An automatic spraying device for inkjet printing equipment

Abstract

The application discloses an automatic spraying device for a spray printing device, and belongs to the technical field of spraying, and solves the problem that existing devices can only process planes and cannot cope with curved surfaces and irregular surfaces, and the device comprises a workbench provided with a rotating clamping piece, a multi-dimensionally adjustable spraying assembly and a processing unit. In operation, the clamping piece fixes a workpiece and is driven to rotate by a motor, a laser scanner scans the workpiece to generate a three-dimensional model, and a dynamic digital twin is constructed in combination with device parameters; the processing unit drives a spray head to dynamically adjust a position and an attitude under the cooperation of the motor through surface parameterization, trajectory planning and model predictive control, ensures an optimal spray printing distance, and a multi-sensor fusion and error monitoring mechanism can correct deviations in real time, and the model and the trajectory are automatically updated when an overrun time is reached. The application realizes accurate spray printing of special-shaped workpieces such as curved surfaces and concave surfaces, improves spraying precision and stability, widens application scenarios, and reduces the need for manual intervention.

Description

Technical Field

[0001] This invention relates to the field of spraying technology, and more specifically, to an automatic spraying device for inkjet printing equipment. Background Technology

[0002] In the field of advertising production, whether it's large outdoor billboards, promotional posters in shopping malls, or display backdrops at event venues, inkjet printing equipment can quickly and efficiently produce images, providing strong visual support for advertising campaigns. In the decoration and renovation industry, it can create various exquisite patterns and textures on surfaces such as walls, floors, and ceilings, meeting personalized and diverse decoration needs and enhancing the aesthetics and artistic atmosphere of spaces. Regarding product labeling, inkjet printing equipment can accurately print brand logos, product information, barcodes, and other content onto product surfaces, aiding in product identification, sales, and management.

[0003] The main component of an inkjet printer is the printhead, whose primary function is to precisely spray ink onto the printing medium to form the desired image and text. Typically, to achieve precise horizontal movement of the printhead, it is mounted on a specialized drive unit. This drive unit generally uses mechanical structures such as linear guides and lead screws, working in conjunction with stepper motors or servo motors. During printing, the printing medium, such as paper, board, or fabric, usually moves in two ways: one is relatively fixed, where the printing medium remains stationary while the printhead reciprocates horizontally along a preset path; the other is moving at a constant speed in a specific direction, in which case the printhead also moves accordingly horizontally. Simultaneously, based on image data transmitted from the computer, the timing and amount of ink ejection are precisely controlled. The image processing software in the computer digitizes the input image, breaking it down into individual pixels and generating corresponding control commands based on the color information of each pixel. Upon receiving these commands, the inkjet printer control system controls the printhead to eject the appropriate color and quantity of ink at the appropriate time, thus accurately copying the image from the computer onto the printing medium, completing the inkjet printing of the flat image.

[0004] Existing inkjet printing equipment has certain limitations when dealing with workpieces of special shapes. When working with convex, curved surfaces, the printhead can only move horizontally, and the distance between it and the printing surface constantly changes with the undulations of the curved surface. During the printhead's movement, when it encounters a protrusion, the distance between the printhead and the printing surface decreases sharply, increasing the risk of the printhead colliding with the protrusion. A collision can damage the printhead, disrupt the normal operation of the printing equipment, and may even cause printing interruptions, resulting in wasted time and materials. Conversely, when working with concave surfaces, the situation is quite different. During printing, because the concave surface is farther from the printhead, the ink is affected by air resistance, gravity, and other factors as it is ejected onto the printing medium, causing deviations in the ink's trajectory and preventing it from accurately reaching the intended position. This not only results in blurry or distorted images, affecting print quality, but may also lead to ink waste and increased printing costs. Summary of the Invention

[0005] To address the problems existing in the prior art, the purpose of this invention is to provide an automatic spraying device for inkjet printing equipment, which can achieve precise processing of irregularly shaped workpieces. By combining a multi-dimensional motion structure driven by a motor with laser scanning modeling technology, the device can perform all-round spraying on curved and concave workpieces such as cylindrical tubes, gourds, and spheres. The clamping component drives the workpiece to rotate and dynamically adjusts the horizontal and vertical directions of the nozzle, which can adapt to the changes in the surface of the workpiece in real time, thus solving the pain point of traditional equipment being unable to handle irregular surfaces.

[0006] To solve the above problems, the present invention adopts the following technical solution.

[0007] An automatic spraying device for inkjet printing equipment includes a worktable and a processing unit. A motor is positioned at the center of the upper end of the worktable, and the output shaft of the motor is connected to a clamping component for fixing the workpiece to be sprayed. A spraying assembly is positioned above the clamping component, comprising a printhead, a laser scanner, a third motor, and a fourth motor. The laser scanner is located below the printhead and is used to scan the workpiece surface. The third motor drives the printhead to move horizontally, and the fourth motor drives the printhead to move vertically. A lifting assembly is also installed on the worktable, including a second motor, for driving the entire spraying assembly to move vertically up and down. Ultrasonic sensors are located above and below the printhead to monitor the distance between the printhead and the workpiece.

[0008] The processing unit includes the following modules:

[0009] The modeling module is used to perform 3D scanning and point cloud data processing of the workpiece. It generates a 3D model with normal vectors through point cloud registration and surface reconstruction algorithms, and constructs a dynamic digital twin by combining the equipment dynamic parameters.

[0010] The planning module performs surface parameterization and distance field calculation based on the digital twin, and solves the optimal motion trajectory of the nozzle with the spraying process constraints and performance indicators as optimization objectives, and generates a multi-dimensional spraying plan including path, speed and attitude.

[0011] The optimization module adopts a model-based predictive control strategy, combined with a feedforward compensation mechanism, to generate motor control commands in real time based on the current system status and the spraying plan.

[0012] The correction module calculates the error between the actual state and the predicted state in real time through multi-sensor data fusion and state estimation algorithms, and feeds it back to the control system to achieve online parameter correction.

[0013] The update module is used to monitor system errors and trigger model updates and trajectory replanning when the error exceeds the limit, correcting the dynamic digital twin and control parameters.

[0014] Furthermore, the modeling module also includes:

[0015] The workpiece is scanned from all directions by a laser scanner to obtain raw point cloud data containing the spatial coordinate information of each point on the surface of the workpiece.

[0016] The original point cloud data is preprocessed by filtering out noise points, duplicate points and redundant data, and filling in missing data areas to obtain purified and complete point cloud data.

[0017] A point cloud registration algorithm is used to spatially align and fuse the multi-view point clouds in the purified and completed point cloud data, eliminate coordinate deviations caused by different scanning perspectives, and form a point cloud set that completely reflects the overall shape of the workpiece under a unified coordinate system.

[0018] Based on the point cloud set, a surface reconstruction algorithm is used to connect the discrete point cloud into a continuous surface, generating a preliminary three-dimensional surface model that can reflect the geometric features of the workpiece surface.

[0019] For the preliminary three-dimensional surface model, the normal vector of each surface point is calculated to determine its spatial orientation, forming a three-dimensional model with normal vector information;

[0020] Collect equipment dynamic parameters, combine the equipment dynamic parameters with the three-dimensional model with normal vector information, and construct a dynamic digital twin that can reflect the interaction state between the equipment and the workpiece in real time by mapping the dynamic relationship between the equipment motion and the three-dimensional model.

[0021] Furthermore, the planning module also includes:

[0022] Surface parameterization is performed on the 3D model with normal vectors to generate a quantifiable model containing the position of each point and the normal vector;

[0023] Based on this model, the distance distribution between each point on the workpiece surface and the movable space of the nozzle is calculated, the minimum, maximum and preset target distance range between the nozzle and the surface are determined, and distance field data is generated.

