Workpiece full-coverage spraying path planning method and system based on model driving and visual guidance, and storage medium

Abstract

The application is suitable for the field of robot automatic spraying technology, and provides a workpiece full-coverage spraying path planning method and system based on model driving and visual guidance and a storage medium, the method comprising: acquiring a three-dimensional model of a target workpiece; generating an initial spraying path covering the surface of the workpiece based on the three-dimensional model, the path avoiding surrounding obstacles; controlling a spraying robot to perform a spraying operation along the initial path and collecting actual spraying data of the surface of the workpiece in real time; comparing the actual spraying data with the three-dimensional model, calculating an actual spraying coverage rate and identifying an uncovered area; generating a supplementary spraying path according to the uncovered area; and controlling the spraying robot to perform a supplementary spraying operation along the supplementary spraying path.The application realizes closed-loop control of full-coverage spraying by real-time detection and evaluation after spraying, automatic identification of missed spraying areas and generation of a supplementary spraying path, solves the problems of inaccurate identification of missed spraying areas and low efficiency of supplementary spraying in the prior art, and is suitable for multi-process automatic spraying of large metal structural parts.

Description

Technical Field

[0001] This invention belongs to the field of automated anti-corrosion spraying technology, and is particularly applicable to automated robotic spraying of sandblasting, zinc spraying, painting and powder coating of hydraulic metal structural components, building machinery structural components and building steel structural components. Specifically, it relates to a workpiece full-coverage spraying path planning method, system and storage medium based on model-driven and vision-guided methods. Background Technology

[0002] With the continuous development of industrial automation, painting robots have been widely used in surface coating operations for various workpieces. In the field of corrosion protection for large metal structural components, such as hydraulic metal structural components, construction machinery structural components, and building steel structural components, which serve in complex corrosive environments such as outdoor and aquatic environments for extended periods, it is essential to ensure their service life through sandblasting for rust removal, zinc spraying, and multi-layer painting or powder coating composite anti-corrosion processes. The widespread adoption of automated robotic painting to replace traditional manual painting in existing anti-corrosion production lines has become an industry trend.

[0003] Currently, robotic painting operations typically employ teach-in programming or offline programming to plan the painting path. However, existing technologies have the following shortcomings: First, the teaching programming method requires the operator to hold a teaching pendant to guide the robot to record the path point by point. For non-standard large and complex workpieces, the teaching process is time-consuming and inefficient, and it is difficult to guarantee the smoothness and consistency of the spraying trajectory.

[0004] Secondly, while offline programming allows for pre-planning of paths on a computer, it cannot accommodate deviations between the actual placement of the workpiece and the theoretical model. When there are positioning errors on the workpiece in the fixture, a misalignment will occur between the pre-planned path and the actual workpiece surface, resulting in unstable coating quality.

[0005] Third, most existing spraying path planning methods are based on the theoretical three-dimensional model of the workpiece to generate the path, without fully considering the limitations of obstacles around the workpiece (such as the brackets that support the workpiece, fixed equipment in the spraying booth, etc.) on the robot's movement. This can easily lead to interference or collision between the robot body and obstacles, posing safety hazards.

[0006] Fourth, existing technologies lack effective means to detect coating coverage. In actual spraying operations, due to the complex shape of the workpiece and the limited accessibility of the robot, there are often areas on the workpiece surface that have not been sprayed (such as dead corners and obstructed surfaces). These uncovered areas are difficult to detect and re-spray in a timely manner, affecting the overall anti-corrosion quality of the workpiece.

[0007] Fifth, when the spraying operation is interrupted due to abnormal conditions such as malfunction or power outage, the existing system usually needs to be re-taught or re-planned, and cannot continue the operation from the interrupted position, resulting in a waste of time and materials.

[0008] Therefore, there is an urgent need for an intelligent spraying path planning method that can automatically plan the spraying path, detect the spraying coverage in real time, and automatically generate the touch-up spraying path, so as to improve the automation level and quality stability of the spraying operation. Summary of the Invention

[0009] The purpose of this invention is to provide a model-driven and vision-guided method, system, and storage medium for full-coverage spraying path planning of workpieces, aiming to solve the following technical problems: 1) Solve the problem that missed areas cannot be accurately identified in the existing spraying robot path planning due to the lack of real-time detection and evaluation after spraying; 2) Solve the problem of low efficiency in existing technologies that require manual intervention or complete replanning of the path after detecting missed spraying; 3) Solve the problem of lack of a unified model reference and positional deviation transmission mechanism between processes in the multi-process spraying of large workpieces; 4) Solve the problem that existing painting robots need to be re-taught or re-planned after the painting operation is interrupted due to abnormality.

[0010] 5) Solve the problem of excessive reliance on manual labor, low efficiency, and unstable quality in the existing non-standard hydraulic metal structure spraying.

