum docking accuracy requirements are met based on the outline data of the A-type and B-type cylindrical sections. If the set minimum docking accuracy requirements are met, the digital twin docking system generates the docking device operation path based on the spatial coordinate position data and attitude deviation data of the A-type and B-type cylindrical sections. The digital twin model can perform docking pre-rehearsal based on the generated operation path and automatically optimize the docking path.

Images