[0024] By combining distance field data, the constraints and performance indicators of the spraying process are clarified, and the constraints and optimization target parameters are formed;

[0025] Based on the surface parametric model, distance field data, and constraint and optimization parameters, a nozzle motion trajectory optimization problem is constructed to satisfy all process constraints and achieve optimal performance.

[0026] Solve the trajectory optimization problem to obtain a continuous motion path that covers the entire area to be sprayed and meets the preset target distance constraint;

[0027] Based on the motion path, speed limit, and surface complexity at each point along the path, the speed value of the nozzle along the path is planned, and speed planning data is generated.

[0028] Based on the surface normal vectors at each point along the path and the motion path, the nozzle posture is planned to match the spraying direction with the normal vectors, and posture planning data is generated.

[0029] By integrating motion path, speed, and attitude planning data, a multi-dimensional spraying plan is formed, including path coordinates, speed at each point, and attitude angle.

[0030] Furthermore, the optimization module also includes:

[0031] The system acquires the current status and spraying plan in real time and integrates them into a set of status and plan parameters.

[0032] Based on this set, the device motion model of the dynamic digital twin is invoked to simulate future motion trends and generate a future state prediction sequence containing the position, velocity trajectory and nozzle status of the driving components.

[0033] Based on the constraints of the spraying process, the predicted sequence of future states is compared with the planned target values ​​to analyze the magnitude and direction of the deviation and form a deviation assessment result.

[0034] By combining the deviation assessment results with the changing patterns of the planned targets, a feedforward compensation mechanism is used to calculate the compensation amount to offset the deviation and generate feedforward compensation parameters.

[0035] Using the predicted sequence, deviation results, and compensation parameters as inputs, a model predictive control strategy is adopted to adjust the control quantity of the drive component, generating preliminary drive control commands that can track the plan and meet the constraints.

[0036] Verify whether the initial command exceeds the physical limitations of the equipment. If it does, correct it based on the equipment's dynamic parameters to obtain the final drive control command that meets the capabilities.

[0037] The final command is sent to each drive component to execute the action, and the command and execution time are recorded for use in the next control cycle to obtain the system status.

[0038] Furthermore, the calibration module also includes:

[0039] Simultaneously collect real-time sensing data from multiple sensors to form a raw sensing data set;

[0040] The original sensor data set is preprocessed by removing outliers, smoothing high-frequency fluctuation signals, and unifying data timestamps to eliminate noise and interference in the data and obtain purified sensor data.

[0041] Based on the purified sensor data, a data fusion method is adopted, which combines the accuracy weights and spatiotemporal correlations of each sensor to complement and verify data from different sources, generating comprehensive actual state data that reflects the relative position of the nozzle and the workpiece, the actual shape of the workpiece, and the real-time motion state of the motor.

[0042] The predicted state data generated based on the digital twin in the control system is called up, and the comprehensive actual state data is compared with the predicted state data parameter by parameter;

[0043] Based on the comparison results, the position deviation, distance deviation, and velocity deviation between the actual state and the predicted state are calculated to determine the magnitude and direction of the error of each parameter and form an error vector.

[0044] By using a state estimation algorithm, combined with the error vector and the dynamic characteristics of the system, the root causes of the error are analyzed and the current true state of the system is accurately estimated, so as to obtain the corrected actual state and error assessment results.

[0045] Based on the error assessment results and the sensitivity of the equipment's control parameters, the required system parameter correction amount is calculated. The system parameter correction amount includes the correction value of the motor control parameters and the compensation amount of the nozzle movement trajectory.

[0046] The system parameter correction is fed back to the control system in real time to update the relevant parameters in the control algorithm, so that the control system can generate motor control commands based on the corrected parameters.

[0047] Furthermore, the update module also includes:

[0048] A system error threshold is pre-set. The error threshold is determined based on the accuracy requirements of the spraying process and the stability standards of equipment operation. It includes the maximum allowable values ​​of various errors such as position deviation, speed deviation, and shape fitting deviation, forming an error judgment benchmark.

[0049] The system collects various error data during operation in real time, including residual errors that still exist after online correction and newly emerging dynamic errors. The error data is continuously compared with the error judgment benchmark to generate error monitoring results that include the comparison relationship between the current error value and the threshold.

[0050] Based on the error monitoring results, determine whether the system error exceeds the error threshold; if it does, output an over-limit trigger signal.

[0051] In response to the over-limit trigger signal, the cause of the over-limit error is analyzed by combining the current state of the digital twin, the historical adjustment records of control parameters, and the equipment operation log, and an error root cause analysis report is generated.

[0052] Based on the error root cause analysis report, the digital twin update mechanism is triggered, the latest workpiece surface scanning data and equipment operating parameters are called, the three-dimensional model of the original digital twin is reconstructed and the dynamic parameters are corrected, and the updated digital twin is generated.

[0053] Furthermore, the update module also includes:

[0054] Based on the updated digital twin, combined with the current spraying progress and unfinished areas, the movement path, speed and attitude of the nozzle are replanned to generate a replanned spraying trajectory that adapts to the corrected workpiece model and equipment status.

[0055] Based on the replanned spraying trajectory and the updated digital twin, the adaptability of the original control parameters to the current system state is analyzed, the control parameter correction value is calculated, and a control parameter update scheme is formed.

[0056] The updated digital twin, the replanned spraying trajectory, and the updated control parameters are synchronized to each module of the system to replace the original data, and the error monitoring process is restarted to form a closed-loop control.

[0057] Furthermore, the lifting assembly also includes a support sleeve rod, which is fixedly installed on the side of the workbench. The second motor is installed on the upper end of the support sleeve rod, and a lead screw is rotatably installed inside the support sleeve rod. The output shaft of the second motor passes through the upper end of the support sleeve rod and is fixedly connected to the lead screw. The spraying assembly is mounted on the lead screw.

[0058] Furthermore, the spraying assembly also includes a lifting rod, which is mounted on a lead screw and threadedly connected to it. A motor is mounted on the upper end of the lifting rod, and a lead screw is fixedly connected to the output shaft of the motor. The end of the lead screw away from the motor is rotatably connected to the lifting rod. A moving block is threadedly connected to the lead screw, and a support plate is fixedly connected to one side of the moving block. A motor is fixedly mounted on the upper surface of the support plate, and the output shaft of the motor moves through the support plate and is fixedly connected to the lead screw. A lifting plate is threadedly connected to the lead screw, and a spray head is mounted on the side of the lifting plate near the clamping component. The ultrasonic sensor and the laser scanner are both fixedly connected to the lifting plate.

[0059] Furthermore, a limiting plate is fixedly connected to the lower end of the support plate, and a sliding groove is provided on the limiting plate. A slider is slidably arranged in the sliding groove, and one side of the slider is fixedly connected to the lifting plate; the clamping component is a three-jaw chuck.

[0060] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0061] (1) This solution breaks through the limitations of planar inkjet printing and realizes precise processing of irregular workpieces. By combining the multi-dimensional motion structure driven by the motor with laser scanning modeling technology, the equipment can perform all-round inkjet printing on curved and concave workpieces such as round tubes, gourds, and spheres. The clamping parts drive the workpiece to rotate and dynamically adjust the horizontal and vertical directions of the nozzle, which can adapt to the changes in the surface of the workpiece in real time. This solves the problem that traditional equipment cannot handle irregular surfaces and greatly expands the application scenarios of inkjet printing equipment.