[0011] This invention is implemented as follows: a workpiece full-coverage spraying path planning method based on model-driven and vision-guided methods, the workpiece full-coverage spraying path planning method comprising the following steps: S1: Obtain the 3D model of the target workpiece; S2: Based on the three-dimensional model, an initial spraying path is generated to cover the surface of the target workpiece, and the initial spraying path avoids the target workpiece and surrounding obstacles; S3: Control the spraying robot to perform spraying operations on the target workpiece along the initial spraying path, and collect the actual spraying data on the surface of the target workpiece in real time; S4: Compare the actual spraying data with the three-dimensional model, calculate the actual spraying coverage of the target workpiece surface, and identify the uncovered areas on the target workpiece surface that were not sprayed. S5: Generate a respray path based on the uncovered area; S6: Control the painting robot to perform a touch-up spraying operation on the uncovered area along the touch-up spraying path.

[0012] This application achieves fully automated closed-loop control from model acquisition, path generation, spraying execution, coverage detection, to respraying. Specifically, step S1, acquiring the 3D model, provides a precise workpiece geometric reference for subsequent path planning; step S2 automatically generates an initial spraying path based on the 3D model to avoid obstacles, ensuring the safety and efficiency of the spraying operation; step S3, by collecting spraying data in real time, provides a basis for coverage determination; step S4 accurately identifies uncovered areas through data comparison, solving the problem of not being able to detect spraying defects in a timely manner in existing technologies; steps S5 and S6 automatically generate respraying paths and execute respraying, ensuring full coverage spraying of the workpiece surface and significantly improving spraying quality.

[0013] A further technical solution of the present invention is: obtaining the three-dimensional model of the target workpiece in step S1 includes the following steps: S11. Import the CAD design model of the target workpiece into the system as the constructed three-dimensional model; and / or, scan the target workpiece using a three-dimensional vision scanning device to obtain the actual point cloud data of the target workpiece, and filter the actual point cloud data to remove noise and outliers. S12. Register and fuse the filtered point cloud data to generate a complete surface cloud model of the target workpiece. S13. Perform triangular meshing on the surface point cloud model to generate the three-dimensional model.

[0014] This application provides two flexible methods for acquiring 3D models: the CAD model import method is suitable for workpieces with known design data and is highly efficient; the visual scanning method is suitable for situations where there is no design model or the model differs from the actual workpiece. Through filtering, registration and fusion, and meshing, a high-precision 3D model can be constructed, providing an accurate data foundation for subsequent path planning.

[0015] A further technical solution of the present invention is: when the three-dimensional model is constructed by visual scanning, the following steps are included before step S2: S201. The target workpiece is located in actual position using a laser three-dimensional vision scanning device, and the deviation between the actual placement position of the target workpiece and the theoretical position of the three-dimensional model is obtained. S202. Perform position calibration compensation on the three-dimensional model based on the deviation; The step S2, which generates the initial spraying path, also includes the following steps: S21. Based on the three-dimensional model, the spraying path program for the target workpiece is automatically generated using offline programming software; or, the target workpiece is scanned by laser three-dimensional vision to form a point cloud model, and the spraying path program is automatically generated based on the point cloud model. S22. Identify obstacle information around the target workpiece, the obstacles including a bracket that carries the target workpiece and fixed equipment in the environment where the target workpiece is located; S23. The acquired obstacle information and the spraying path program are combined to generate the latest initial spraying path, so that the initial spraying path avoids the obstacle and prevents the spraying robot from interfering with or colliding with the obstacle.

[0016] By employing visual positioning and position calibration compensation, the deviation between the actual placement of the workpiece and the theoretical model is eliminated, ensuring consistency between the spraying path and the actual position of the workpiece. Simultaneously, by identifying obstacle information and generating avoidance paths, collisions between the robot and brackets, fixed equipment, etc., are effectively prevented, improving the safety of the spraying operation.

[0017] The automatic generation program for spraying paths from the point cloud model includes the following steps: S211. Scan the target workpiece using a laser three-dimensional vision scanning device to obtain the surface point cloud data of the target workpiece; S212. Preprocess the surface point cloud data to generate a three-dimensional point cloud model of the target workpiece; S213. Extract surface features from the three-dimensional point cloud model to identify the surface area to be coated on the target workpiece; S214. Based on the geometric features of the surface area to be sprayed and the kinematic parameters of the spraying robot, automatically generate a spraying path program that covers the surface area to be sprayed.

[0018] It enables the automatic generation of spraying paths from point cloud data without manual intervention, and is applicable to workpieces with various complex shapes, greatly improving the efficiency and adaptability of path planning.

[0019] Identifying obstacle information around the target workpiece includes the following steps: S221. Obtain a three-dimensional environment model of the environment in which the target workpiece is located; S222. Mark the position and outline information of the bracket and the fixing device in the three-dimensional environment model; S223. Use the position and outline information of the bracket and the fixing device as the obstacle information; The system can accurately identify all obstacles around the workpiece, providing complete environmental information for path avoidance.

[0020] The initial spraying path avoids the obstacle, which includes the following steps: S231. Calculate the envelope space of the spraying robot at each path point of the initial spraying path; S232. Determine whether the envelope space spatially interferes with any obstacle in the obstacle information; S233. If so, adjust the position or orientation of the path point until the envelope space does not interfere with any obstacles.

[0021] By employing the envelope space interference detection method, it is possible to accurately determine whether there is a collision risk at each path point of the robot, and ensure that there is no collision throughout the entire process by adjusting the position or attitude of the path points, thereby improving the safety and reliability of path planning.

[0022] A further technical solution of the present invention is: the following steps are included before step S3: S301. Based on the material of the target workpiece and the process requirements of the spraying operation, call the corresponding spraying process parameter set from the preset process parameter database. S302. The spraying process parameter set is sent to the spraying robot, wherein the spraying process parameter set includes at least one or more of spraying pressure, spraying flow rate, spraying speed and spraying distance; It enables automatic matching and recall of spraying process parameters, ensuring that workpieces of different materials and with different process requirements can obtain the optimal spraying parameters, thereby improving the stability and consistency of spraying quality.

[0023] Step S3 also includes the following steps: S31. During the spraying operation performed by the spraying robot, the surface of the target workpiece is scanned in real time using a three-dimensional vision scanning device to obtain the actual point cloud data of the target workpiece surface; and / or, the distribution data of the spraying material on the surface of the target workpiece is detected in real time using a spectral detection device.

[0024] By using three-dimensional visual scanning and / or spectral detection to collect spraying data in real time, it is possible to accurately obtain spraying status information on the workpiece surface, providing accurate data support for coverage determination.

[0025] A further technical solution of the present invention is: the calculation of the actual spray coverage and identification of uncovered areas in step S4 includes the following steps: S41. Register the actual spraying data with the three-dimensional model to determine the actual spraying state of each area on the surface of the target workpiece. S42. Based on the actual spraying state, calculate the ratio of the sprayed area to the total surface area of ​​the target workpiece to obtain the actual spraying coverage rate; S43. Mark the areas where no sprayed material was detected in the actual spraying state as uncovered areas; By using point cloud registration technology to precisely align the actual spraying data with the 3D model, the spraying status of each area on the workpiece surface can be accurately determined, thereby accurately calculating the coverage and identifying uncovered areas.

[0026] The point cloud registration of the actual spraying data and the 3D model includes the following steps: S411. Obtain the actual point cloud data from the actual spraying data; S412. Using the three-dimensional model as the target point cloud and the actual point cloud data as the source point cloud; S413. The source point cloud and the target point cloud are registered using the iterative nearest point algorithm to determine the spatial transformation matrix between the source point cloud and the target point cloud; S414. Based on the spatial transformation matrix, map the source point cloud to the coordinate system of the target point cloud; By employing the iterative nearest point algorithm for point cloud registration, the spatial transformation relationship between the actual spraying data and the 3D model can be accurately determined, providing technical support for the accurate determination of the spraying status.

[0027] Obtaining the actual coating coverage includes the following steps: S421. Based on the actual spraying state, determine the sprayed areas that have been covered by the spraying material in each region of the target workpiece surface; S422. Calculate the sum of the areas of the already sprayed areas to obtain the total sprayed area; S423. Obtain the total surface area of ​​the target workpiece in the three-dimensional model; S424. Calculate the ratio of the total area already sprayed to the total surface area, and use it as the actual spray coverage rate; It enables precise quantitative calculation of coating coverage, providing objective quantitative indicators for coating quality assessment.

[0028] Step S4 also includes the following steps: S44. Highlight or color-mark the uncovered areas in the visualization interface to generate a coverage visualization report; S45. The visualization report is used to guide manual touch-up spraying or can be invoked in step S5. Presenting uncovered areas in a visual manner allows operators to intuitively understand the coating quality and provides a data foundation for generating automatic touch-up spray paths.

[0029] The coverage visualization report includes at least one of the following: A three-dimensional model view of the target workpiece, wherein the painted area is displayed in a first color and the uncovered area is displayed in a second color; The numerical value of the actual spray coverage; Location coordinates and area data of each uncovered area.

[0030] Coverage visualization reports present coating quality information in an intuitive way, making it easy for operators to quickly locate and address uncovered areas.

[0031] A further technical solution of the present invention is: the generation of the supplementary spray path in step S5 includes the following steps: S51. Obtain the position coordinates and geometric features of the uncovered area in the three-dimensional model; S52. Based on the location coordinates and geometric features of the uncovered area, and combined with the reachable space and spraying parameters of the spraying robot, generate a respraying path to cover the uncovered area; S53. The re-spraying path does not overlap with the initial spraying path, and the overlap area between the re-spraying path and the initial spraying path meets the preset overlap width requirement. The generation of the touch-up spray path takes into account the specific location and geometric features of the uncovered areas, ensuring that the touch-up spray operation accurately covers the defective areas. At the same time, the touch-up spray path does not coincide with the initial path and meets the overlap width requirements, avoiding the problems of repeated spraying or insufficient overlap.