[0062] (2) This solution uses dynamic digital twin and multi-sensor fusion technology to accurately construct a three-dimensional model of the workpiece and correct the nozzle trajectory in real time. The laser scanner and ultrasonic sensor work together to obtain workpiece shape data through pre-scanning and dynamically monitor position deviation during printing. Combined with model prediction control and feedforward compensation mechanism, it ensures that the nozzle always maintains the best distance, thereby ensuring the printing quality.

[0063] (3) Through error monitoring and dynamic correction mechanism, the equipment can autonomously cope with sudden situations such as workpiece loosening and displacement, automatically trigger model update and trajectory replanning, and maintain stable operation without manual intervention. The real-time iteration of digital twin and online optimization of control parameters enable the system to adapt to the inkjet printing needs of workpieces of different materials and shapes, reduce the dependence on operator skills, and improve efficiency and yield in mass production. Attached Figure Description

[0064] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0065] Figure 1 This is an external view of the overall structure of the present invention;

[0066] Figure 2 This is a schematic diagram of the lifting assembly of the present invention;

[0067] Figure 3 This is a schematic diagram of the structure of the spraying assembly of the present invention;

[0068] Figure 4 This is a schematic diagram of the structure of the lifting plate of the present invention;

[0069] Figure 5 This is a flowchart of an automatic spraying device for inkjet printing equipment.

[0070] Explanation of the labels in the diagram:

[0071] 1. Workbench; 2. Motor 1; 3. Clamping component; 4. Support sleeve; 5. Motor 2; 6. Lead screw 1; 7. Lifting rod; 8. Motor 3; 9. Lead screw 2; 10. Moving block; 11. Support plate; 12. Motor 4; 13. Lead screw 3; 14. Lifting plate; 15. Laser scanner; 16. Nozzle; 17. Limiting plate; 18. Slide groove; 19. Slider; 20. Ultrasonic sensor. Detailed Implementation

[0072] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0073] Please see Figures 1 to 5An automatic spraying device for inkjet printing equipment includes a worktable 1 and a processing unit. A motor 2 is located at the center of the upper end of the worktable 1. The output shaft of motor 2 is connected to a clamping member 3, which is used to fix the workpiece to be sprayed. A spraying assembly is located above the clamping member 3. This spraying assembly includes a nozzle 16, a laser scanner 15, a third motor 8, and a fourth motor 12. The laser scanner 15 is located below the nozzle 16 and is used to scan the surface of the workpiece. The third motor 8 drives the nozzle 16 to move horizontally, and the fourth motor 12 drives the nozzle 16 to move vertically. A lifting assembly is also installed on the worktable 1. This lifting assembly includes a second motor 5, which drives the entire spraying assembly to move vertically. Ultrasonic sensors 20 are located above and below the nozzle 16 to monitor the distance between the nozzle 16 and the workpiece.

[0074] The lifting assembly also includes a support sleeve 4, which is fixedly installed on the side of the workbench 1. A second motor 5 is installed on the upper end of the support sleeve 4. A first lead screw 6 is rotatably installed inside the support sleeve 4. The output shaft of the second motor 5 passes through the upper end of the support sleeve 4 and is fixedly connected to the first lead screw 6. The spraying assembly is set on the first lead screw 6.

[0075] The spraying assembly also includes a lifting rod 7, which is mounted on a lead screw 6 and the two are threaded together. A motor 8 is mounted on the upper end of the lifting rod 7. A lead screw 9 is fixedly connected to the output shaft of the motor 8. The end of the lead screw 9 away from the motor 8 is rotatably connected to the lifting rod 7. A moving block 10 is threadedly connected to the lead screw 9. A support plate 11 is fixedly connected to one side of the moving block 10. A motor 12 is fixedly mounted on the upper surface of the support plate 11. The output shaft of the motor 12 moves through the support plate 11 and is fixedly connected to a lead screw 13. A lifting plate 14 is threadedly connected to the lead screw 13. A spray nozzle 16 is mounted on the side of the lifting plate 14 near the clamping member 3. An ultrasonic sensor 20 and a laser scanner 15 are both fixedly connected to the lifting plate 14.

[0076] In use, the workpiece is first placed on the clamping member 3 to fix the bottom of the workpiece. Then, the motor 2 is turned on to drive the clamping member 3 to rotate at a constant speed, thereby causing the workpiece to rotate. During the rotation of the workpiece, the laser scanner 15 scans the outer surface of the workpiece and obtains the three-dimensional data of the outer surface of the workpiece based on the scanning results. When the printhead 16 is printing on the outer surface of the workpiece, the position of the printhead 16 can be dynamically adjusted in real time according to the obtained data. For example, when encountering a protruding part of the workpiece, the printhead 16 can be controlled to move away from the workpiece; when encountering a concave part of the workpiece, the printhead 16 can be moved closer to the workpiece, so that the printhead 16 and the workpiece are always at the optimal printing distance, thereby ensuring the printing effect and realizing the printing operation of irregularly shaped workpieces. During printing, motor 2 (5) drives lead screw 1 (6) to rotate, thereby moving lifting rod 7 up and down to position the printing assembly at a suitable height. Then, motor 3 (8) drives lead screw 2 (9) to rotate, thereby moving moving block 10. The movement of moving block 10 brings the nozzle 16 closer to the outer surface of the workpiece, stopping when the preset printing distance is reached. Then, motor 1 (2) drives the workpiece to rotate at a constant speed, and the nozzle 16 performs the printing operation. When the printing height needs to be adjusted, motor 4 (12) drives lead screw 3 (13) to rotate, thereby controlling the lifting plate 14 to move up and down, which in turn controls the up and down movement of the nozzle 16. During the printing process, if the distance the nozzle 16 moves up and down exceeds the stroke of lead screw 3 (13), motor 2 (5) can control lead screw 1 (6) to control the up and down movement of the printing assembly. During the inkjet printing process, the ultrasonic sensor 20 performs real-time detection on the outer surface of the workpiece. If the detection result does not match the three-dimensional data previously acquired by the laser scanner 15, the three-dimensional data is automatically corrected based on the data acquired by the ultrasonic sensor 20. This may be because the workpiece has become loose or shifted during the inkjet printing process. By adjusting the three-dimensional data in a timely manner, the quality of inkjet printing on the workpiece is ensured.

[0077] In some embodiments, the processing unit includes the following modules:

[0078] The modeling module is used to perform 3D scanning and point cloud data processing of the workpiece. It generates a 3D model with normal vectors through point cloud registration and surface reconstruction algorithms, and constructs a dynamic digital twin by combining the equipment dynamic parameters.

[0079] The planning module performs surface parameterization and distance field calculation based on the digital twin. With the constraints of the spraying process and performance indicators as optimization objectives, it solves the optimal motion trajectory of the nozzle 16 and generates a multi-dimensional spraying plan including path, speed and attitude.

[0080] The optimization module adopts a model-based predictive control strategy, combined with a feedforward compensation mechanism, to generate motor control commands in real time based on the current system status and the spraying plan.

[0081] The correction module calculates the error between the actual state and the predicted state in real time through multi-sensor data fusion and state estimation algorithms, and feeds it back to the control system to achieve online parameter correction.

[0082] The update module is used to monitor system errors and trigger model updates and trajectory replanning when the error exceeds the limit, correcting the dynamic digital twin and control parameters.

[0083] The modeling module also includes: scanning the workpiece from all angles using a laser scanner 15 to obtain raw point cloud data containing the spatial coordinate information of each point on the surface of the workpiece;

[0084] The raw point cloud data is preprocessed by filtering out noise points, duplicate points and redundant data, and filling in missing data areas to obtain purified and complete point cloud data.