[0032] Step S51 further includes the following steps: S511. Extract the surface mesh data corresponding to the uncovered area from the three-dimensional model; S512. Calculate the geometric center coordinates, boundary contour, and area of ​​the uncovered area based on the surface mesh data; S513. Determine the surface curvature change information of the uncovered area based on the normal distribution of the surface mesh data; Surface mesh data was extracted from uncovered areas and analyzed for their geometric features and curvature variations, providing detailed data support for the accurate planning of the spraying path.

[0033] Step S52 also includes the following steps: S521. Based on the geometric characteristics of the uncovered area, determine the coverage strategy of the respraying path. The coverage strategy includes a unidirectional spraying strategy, a reciprocating spraying strategy, or a spiral spraying strategy. S522. Generate an initial touch-up spraying path to cover the uncovered area based on the coverage strategy and the spraying parameters; S523. Perform smoothing optimization on the initial spray path to generate the spray path.

[0034] The optimal coverage strategy is selected based on the geometric characteristics of the uncovered area, and a high-quality touch-up spraying path is generated through smoothing optimization, ensuring the efficiency and spraying quality of the touch-up spraying operation.

[0035] A further technical solution of the present invention is: the workpiece full-coverage spraying path planning method further includes the following steps: When the spraying operation is interrupted due to an abnormal situation, the current spraying progress information is saved, including the areas that have been sprayed and the areas that have not been sprayed. Once the abnormal situation is resolved, the spraying operation can continue from the interrupted position based on the spraying progress information, without the need for re-path planning or teaching. When the spraying operation is interrupted due to an abnormality, the progress information is saved and the operation can continue from the break point after the operation is restored. There is no need to replan the path or teach the operator, which significantly reduces the waste of time and materials and improves production efficiency.

[0036] When the target workpiece needs to undergo multiple spraying processes, the workpiece full-coverage spraying path planning method further includes the following steps: The three-dimensional model is used as a reference model and shared and reused among the multiple spraying processes. When a deviation is detected between the actual position of the target workpiece and the reference model during any process, the deviation is recorded. When the reference model is called in subsequent processes, position compensation is performed based on the recorded deviation; The 3D model is shared and reused across multiple processes, and positional deviation information is recorded and transmitted to ensure that the spraying path of each process matches the actual workpiece position, thereby improving the consistency and quality stability of multi-process spraying.

[0037] The multiple spraying processes include at least one or two of sandblasting, zinc spraying, and painting processes, and the spraying materials and process parameters corresponding to each spraying process are different; this application is applicable to a variety of different spraying processes such as sandblasting, zinc spraying, and painting, and has wide applicability.

[0038] The workpiece full-coverage spraying path planning method also includes the following steps: The spraying path data, actual spraying coverage data, and uncovered area data generated in each process are associated with the baseline model to generate full-process spraying quality traceability data for the target workpiece; the spraying data of each process are associated with the baseline model to form full-process quality traceability data, which facilitates subsequent quality analysis and process optimization.

[0039] The target workpiece is a large metal structural component, including hydraulic metal structural components, construction machinery structural components, or building steel structural components. This application is particularly applicable to the spraying operation of large metal structural components, solving technical problems such as the complex shape, poor accessibility, and high requirements for spraying quality of large workpieces.

[0040] Another objective of this invention is to provide a workpiece full-coverage spraying path planning system based on model-driven and vision-guided methods, the workpiece full-coverage spraying path planning system based on model-driven and vision-guided methods comprising: The model acquisition module is used to acquire the three-dimensional model of the target workpiece; The path generation module is used to generate an initial spraying path covering the surface of the target workpiece based on the three-dimensional model, wherein the initial spraying path avoids the target workpiece and surrounding obstacles. The vision acquisition module is used to acquire the actual spraying data of the target workpiece surface in real time during the spraying operation performed by the spraying robot along the initial spraying path; The coverage determination module is used to compare the actual spraying data with the three-dimensional model, calculate the actual spraying coverage of the target workpiece surface, and identify the uncovered areas on the target workpiece surface that have not been sprayed. The respray path generation module is used to generate a respray path based on the uncovered area; The control module controls the painting robot to perform painting operations along the initial painting path and to perform repainting operations along the repainting path. All modules of the system work collaboratively to achieve full automation from model acquisition to repainting execution. The model acquisition module provides the basic workpiece geometric data; the path generation module automatically plans obstacle avoidance paths; the vision acquisition module acquires the painting status in real time; the coverage determination module accurately identifies uncovered areas; the repainting path generation module automatically plans the repainting path; and the control module uniformly schedules and executes the processes, forming a complete closed-loop control system.

[0041] A further technical solution of the present invention is: the coverage determination module includes: A point cloud registration unit is used to perform point cloud registration between the actual spraying data and the three-dimensional model. The coverage calculation unit is used to calculate the ratio of the sprayed area to the total surface area of ​​the target workpiece based on the registration result, so as to obtain the actual spraying coverage. The uncovered area marking unit is used to mark areas where no sprayed material was detected as uncovered areas; the coverage determination module, through the coordinated work of the point cloud registration, coverage calculation and uncovered area marking units, can accurately determine the spraying quality and locate defective areas.