[0085] A point cloud registration algorithm is used to spatially align and fuse the multi-view point clouds in the purified and completed point cloud data, eliminating coordinate deviations caused by different scanning perspectives and forming a point cloud set that fully reflects the overall shape of the workpiece under a unified coordinate system.

[0086] Based on point cloud sets, a surface reconstruction algorithm is used to connect discrete point clouds into continuous surfaces, generating a preliminary three-dimensional surface model that can reflect the geometric features of the workpiece surface.

[0087] For the initial three-dimensional surface model, the normal vector of each surface point is calculated to determine its spatial orientation, forming a three-dimensional model with normal vector information;

[0088] Collect equipment dynamic parameters, combine the equipment dynamic parameters with a 3D model with normal vector information, and construct a dynamic digital twin that can reflect the interaction state between the equipment and the workpiece in real time by mapping the dynamic relationship between the equipment motion and the 3D model.

[0089] By adopting the above technical solution, when the laser scanner 15 performs an all-around scan of the workpiece, it emits a laser beam towards the workpiece surface. After the laser beam contacts the workpiece surface, it reflects back to the scanner. The laser scanner 15 calculates the spatial coordinates of each point on the workpiece surface by recording the time difference or phase difference between laser emission and reception, combined with its own position information, and thus collects raw point cloud data containing this coordinate information. After obtaining the raw point cloud data, preprocessing is required. Due to environmental interference during the scanning process, such as changes in light, dust obstruction, or limitations in the accuracy of the equipment itself, the raw data may contain noise points, duplicate points, and redundant data. These invalid data are removed through filtering operations. At the same time, for data missing areas caused by workpiece surface obstruction or scanning angle limitations, the missing data is filled in according to the distribution pattern of surrounding valid points, finally obtaining purified and complete point cloud data.

[0090] In the point cloud registration stage, since the laser scanner 15 typically needs to scan the workpiece from multiple perspectives to obtain complete data, the point cloud data from different perspectives are in their own independent coordinate systems, resulting in spatial positional deviations. When using a point cloud registration algorithm, points with similar features are first found in the point clouds from different perspectives as matching points. Then, by adjusting the spatial positions of the point clouds from each perspective, the coordinates of the matching points are made to overlap as much as possible, thereby achieving spatial alignment and fusion of the point clouds from multiple perspectives, eliminating coordinate deviations, and forming a point cloud set that can completely reflect the overall shape of the workpiece under a unified coordinate system. This set provides complete discrete point information for surface reconstruction.

[0091] Surface reconstruction is performed based on the registered point cloud set. Discrete point clouds can only reflect the local position of the workpiece surface and cannot represent the continuous morphology of the surface. When using a surface reconstruction algorithm, adjacent points are connected by fitting the surface equation according to the distribution density and spatial relationship of the point cloud to form a continuous surface. For example, for gentle regions with small curvature, larger facets are used to connect adjacent points; for complex regions with large curvature, smaller facets are used to ensure accuracy, ultimately generating a preliminary 3D surface model that can represent the geometric features of the workpiece surface, such as concavity, convexity, and curvature. The normal vector of each surface point is calculated for the preliminary 3D surface model. The normal vector is a vector perpendicular to the tangent direction of the surface at that point, reflecting the orientation of the point in space. During calculation, the direction perpendicular to the surface at that point is determined by analyzing the surface morphology within a certain range around the surface point, thus obtaining the normal vector of each surface point. Combining these normal vectors with the preliminary 3D surface model forms a 3D model with normal vector information. This model provides geometric and orientation information of the workpiece for constructing a dynamic digital twin.

[0092] When constructing a dynamic digital twin, the dynamic parameters of the equipment are first collected, including the movement range of motors 1 (2), 2 (5), 3 (8), and 4 (12), such as the rotation angle range, movement distance range, speed limit, and the response time of nozzle 16. These parameters are then combined with a 3D model containing normal vector information. By establishing association rules between the equipment movement and the workpiece model, such as the correspondence between the movement distance of nozzle 16 and the change in coordinates of points on the workpiece surface, and the relationship between the workpiece rotation angle and the update of the model's viewpoint, the dynamic interaction state between the two is mapped. When the equipment moves, the digital twin can synchronously update the display state of the workpiece model; conversely, the movement requirements of the equipment can be inferred from the state of the workpiece model. Ultimately, a dynamic digital twin that can reflect the real-time interaction state between the equipment and the workpiece is constructed.

[0093] In some embodiments, the planning module further includes:

[0094] Surface parameterization is performed on the 3D model with normal vectors to generate a quantifiable model containing the position of each point and the normal vector;

[0095] Based on this model, the distance distribution between each point on the workpiece surface and the movable space of the nozzle 16 is calculated, the minimum, maximum and preset target distance range between the nozzle 16 and the surface are determined, and distance field data is generated.

[0096] By combining distance field data, the constraints and performance indicators of the spraying process are clarified, and the constraints and optimization target parameters are formed;

[0097] Based on the surface parameterization model, distance field data, and constraints and optimization parameters, a motion trajectory optimization problem for nozzle 16 is constructed to satisfy all process constraints and achieve optimal performance.

[0098] Solve the trajectory optimization problem to obtain a continuous motion path that covers the entire area to be sprayed and meets the preset target distance constraint;

[0099] Based on the motion path, speed limit, and surface complexity at each point along the path, the speed value of nozzle 16 along the path is planned, and speed planning data is generated.

[0100] Based on the surface normal vectors at each point along the path and the motion path, the attitude of the nozzle 16 is planned to make the spraying direction match the normal vector, and attitude planning data is generated.

[0101] By integrating motion path, speed, and attitude planning data, a multi-dimensional spraying plan is formed, including path coordinates, speed at each point, and attitude angle.

[0102] By employing the above technical solution, when parameterizing the surface of a 3D model with normal vectors, the complex 3D surface needs to be decomposed into a large number of regular small units, such as triangular or quadrilateral meshes, with each unit corresponding to a small region on the workpiece surface. Each unit is marked with its coordinate position in 3D space, and the normal vector information of the center point of that region is recorded. This transforms the 3D surface into a quantifiable model containing the coordinates of each point and the direction of the normal vector. Such a model can accurately reflect the geometry and orientation of each part of the workpiece surface, providing a structured data foundation for subsequent distance calculations.

[0103] When calculating distance field data based on the aforementioned quantifiable model, the position coordinates of each point in the model are used as a reference. Combined with the movement range of the nozzle 16, the movement limits of motor 3 (8), motor 4 (12), and motor 2 (5) are used to jointly determine the distance from each point to all spatial positions reachable by the nozzle 16. By analyzing these distances, the minimum distance between the nozzle 16 and each point on the workpiece surface, the closest distance at which the nozzle 16 will not collide with the workpiece, the maximum distance, and the preset target distance are determined. The preset target distance is the distance set in advance for the best spraying effect. This generates distance field data that reflects the distance distribution at different positions. This data intuitively presents the reasonable movement range of the nozzle 16 near various parts of the workpiece.

[0104] When defining spraying process constraints and performance indicators by combining distance field data, process constraints need to be determined based on the distance range in the distance field. These constraints include ensuring the distance between the nozzle 16 and the workpiece surface is within a preset target distance range, that the nozzle 16's movement speed does not exceed the motor's maximum speed limit, and that the movement path completely covers all unit areas to be sprayed. Performance indicators are set around spraying quality and efficiency, such as requiring the coating uniformity on the workpiece surface to reach a specific standard, controlling the entire spraying process time within a specified range, and ensuring the coating thickness deviation does not exceed the allowable value. These constraints and indicators are then translated into specific parameters, such as the upper speed limit, the coordinate range of the covered area, and the thickness deviation threshold, forming the constraint conditions and optimization target parameters.