[0042] The visual acquisition module includes: At least one three-dimensional vision scanning device is disposed around the target workpiece to scan the surface of the target workpiece to obtain actual point cloud data; And / or, at least one spectral detection device is used to detect the distribution data of the sprayed material on the surface of the target workpiece; the visual acquisition module adopts three-dimensional visual scanning and / or spectral detection methods to comprehensively acquire the sprayed state information of the workpiece surface.

[0043] The supplementary spraying path generation module includes: The feature extraction unit is used to obtain the position coordinates and geometric features of the uncovered area in the three-dimensional model; The path planning unit is used to generate a respray path to cover the uncovered area based on the location coordinates and geometric features of the uncovered area, combined with the reachability space of the painting robot and the painting parameters. The respray path generation module automatically generates an accurate respray path through the cooperation of the feature extraction and path planning units.

[0044] Another object of the present invention is to provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in any of the preceding claims.

[0045] Another object of the present invention is to provide a computer program product, including a computer program or instructions that, when executed by a processor, implement the method described in any of the above-mentioned embodiments.

[0046] The beneficial effects of this invention are: by comparing the actual spraying data with the three-dimensional model after the spraying operation is completed, the spraying coverage rate can be automatically calculated and the uncovered areas can be accurately identified. Then, the respraying path can be automatically generated for respraying without manual inspection and intervention, realizing closed-loop control of full spraying coverage and effectively avoiding missed spraying defects.

[0047] By sharing and reusing the same reference model across multiple spraying processes and transmitting positional deviation information between processes, it ensures that the spraying path of each process is based on a unified coordinate system, avoiding cumulative errors caused by independent positioning of each process. It is particularly suitable for anti-corrosion spraying operations of large metal structural parts that require multiple processes such as sandblasting, zinc spraying, and painting.

[0048] It supports the automatic generation of point cloud models through laser 3D vision scanning and the automatic generation of spraying paths based on these models, eliminating the need for manual instruction or complex offline programming operations. This significantly reduces the technical requirements for operators and shortens the path planning time.

[0049] With the emergency stop memory function, the spraying operation can resume from the interrupted position after an abnormal interruption, without the need for re-teaching or replanning the path, which greatly improves the efficiency and continuity of the spraying operation.

[0050] By linking the spraying path data, coverage data, and uncovered area data of each process to the same benchmark model, a full-process spraying quality traceability data is formed, providing a data foundation for subsequent quality analysis and process optimization.

[0051] By calculating the robot's envelope space and performing interference detection during the path planning phase, all potential collision risks can be eliminated before the painting operation begins, ensuring the safety of the painting operation. Attached Figure Description

[0052] Figure 1 This is a flowchart illustrating the workpiece full-coverage spraying path planning method based on model-driven and vision-guided methods provided in this embodiment of the invention.

[0053] Figure 2 This is a schematic diagram of the process for obtaining a three-dimensional model of a target workpiece according to an embodiment of the present invention.

[0054] Figure 3 This is a schematic diagram of the process for generating the initial spraying path provided in an embodiment of the present invention.

[0055] Figure 4 This is a schematic diagram of the process for calculating the actual spray coverage and identifying uncovered areas provided in an embodiment of the present invention.

[0056] Figure 5 This is a schematic diagram of the process for generating a touch-up spray path provided in an embodiment of the present invention.

[0057] Figure 6 This is a structural block diagram of a workpiece full-coverage spraying path planning system based on model-driven and vision-guided methods provided in an embodiment of the present invention. Detailed Implementation

[0058] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0059] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0060] Example 1 This embodiment provides a model-driven and vision-guided method for planning the full-coverage spraying path of a workpiece, such as... Figure 1 As shown, the method includes the following steps S1 to S6.

[0061] Step S1: Obtain the three-dimensional model of the target workpiece. In this embodiment, the target workpiece is a large gate component in hydraulic metal structures, with dimensions of 28,000 mm in length, 4,500 mm in width, and 3,600 mm in height, and a weight of approximately 50 tons. This workpiece requires multiple coating processes, including sandblasting for rust removal, zinc spraying for corrosion protection, and paint spraying for protection.

[0062] like Figure 2As shown, step S1 specifically includes the following sub-steps: S11, importing the CAD design model of the target workpiece into the system as the constructed 3D model. For this large gate component, the design unit has provided a complete CAD 3D model, which can be directly imported into the system as a reference model. Simultaneously, for cases where there is no CAD model or the CAD model differs from the actual workpiece, the target workpiece can be scanned using a 3D vision scanning device to obtain the actual point cloud data of the target workpiece, and the actual point cloud data can be filtered to remove noise and outliers. In this embodiment, a laser 3D scanner is used to scan the workpiece with a scanning accuracy of ±0.1mm. The obtained raw point cloud data is filtered to remove outliers using a statistical filtering algorithm and to remove noise points using a radius filtering algorithm. S12, registering and fusing the filtered point cloud data to generate a complete surface point cloud model of the target workpiece. Since a single-view scanning cannot cover the entire surface of a large workpiece, this embodiment scans the workpiece from multiple views and uses the Iterative Closest Point (ICP) algorithm to register and fuse the multi-view point cloud data to generate a complete surface point cloud model. S13. Triangulate the surface point cloud model to generate the 3D model. In this embodiment, the Delaunay triangulation algorithm is used to triangulate the point cloud model, generating a 3D mesh model that can be directly called by the path planning algorithm. It is compatible with both drawing modeling and physical object scanning modeling modes, and can complete the digital modeling of non-standard workpieces without relying on complete design drawings. It is suitable for various specifications of hydraulic engineering and heavy steel structures in buildings, and solves the problem of large workpieces lacking a unified 3D model and having irregular shapes that are difficult to program.

[0063] Step S2: Based on the 3D model, an initial spraying path is generated to cover the surface of the target workpiece. The initial spraying path avoids the target workpiece and surrounding obstacles. In this embodiment, when the 3D model is constructed through visual scanning, before executing step S2, the actual position of the target workpiece is located using a laser 3D visual scanning device. The deviation between the actual placement position of the target workpiece and the theoretical position of the 3D model is obtained, and then the position of the 3D model is calibrated and compensated based on this deviation. Specifically, after the workpiece is in place, at least three non-coplanar feature points on the workpiece are scanned using a laser 3D visual scanning device fixed in the sandblasting room. The coordinates of the scanned feature points are compared with the theoretical coordinates of the corresponding feature points in the 3D model. A spatial transformation matrix (including rotation matrix and translation vector) is calculated, and then the position of the 3D model is calibrated based on the spatial transformation matrix to align the model coordinate system with the actual workpiece coordinate system.

[0064] like Figure 3 As shown, generating the initial spraying path in step S2 specifically includes the following sub-steps: S21. Based on the calibrated 3D model, the spraying path program for the target workpiece is automatically generated using offline programming software. In this embodiment, RobotStudio offline programming software is used. After importing the calibrated 3D model, the software automatically plans the spraying path according to the geometric features of the workpiece surface. For cases where there is no CAD model or the CAD model differs significantly from the actual workpiece, a point cloud model can also be formed by scanning the target workpiece with laser 3D vision, and the spraying path program can be automatically generated based on the point cloud model. Specifically: the target workpiece is scanned using a laser 3D vision scanning device to obtain the surface point cloud data of the target workpiece; the surface point cloud data is preprocessed (including noise reduction, smoothing, downsampling, etc.) to generate a 3D point cloud model of the target workpiece; surface features are extracted from the 3D point cloud model to identify the surface area to be sprayed on the target workpiece; based on the geometric features of the surface area to be sprayed (such as planes, curved surfaces, edges, etc.) and the kinematic parameters of the spraying robot (such as joint limits, reachable space, maximum speed, etc.), a spraying path program covering the surface area to be sprayed is automatically generated.

[0065] S22. Identify obstacle information around the target workpiece, including the bracket supporting the target workpiece and fixed equipment in the environment where the target workpiece is located. Specifically: Obtain a three-dimensional environment model of the environment where the target workpiece is located. This three-dimensional environment model can be obtained by performing a three-dimensional scan of the environment or constructed based on the workshop design drawings; mark the position and outline information of the bracket and fixed equipment (such as tracks, support columns, dust removal pipes, etc. in the sandblasting room) in the three-dimensional environment model; use the position and outline information of the bracket and fixed equipment as obstacle information.

[0066] S23. The acquired obstacle information and the spraying path program are combined to generate the latest initial spraying path, ensuring that the initial spraying path avoids obstacles and prevents interference or collision between the spraying robot and obstacles. Specifically: the envelope space of the spraying robot at each path point of the initial spraying path (including the envelope of the robot body, spray gun, and pipeline) is calculated; it is determined whether the envelope space interferes with any obstacle in the obstacle information; if so, the position or orientation of the path point is adjusted until the envelope space does not interfere with any obstacle. In this embodiment, a fast collision detection algorithm based on bounding boxes is used for preliminary interference judgment. For areas where interference may occur, a precise collision detection algorithm based on meshes is used for verification and path adjustment.

[0067] Before performing step S3, the following steps are also included: S301. Based on the material of the target workpiece and the process requirements of the spraying operation, the corresponding spraying process parameter set is retrieved from the preset process parameter database. In this embodiment, for the sandblasting process, the retrieved process parameters include: sandblasting pressure 0.7-0.8MPa, single-gun air consumption 12-16m³ / min, and nozzle diameter 11-18mm; for the zinc spraying process, the retrieved process parameters include: spraying voltage, spraying current, and wire feeding speed; for the painting process, the retrieved process parameters include: spraying pressure, spraying flow rate, spraying speed, and spraying distance.

[0068] S302. Send the spraying process parameter set to the spraying robot to control the spraying robot to perform the spraying operation according to the set process parameters.