[0105] When constructing the nozzle 16 motion trajectory optimization problem based on the surface parametric model, distance field data, and constraints and optimization parameters, the boundary of the area to be sprayed is determined based on the quantifiable model, the feasible motion space of the nozzle 16 is delineated according to the distance field data, and the allowable range of the trajectory is defined by the constraints. Simultaneously, performance indicators are used as the standard for evaluating the quality of the trajectory. In this way, a problem to be solved is: under the premise of satisfying all process constraints, find a path that enables the nozzle 16 motion trajectory to achieve optimal performance.

[0106] When solving the trajectory optimization problem, starting from the starting point of the workpiece's area to be coated, a path is planned within the feasible space defined by the distance field data, based on the positional distribution of each unit in the quantifiable model. The path must sequentially pass through all units to be coated, and each segment of the path must ensure that the distance between the nozzle 16 and the corresponding workpiece surface is within a preset target distance range. By continuously adjusting the path direction and avoiding areas exceeding the distance constraints, a continuous motion path that completely covers the entire area to be coated is ultimately formed.

[0107] When planning speed values ​​based on the motion path, the surface complexity of the workpiece corresponding to each point on the path is first analyzed. This is determined by the change in the normal vector of each element in the quantifiable model: areas with drastic changes in the normal vector belong to complex areas, such as parts of a surface with high curvature; areas with gentle changes in the normal vector belong to simple areas. Combined with speed limit parameters, such as for planar or gently curved surfaces, a lower motion speed is planned in complex areas to ensure that the nozzle 16 has enough time to accurately adjust its position and ensure uniform spraying; a higher motion speed is planned in simple areas to improve overall spraying efficiency. This generates speed planning data containing the speed values ​​of each point on the path.

[0108] When planning the attitude of the nozzle 16 based on the surface normal vectors at each point along the path and the motion path, the normal vectors of each point in the quantifiable model are used as a reference. If the spraying direction is required to be perpendicular to the workpiece surface, the axis direction of the nozzle 16 is adjusted to be in the same direction as the normal vector of that point. At the same time, combined with the direction of the motion path, it is ensured that the attitude change of the nozzle 16 during the movement is smooth and without abrupt changes, avoiding the impact of abrupt attitude changes on the spraying quality, and thus generating attitude planning data containing the attitude angles of each point.

[0109] When integrating motion path, speed, and attitude planning data, the coordinate information, corresponding speed values, and attitude angles of each point on the path are matched one by one to form a complete dataset. This dataset covers the position, speed, and orientation of the nozzle 16 at every moment during the spraying process, thus forming a multi-dimensional spraying plan that includes path coordinates, speeds at each point, and attitude angles.

[0110] In some embodiments, the optimization module further includes:

[0111] The system acquires the current status and spraying plan in real time and integrates them into a set of status and plan parameters.

[0112] Based on this set, the device motion model of the dynamic digital twin is invoked to simulate future motion trends and generate a future state prediction sequence containing the position, velocity trajectory and nozzle 16 states.

[0113] Based on the constraints of the spraying process, the predicted sequence of future states is compared with the planned target values ​​to analyze the magnitude and direction of the deviation and form a deviation assessment result.

[0114] By combining the deviation assessment results with the changing patterns of the planned targets, a feedforward compensation mechanism is used to calculate the compensation amount to offset the deviation and generate feedforward compensation parameters.

[0115] Using the predicted sequence, deviation results, and compensation parameters as inputs, a model predictive control strategy is adopted to adjust the control quantity of the drive component, generating preliminary drive control commands that can track the plan and meet the constraints.

[0116] Verify whether the initial command exceeds the physical limitations of the equipment. If it does, correct it based on the equipment's dynamic parameters to obtain the final drive control command that meets the capabilities.

[0117] The final command is sent to each drive component to execute the action, and the command and execution time are recorded for use in the next control cycle to obtain the system status.

[0118] By adopting the above technical solution, when acquiring the current status and spraying plan in real time, the system status data, such as the current position and rotation speed of motor 1 (2), motor 2 (5), motor 3 (8), and motor 4 (12), as well as the real-time distance between the nozzle 16 and the workpiece, are collected by the motor position sensor and ultrasonic sensor 20 on the equipment. The motor position sensor in this application is an encoder, which can acquire the rotation angle and speed of each motor; each motor is equipped with an encoder. The pre-generated spraying plan, including the target path of the nozzle 16, the target speed at each point, and the target posture, is retrieved from the control system. This real-time status data is integrated with the spraying plan data to form a set containing current status parameters and plan parameters. When the equipment motion model of the dynamic digital twin is called based on this parameter set, the model simulates the motion process in the near future based on the equipment's dynamic characteristics, such as the motor's motion response speed and the nozzle 16's inertia, combined with the current status parameters. For example, based on the current speed and position of the motor, its rotation angle and position changes in the future are predicted, and then the position, speed change trajectory of the nozzle 16 and the motion state of each driving component are calculated. Finally, a future state prediction sequence containing the position, speed trajectory and state of the driving component and the nozzle 16 is generated. This sequence reflects the expected result of the system moving according to the current trend.

[0119] When comparing the predicted future state sequence with the planned target value based on the spraying process constraints, the process constraints include the upper speed limit of the driving components, the optimal distance range between the nozzle 16 and the workpiece, and the allowable deviation of the nozzle 16's attitude. Each parameter in the predicted sequence is compared with the target value in the spraying plan (such as target path coordinates and target speed), the difference between the two is calculated, and the specific magnitude and direction of the deviation are analyzed to form a clear deviation assessment result. Combining the deviation assessment result with the changing patterns of the planned target, when calculating the compensation amount using a feedforward compensation mechanism, the changing patterns of the planned target refer to known trends such as speed adjustments at path corners and distance changes at concave and convex points on curved surfaces in the spraying plan. Based on the deviation assessment result, the types of deviations that may occur in the future are determined, such as the nozzle 16 being too close due to workpiece protrusion. Based on the changing patterns of the target, the compensation amount that needs to be adjusted is calculated in advance, such as controlling motor 38 to drive the nozzle 16 to move outward a certain distance to offset the predicted deviation. These compensation amounts are compiled into feedforward compensation parameters to correct the control quantities of the driving components.

[0120] Using the predicted sequence, deviation results, and compensation parameters as input, when adjusting the control quantities of the drive components using a model predictive control strategy, this strategy comprehensively analyzes the possible deviations in the future state prediction sequence, the specific deviations in the deviation evaluation results, and the adjustment amounts in the feedforward compensation parameters. By adjusting the control quantities of each drive component, the future motion state can accurately track the spraying plan. For example, if it is predicted that nozzle 16 will deviate from the path, the speed of the corresponding motor is increased to adjust the position of nozzle 16, while ensuring that the adjusted control quantities do not violate process constraints, ultimately generating preliminary drive control commands.

[0121] When verifying whether the initial command exceeds the physical limitations of the equipment, these physical limitations include the maximum torque and maximum rotation angle of the motor, and the movement range of the nozzle 16. The motion parameters of the drive components corresponding to the initial drive control command are compared with the physical limitation parameters. If the limits are not exceeded, the command is directly adopted; if the limits are exceeded, the command is modified based on the equipment's dynamic parameters, such as reducing the motor speed to decrease torque, until the command conforms to the equipment's physical capabilities, resulting in the final drive control command. When the final command is sent to each drive component for execution, the control system transmits the command to the corresponding motor through the circuit, driving the motor to rotate according to the command. For example, motor 2 drives the workpiece to rotate, and motor 8 drives the nozzle 16 to move horizontally, causing the nozzle 16 to move as planned. Simultaneously, the content and execution time of the command are automatically recorded. These records serve as reference data for obtaining the current system state in the next control cycle, helping the next collected state data to more accurately reflect the actual operating status of the equipment, forming a closed-loop control process.