[0069] Laser vision position compensation eliminates workpiece transfer and positioning errors, while automatically identifying tooling and equipment obstacles and optimizing the trajectory, completely solving the problems of traditional offline programming not considering on-site placement deviations and robot collisions damaging equipment; it supports dual-path automatic programming of CAD models and on-site point clouds, completely replacing manual point-by-point teaching, and reducing single workpiece programming time by more than 90%.

[0070] Step S3: Control the spraying robot to perform spraying operations on the target workpiece along the initial spraying path, and collect the actual spraying data of the target workpiece surface in real time. In this embodiment, the spraying robot is an overhead crane (gantry) 8-axis sandblasting robot, which performs sandblasting operations along the preset initial spraying path. During the spraying operation, multiple 3D vision scanning devices fixed in the sandblasting chamber scan the workpiece surface in real time to obtain the actual point cloud data of the workpiece surface. At the same time, the distribution data of the sprayed material on the workpiece surface is detected in real time by a spectral detection device to determine whether each area has been effectively sprayed.

[0071] Step S4: Compare the actual spraying data with the 3D model to calculate the actual spraying coverage of the target workpiece surface, and identify the uncovered areas on the target workpiece surface that were not sprayed; for example... Figure 4 As shown, step S4 specifically includes the following sub-steps: S41. Register the actual spraying data with the 3D model to determine the actual spraying state of each area on the surface of the target workpiece. Specifically: acquire the actual point cloud data from the actual spraying data; use the 3D model as the target point cloud and the actual point cloud data as the source point cloud; use the iterative nearest point algorithm to register the source point cloud and the target point cloud to determine the spatial transformation matrix between the source point cloud and the target point cloud; based on the spatial transformation matrix, map the source point cloud to the coordinate system of the target point cloud, thereby achieving accurate alignment between the actual spraying data and the 3D model.

[0072] S42. Based on the actual spraying status, calculate the ratio of the sprayed area to the total surface area of ​​the target workpiece to obtain the actual spraying coverage rate. Specifically: Based on the actual spraying status (determining whether each area is covered by the spraying material through point cloud color information or spectral detection data), determine the sprayed areas covered by the spraying material in each region of the target workpiece surface; calculate the sum of the areas of the sprayed areas to obtain the total sprayed area; obtain the total surface area of ​​the target workpiece in the 3D model; calculate the ratio of the total sprayed area to the total surface area as the actual spraying coverage rate.

[0073] S43. Mark areas where no sprayed material was detected in the actual spraying process as uncovered areas.

[0074] S44. Highlight or color-code the uncovered areas in the visualization interface (e.g., painted areas are displayed in green, and uncovered areas are displayed in red), and generate a coverage visualization report. This visualization report includes: a 3D model view of the target workpiece (where painted areas are displayed in a first color and uncovered areas in a second color); the actual paint coverage value; and the location coordinates and area data of each uncovered area.

[0075] S45. This visualization report is used to guide manual touch-up spraying or can be called by step S5 to automatically generate a touch-up spraying path.

[0076] It enables digital quantitative detection of coating coverage, replacing subjective judgment by human eyes; it visually and intuitively displays missed spraying areas, accurately locates difficult-to-spray areas such as inner cavities, bottom surfaces, and rib plate obstructions, and provides accurate coordinate data support for automated touch-up spraying.

[0077] Step S5: Generate a respray path based on the uncovered area; such as... Figure 5 As shown, step S5 specifically includes the following sub-steps: S51. Obtain the position coordinates and geometric features of the uncovered area in the 3D model. Specifically: extract the surface mesh data corresponding to the uncovered area from the 3D model; calculate the geometric center coordinates, boundary contour, and area of ​​the uncovered area based on the surface mesh data; determine the surface curvature variation information of the uncovered area based on the normal distribution of the surface mesh data.

[0078] S52. Based on the location coordinates and geometric features of the uncovered area, combined with the reachable space of the painting robot and the painting parameters, generate a respray path to cover the uncovered area. Specifically: Based on the geometric features of the uncovered area (such as area size, shape regularity, curvature variation, etc.), determine the coverage strategy of the respray path. For areas with large areas and regular shapes, adopt a unidirectional spraying strategy or a reciprocating spraying strategy; for areas with small areas or irregular shapes, adopt a spiral spraying strategy. Based on the coverage strategy and painting parameters, generate an initial respray path to cover the uncovered area. Perform smoothing optimization on the initial respray path (such as using B-spline curve fitting) to generate a smooth respray path.

[0079] S53. The touch-up spraying path does not overlap with the initial spraying path, and the overlap area between the touch-up spraying path and the initial spraying path meets the preset overlap width requirement (in this embodiment, the preset overlap width is 30%-50% of the path spacing) to ensure a smooth transition between the touch-up spraying area and the already sprayed area, without obvious overlap marks or missed spraying gaps.

[0080] For areas with different shapes of missed spraying, an adaptive matching spraying strategy is used to automatically generate a smooth and vibration-free respray trajectory, ensuring uniform coating thickness in the resprayed area and eliminating the problems of uneven thickness and rework due to missed spraying.