[0122] In some embodiments, the calibration module further includes:

[0123] Simultaneously collect real-time sensing data from multiple sensors to form a raw sensing data set;

[0124] The raw sensor data set is preprocessed by removing outliers, smoothing high-frequency fluctuation signals, and unifying data timestamps to eliminate noise and interference in the data and obtain purified sensor data.

[0125] Based on the purified sensor data, a data fusion method is adopted, which combines the accuracy weights and spatiotemporal correlations of each sensor to complement and verify data from different sources, generating comprehensive actual state data that reflects the relative position of the nozzle 16 and the workpiece, the actual shape of the workpiece, and the real-time motion state of the motor.

[0126] The predicted state data generated based on the digital twin in the control system is called up, and the comprehensive actual state data is compared with the predicted state data parameter by parameter.

[0127] Based on the comparison results, the position deviation, distance deviation, and velocity deviation between the actual state and the predicted state are calculated to clarify the magnitude and direction of the error of each parameter and form an error vector.

[0128] By using a state estimation algorithm, combined with the error vector and the dynamic characteristics of the system, the root causes of the error are analyzed and the current true state of the system is accurately estimated, so as to obtain the corrected actual state and error assessment results.

[0129] Based on the error assessment results and the sensitivity of the equipment's control parameters, the required system parameter correction amounts are calculated. These system parameter correction amounts include the correction values ​​for the motor control parameters and the compensation amounts for the movement trajectory of the nozzle 16.

[0130] The system parameter correction is fed back to the control system in real time, updating the relevant parameters in the control algorithm, so that the control system can generate motor control commands based on the corrected parameters.

[0131] By adopting the above technical solution, when simultaneously acquiring real-time data from multiple sensors, the equipment will simultaneously activate all relevant sensors: the ultrasonic sensor 20 monitors the distance between the nozzle 16 and the workpiece in real time, the laser scanner 15 captures the morphological changes on the workpiece surface, and the position of the motors and the real-time motion status of each motor are monitored. This data from different sensors is aggregated in real time to form a raw sensor data set. The purpose of synchronous acquisition is to ensure that all data corresponds to the equipment status at the same point in time. Preprocessing the raw sensor data set is to eliminate interference factors in the data. By setting reasonable thresholds, outliers exceeding the normal range are filtered out; a smoothing algorithm is used to filter out high-frequency fluctuations in the data, making the signal more stable; and the timestamps of all data are unified to ensure strict alignment of data from different sensors in the time dimension. After these processes, purified sensor data is obtained.

[0132] When fusing data from purified sensors, different weights are assigned to each sensor based on their accuracy. For example, the morphological data from the laser scanner 15 has higher accuracy and a greater weight; the distance data from the ultrasonic sensor 20 is more reliable at close range and has a higher weight in the corresponding scenario. Simultaneously, spatiotemporal correlations are considered. For instance, at any given moment, the position change of the nozzle 16 should match the motor motion data and distance data. Data from different sources is cross-validated: for example, the morphological data from the laser scanner 15 is used to verify the reasonableness of the distance data from the ultrasonic sensor 20, and the position of the nozzle 16 calculated using motor motion data is complementary and corrected to the actual measured position. Finally, comprehensive actual state data is generated, which more comprehensively and accurately reflects the relative position of the nozzle 16 and the workpiece, the actual shape of the workpiece, and the real-time motion state of the motor.

[0133] The system retrieves predicted state data generated from the digital twin in the control system, including the predicted position of nozzle 16, distance from the workpiece, and motor speed, and compares this data parameter by parameter with the obtained comprehensive actual state data. For example, the predicted X-axis coordinate of nozzle 16 is compared with the actual measured X-axis coordinate, and the predicted motor speed is compared with the actual speed, ensuring that each key parameter can be analyzed individually, providing a direct basis for subsequent error calculation.

[0134] When calculating errors based on the comparison results, the difference between the actual value and the predicted value is calculated for each parameter: position deviation reflects the degree of deviation between the actual position of the nozzle 16 and the predicted path; distance deviation reflects the difference between the actual distance between the nozzle 16 and the workpiece and the predicted distance; speed deviation shows the difference between the actual speed of the motor and the predicted speed. At the same time, the direction of each deviation is specified, such as position to the left, distance to the right, and speed to the left. These deviations are integrated into an error vector that includes both magnitude and direction, intuitively presenting the deviation of various aspects of the system.

[0135] When applying the state estimation algorithm, the error vector and system dynamic characteristics are considered. For example, the inertia of the equipment causes a delay in changes in motor speed, and the movement of nozzle 16 is restricted by the mechanical structure, resulting in a response lag. The root causes of the errors are analyzed: if the distance deviation continues to increase, it may be due to drift of the ultrasonic sensor 20; if the position deviation increases with the increase of motor speed, it may be due to insufficient motor response speed. Based on these analyses, a more accurate estimate of the current true state of the system is made, such as the actual distance after excluding the influence of sensor drift. Finally, the corrected actual state and error assessment results including error type and severity are obtained.

[0136] When calculating the system parameter correction, it is necessary to combine the error assessment results and the sensitivity of the equipment's control parameters, such as the impact of small changes in the motor speed regulation coefficient on the actual rotational speed and the sensitivity of the trajectory tracking threshold to position deviation. For example, if the speed deviation is large and the motor is sensitive to the regulation coefficient, the correction value of the speed regulation coefficient should be appropriately increased; if the position deviation originates from trajectory planning, the compensation amount for the additional movement of the nozzle 16 should be calculated, such as a fine adjustment of 0.3 mm along the positive X-axis. Finally, the correction values ​​of the motor control parameters and the compensation amount of the nozzle 16's motion trajectory are determined to form the system parameter correction. After the system parameter correction is fed back to the control system in real time, the control system will immediately update the relevant parameters in the control algorithm, such as replacing the original motor speed regulation coefficient and updating the deviation threshold of trajectory tracking. These corrected parameters will play a role when generating motor control commands subsequently: for example, the adjusted speed regulation coefficient enables the motor to reach the target rotational speed more accurately, and the trajectory compensation amount enables the nozzle 16 path to better conform to the actual workpiece shape, thereby reducing the error between the actual state and the predicted state and improving the spraying accuracy.

[0137] In some embodiments, the update module further includes:

[0138] A system error threshold is pre-set. The error threshold is determined based on the accuracy requirements of the spraying process and the stability standards of equipment operation. It includes the maximum allowable values ​​of various errors such as position deviation, speed deviation, and shape fitting deviation, forming an error judgment benchmark.

[0139] The system collects various error data in real time during operation, including residual errors that still exist after online correction and newly emerging dynamic errors. It continuously compares the error data with the error judgment benchmark and generates error monitoring results that include the comparison relationship between the current error value and the threshold.

[0140] Based on the error monitoring results, determine whether the system error exceeds the error threshold. If it does, output an over-limit trigger signal.

[0141] In response to the over-limit trigger signal, the cause of the over-limit error is analyzed by combining the current status of the digital twin, the historical adjustment records of control parameters, and the equipment operation log, and an error root cause analysis report is generated.

[0142] Based on the error root cause analysis report, the digital twin update mechanism is triggered, the latest workpiece surface scanning data and equipment operating parameters are called, the three-dimensional model of the original digital twin is reconstructed and the dynamic parameters are corrected, and the updated digital twin is generated.

[0143] Based on the updated digital twin, combined with the current spraying progress and unfinished areas, the motion path, speed and attitude of the nozzle 16 are replanned to generate a replanned spraying trajectory that adapts to the corrected workpiece model and equipment status.

[0144] Based on the replanned spraying trajectory and the updated digital twin, the adaptability of the original control parameters to the current system state is analyzed, the control parameter correction values ​​are calculated, and a control parameter update scheme is formed.

[0145] The updated digital twin, replanned spray trajectory, and updated control parameters are synchronized to all modules of the system to replace the original data, and the error monitoring process is restarted to form a closed-loop control.

[0146] By adopting the above technical solution, when setting the system error threshold, it is necessary to combine the core precision requirements of the spraying process and the equipment operation stability standards to determine the maximum allowable value for various key errors such as position deviation, speed deviation, and shape fitting deviation. These allowable values ​​are integrated into a clear error judgment benchmark. When collecting various error data in real time during system operation, it includes both residual errors that have not been eliminated after online correction, such as the 0.3 mm position deviation that still exists after multiple parameter adjustments, and new dynamic errors that appear during the spraying process, such as the rotational offset caused by slight loosening of the workpiece and the motion lag of the nozzle 16 due to mechanical wear. The collected error data will be continuously compared with the set error judgment benchmark. For example, the current 0.6 mm position deviation will be compared with the maximum allowable value of 0.5 mm in the benchmark. Finally, an error monitoring result containing information such as the current error value, the corresponding threshold, whether it is close to the threshold, and whether it has exceeded the threshold will be generated. The system error is judged based on the error monitoring result. If all errors are within the threshold range, monitoring will continue; if any error exceeds the corresponding threshold, an over-limit trigger signal will be output immediately, which will start the subsequent model update and trajectory replanning process.

[0147] After responding to the over-limit trigger signal, the root cause of the error needs to be analyzed. This involves retrieving the current state of the digital twin, such as whether there is a significant deviation between the model and the actual workpiece shape, historical adjustment records of control parameters, and equipment operation logs, including whether there is abnormal noise from the motor or data jumps from sensors. Cross-analysis is used to pinpoint the cause: if the digital twin model deviates significantly from the latest scanned workpiece shape, it may be due to insufficient initial modeling accuracy; if the motor speed deviation continues to increase, it may be due to motor aging leading to decreased response performance. This process ultimately generates an error root cause analysis report containing the error type, specific cause, and scope of impact. Based on the error root cause analysis report, the digital twin update mechanism is triggered. If the error originates from the deviation between the model and the actual workpiece, the laser scanner 15 is used to perform a supplementary scan of the workpiece to obtain the latest surface morphology data; if the error is related to equipment performance, the latest motion parameters of each motor are collected, such as the actual rotational speed and the response delay of control commands. Using this new data, the 3D model in the original digital twin is reconstructed, such as correcting the coordinates of protruding parts of the workpiece and correcting equipment dynamic parameters, such as updating the motor speed response coefficient. Finally, an updated digital twin that more accurately reflects the actual situation is generated. When replanning the trajectory based on the updated digital twin, the current spraying progress and the geometric features of the unfinished areas are first determined. Combining the precise shape of the workpiece in the new twin and the latest dynamic characteristics of the equipment, the motion path, speed, and attitude of the nozzle 16 are recalculated to generate a replanned spraying trajectory that can adapt to the corrected workpiece model and equipment state.

[0148] Based on the replanned spraying trajectory and the updated digital twin, the control parameters are adjusted, and the compatibility between the original control parameters and the current system state is analyzed. If there are more sharp bends in the new trajectory, the original speed adjustment coefficient may cause the motor to decelerate too slowly, requiring an increase in the coefficient to improve response sensitivity. If the optimal distance between nozzle 16 and the workpiece changes due to the model update, the distance threshold for trajectory tracking needs to be corrected. Through this type of analysis, specific control parameter correction values ​​are calculated, forming a control parameter update scheme that includes parameter names, original values, new values, and reasons for adjustment. The updated digital twin, the replanned spraying trajectory, and the updated control parameter scheme are synchronized to all relevant modules of the system, including the modeling module, planning module, and optimization module, replacing the original model, trajectory, and parameters. After synchronization, the error monitoring process is restarted, continuously monitoring errors based on the new system state to ensure stable operation of the corrected system, forming a closed-loop control of "monitoring-judgment-correction-re-monitoring," thereby ensuring long-term spraying accuracy and equipment stability.

[0149] In some embodiments, the lower end of the support plate 11 is fixedly connected to a limiting plate 17, the limiting plate 17 is provided with a sliding groove 18, a slider 19 is slidably disposed in the sliding groove 18, and one side of the slider 19 is fixedly connected to the lifting plate 14; the clamping member 3 is a three-jaw chuck.

[0150] During operation, the lead screw 13 rotates, and the lifting plate 14 rotates accordingly. The sliding block 19, in cooperation with the slide groove 18, restricts the rotation of the lifting plate 14 with the lead screw 13, thus ensuring stable up-and-down movement of the lifting plate 14. The clamping component 3 uses a three-jaw chuck, a technology already available, which allows for the fixing and disassembly of the workpiece.

[0151] The above are merely preferred embodiments of the present invention; however, the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. An automatic spraying device for a plotter printing apparatus, comprising a worktable (1) and a processing unit, characterized in that: The upper end center position of the workbench (1) is provided with a motor one (2), the output shaft of the motor one (2) is connected with a clamping piece (3), the clamping piece (3) is used for fixing the workpiece to be sprayed;The clamping piece (3) is provided with a spraying assembly above, the spraying assembly includes a spray head (16), a laser scanner (15), a motor three (8) and a motor four (12), the laser scanner (15) is located below the spray head (16), and is used for scanning the surface of the workpiece;The motor three (8) is used for driving the spray head (16) to move along the horizontal direction, and the motor four (12) is used for driving the spray head (16) to move along the vertical direction;The workbench (1) is also provided with a lifting assembly, and the lifting assembly includes a motor two (5), which is used for driving the whole spraying assembly to lift along the vertical direction;The spray head (16) is provided with an ultrasonic sensor (20) above and below, which is used for monitoring the distance between the spray head (16) and the workpiece; The processing unit comprises the following modules: A modeling module is configured to acquire original point cloud data containing spatial coordinate information of each point on the surface of the workpiece by comprehensively scanning the workpiece through the laser scanner (15). The original point cloud data is preprocessed to remove noise points, repeated points and redundant data therefrom by filtering, and to fill in missing data areas, so as to obtain purified and completed point cloud data. A point cloud registration algorithm is used to align and fuse multi-view point clouds in the purified and completed point cloud data in spatial positions, so as to eliminate coordinate deviations caused by different scanning views and form a point cloud set that completely reflects the overall shape of the workpiece under a unified coordinate system. Based on the point cloud set, a surface reconstruction algorithm is used to connect discrete point clouds into continuous surfaces, so as to generate a preliminary three-dimensional surface model that can reflect the geometric features of the surface of the workpiece. For the preliminary three-dimensional surface model, the normal vector of each surface point is calculated to determine its spatial orientation, so as to form a three-dimensional model with normal vector information. Device dynamics parameters are collected, the device dynamics parameters are combined with the three-dimensional model with normal vector information, a dynamic association between device motion and the three-dimensional model is mapped, and a dynamic digital twin that can reflect the interactive state of the device and the workpiece in real time is constructed. A planning module is configured to parameterize the three-dimensional model with normal vector to generate a quantifiable model containing the position and normal vector of each point. Based on the model, the distance distribution between each point on the surface of the workpiece and the movable space of the spray head (16) is calculated, the minimum, maximum and preset target distance range of the spray head (16) and the surface are determined, and distance field data is generated. In combination with the distance field data, spraying process constraints and performance indicators are determined, and constraint conditions and optimization target parameters are formed. Based on the surface parameterized model, the distance field data and the constraint and optimization parameters, a trajectory optimization problem of the spray head (16) is constructed to meet all process constraints and achieve optimal performance. The trajectory optimization problem is solved to obtain a continuous motion path that covers all the areas to be sprayed and meets the preset target distance constraints. According to the motion path, speed limit and surface complexity of each point on the path, the speed value of the spray head (16) along the path is planned, and speed planning data is generated. According to the surface normal vector of each point on the path and the motion path, the attitude of the spray head (16) is planned to adapt the spraying direction to the normal vector, and attitude planning data is generated; The motion path, speed and attitude planning data are integrated to form a multi-dimensional spraying plan containing path coordinates, point speeds and attitude angles; The optimization module uses a model-based predictive control strategy combined with a feedforward compensation mechanism to generate motor control instructions in real time based on the current state of the system and the spraying plan; The correction module calculates the error between the actual state and the predicted state in real time through multi-sensor data fusion and state estimation algorithm, and feeds back to the control system to realize online parameter correction; The update module is used to monitor system errors and trigger model updating and trajectory re-planning when the error exceeds the limit to correct the dynamic digital twin and control parameters.

2. An automatic spraying device for a spray printing apparatus according to claim 1, wherein The optimization module further comprises: The current state of the system and the spraying plan are integrated into a state and plan parameter set in real time, The device motion model of the dynamic digital twin is called based on the set to simulate future motion trends and generate a future state prediction sequence containing the position, speed trajectory of the driving component and the state of the spray head (16); According to the spraying process constraints, the future state prediction sequence and the planned target value are compared to analyze the size and direction of the deviation, and a deviation evaluation result is formed; Combined with the deviation evaluation result and the change law of the planned target, a feedforward compensation mechanism is used to calculate the compensation amount to offset the deviation and generate feedforward compensation parameters; Using the prediction sequence, deviation result and compensation parameters as input, a model predictive control strategy is used to adjust the driving component control amount to generate preliminary driving control instructions that can track the plan and meet the constraints; Verify whether the preliminary instructions exceed the physical limits of the device. If they do, modify them based on the device dynamics parameters to obtain final driving control instructions that meet the capabilities; The final instructions are sent to each driving component to perform actions, and the instructions and execution time are recorded for use by the next control cycle to obtain the system state.

3. An automatic spraying device for a plotter printing apparatus according to claim 2, wherein The correction module further comprises: Synchronously collect real-time sensing data from multiple sensors to form a set of raw sensing data; Pretreat the raw sensing data set by removing outliers, smoothing high-frequency fluctuation signals, and unifying data timestamps to eliminate noise and interference in the data and obtain purified sensor data; Based on the purified sensor data, a data fusion method is used to complement and verify data from different sources based on the accuracy weights and spatiotemporal correlation of each sensor to generate comprehensive actual state data reflecting the relative position of the spray head (16) and the workpiece, the actual shape of the workpiece, and the real-time motion state of the motor; Call the prediction state data generated by the digital twin in the control system and compare the comprehensive actual state data with the prediction state data parameter by parameter; According to the comparison result, calculate the position deviation, distance deviation and speed deviation between the actual state and the predicted state to determine the error size and direction of each parameter and form an error vector; Using a state estimation algorithm, analyze the error source and accurately estimate the current true state of the system based on the error vector and system dynamic characteristics to obtain the corrected actual state and error evaluation result; According to the error evaluation result, combined with the control parameter sensitivity of the equipment, the system parameter correction amount to be adjusted is calculated, the system parameter correction amount includes the correction value of the motor control parameter and the compensation amount of the nozzle (16) motion trajectory; The system parameter correction amount is fed back to the control system in real time, and the related parameters in the control algorithm are updated, so that the control system generates motor control instructions based on the corrected parameters.

4. An automatic spraying device for a plotter printing apparatus according to claim 3, wherein The updating module further comprises: The system error threshold is set in advance, which is determined based on the spraying process precision requirement and the equipment operation stability standard, and includes the maximum allowable value of various errors such as position deviation, speed deviation and shape fitting deviation, forming an error judgment benchmark; Real-time acquisition of various error data in system operation, including residual error after online correction and newly occurring dynamic error, continuous comparison of the error data with the error judgment benchmark, generation of error monitoring results including current error value and threshold comparison relationship; Based on the error monitoring results, it is judged whether the system error exceeds the error threshold, and if it exceeds, an out-of-limit trigger signal is outputted; In response to the out-of-limit trigger signal, the causes of the out-of-limit error are analyzed based on the current state of the digital twin, the historical adjustment record of the control parameters and the equipment operation log, and an error root cause analysis report is formed; According to the error root cause analysis report, the digital twin updating mechanism is triggered, the latest workpiece surface scanning data and equipment operation parameters are called, the original digital twin is reconstructed and the kinetic parameters are corrected, and the updated digital twin is generated.

5. An automatic spraying device for a plotter printing apparatus according to claim 4, wherein The updating module further comprises: Based on the updated digital twin, combined with the current spraying progress and the unfinished area, the motion path, speed and attitude of the nozzle (16) are re-planned, and the re-planned spraying trajectory adapted to the corrected workpiece model and equipment state is generated; According to the re-planned spraying trajectory and the updated digital twin, the adaptability of the original control parameters to the current system state is analyzed, the control parameter correction value is calculated, and the control parameter updating scheme is formed; The updated digital twin, re-planned spraying trajectory and control parameter updating scheme are synchronized to each module of the system to replace the original data, and the error monitoring process is restarted to form a closed-loop control.

6. An automatic spraying device for a spray printing apparatus according to claim 1, characterized in that: The lifting assembly further comprises a support sleeve rod (4) fixedly installed at the side edge position of the workbench (1), the motor two (5) is installed at the upper end of the support sleeve rod (4), the screw rod one (6) is rotatably installed in the support sleeve rod (4), the output shaft of the motor two (5) is movably penetrated through the upper end of the support sleeve rod (4) and fixedly connected with the screw rod one (6), and the spraying assembly is arranged on the screw rod one (6).

7. An automatic spraying device for a plotter printing apparatus according to claim 6, characterized in that: The spraying assembly further comprises a lifting rod (7) provided on the screw rod (6) and threadedly connected with the screw rod (6), a motor (8) installed on the upper end of the lifting rod (7), a screw rod (9) fixedly connected on the output shaft of the motor (8), the screw rod (9) being rotatably connected with the lifting rod (7) at the end away from the motor (8), a moving block (10) threadedly and drivingly connected on the screw rod (9), a support plate (11) fixedly connected on one side of the moving block (10), a motor (12) fixedly installed on the upper surface of the support plate (11), an output shaft of the motor (12) movably penetrating through the support plate (11) and fixedly connected with a screw rod (13), a lifting plate (14) threadedly and drivingly connected on the screw rod (13), a nozzle (16) installed on one side of the lifting plate (14) close to the clamping piece (3), and the ultrasonic sensor (20) and the laser scanner (15) being fixedly connected with the lifting plate (14).

8. An automatic spraying device for a plotter printing apparatus according to claim 7, characterized in that: The lower end of the support plate (11) is fixedly connected with a limiting plate (17), the limiting plate (17) is provided with a sliding groove (18), the sliding groove (18) is slidably provided with a sliding block (19), and one side of the sliding block (19) is fixedly connected with the lifting plate (14); the clamping piece (3) is a three-jaw chuck.

Images