[0081] Step S6: Control the painting robot to perform a touch-up spraying operation on the uncovered area along the touch-up spraying path. After the touch-up spraying operation is completed, the system executes steps S3 to S4 again to re-inspect the surface of the workpiece after touch-up spraying. After confirming that the actual spraying coverage rate reaches the preset threshold (in this embodiment, the preset coverage rate is above 98%), the painting operation ends.

[0082] In addition, the method of this embodiment also has the following functions: When a painting operation is interrupted due to an abnormal situation (such as equipment failure or sudden power outage), the system automatically saves the current painting progress information, including the areas that have been painted and the areas that have not been painted. Once the abnormal situation is resolved, the system resumes the painting operation from the interrupted position based on the saved painting progress information, without the need for re-path planning or re-teaching.

[0083] Since the target workpiece in this embodiment needs to undergo three spraying processes—sandblasting, zinc spraying, and painting—the method in this embodiment uses the 3D model obtained in step S1 as a reference model, which is shared and reused among the three spraying processes. When a deviation is detected between the actual position of the target workpiece and the reference model in any process, the deviation is recorded. When the reference model is called in subsequent processes, position compensation is performed based on the recorded deviation. At the same time, the spraying path data, actual spraying coverage data, and uncovered area data generated in each process are associated with the reference model to generate full-process spraying quality traceability data for the target workpiece.

[0084] As another core improvement of this invention, this method adds an abnormal breakpoint resume spraying function: if abnormal interruption scenarios such as excessive dust, combustible gas alarm, robot servo failure, power failure, or equipment emergency stop occur during the spraying operation, the system automatically stores the current complete spraying progress data packet in real time. The data packet includes the coordinates of the sprayed area, the unsprayed area, the current robot coordinates, and the remaining spraying steps; after the fault is cleared and the equipment is reset and restarted, the system directly reads the spraying progress data packet, automatically locates the interruption stop point, and does not require remodeling, replanning the initial path, or manual secondary teaching. The robot continues to complete the remaining spraying operation from the interruption position, which greatly reduces paint waste and production waiting time.

[0085] For multi-stage continuous anti-corrosion production lines (sandblasting section → zinc spraying section → multi-stage painting section → powder coating section), this method adds a cross-stage model sharing and deviation compensation mechanism: 1) The initial 3D model of the workpiece is used as a unified reference model for the entire process. The control systems of each section, such as sandblasting, zinc spraying, painting, and powder spraying, share the same reference model, eliminating the need for separate modeling for each process. 2) When the workpiece is transferred from the sandblasting room to the zinc spraying room, paint spraying room, drying room and powder spraying room, the visual scanning collects the workpiece placement deviation and binds the offset deviation data to the reference model for storage. 3) When the workpiece enters the next spraying process, the system automatically retrieves the position deviation stored in the previous process, compensates the reference model for coordinates in real time, eliminates workpiece transfer and positioning errors, and ensures uniform path planning accuracy for each process.

[0086] Meanwhile, this method has a full-process anti-corrosion quality traceability function: it links and binds the spraying path program, real-time spraying coverage, coordinates of missed spraying areas, spraying process parameters, and equipment operating parameters of the workpiece sandblasting, zinc spraying, each painting and powder spraying process to a unified benchmark 3D model, forming a single workpiece full-process spraying quality traceability database; the database supports local storage and uploading to the factory MES system via Ethernet OPC-UA / MODBUS TCP protocol, realizing digital full life-cycle traceability of workpiece anti-corrosion quality and meeting the information integration requirements of smart manufacturing factories.

[0087] This invention is applicable to large metal structural components, including three main product categories: hydraulic metal structural components with a maximum size of 28m, construction machinery structural components within 10m, and construction steel structural components within 16m; the applicable processes include a full range of anti-corrosion processes such as Sa2.5~Sa3.0 grade sandblasting and rust removal, thermal spraying zinc, multi-layer painting, and electrostatic powder coating.

[0088] Example 2 This embodiment provides a workpiece full-coverage spraying path planning system based on model-driven and vision-guided methods, such as... Figure 6As shown, the system includes: The model acquisition module is used to acquire a 3D model of the target workpiece. This module includes a CAD model import unit and / or a 3D vision scanning unit. The CAD model import unit is used to import the CAD design model of the target workpiece into the system, and the 3D vision scanning unit is used to scan the target workpiece using a 3D vision scanning device to obtain actual point cloud data and generate a 3D model.

[0089] The path generation module is used to generate an initial spraying path covering the surface of the target workpiece based on a 3D model. The initial spraying path avoids obstacles around the target workpiece. This module includes an offline programming unit and / or a point cloud path generation unit, as well as an obstacle recognition unit and a collision avoidance unit.

[0090] A vision acquisition module is used to acquire actual spraying data on the surface of the target workpiece in real time during the spraying operation performed by the spraying robot along the initial spraying path. This module includes at least one three-dimensional vision scanning device (set around the target workpiece to scan the surface of the target workpiece to obtain actual point cloud data) and / or at least one spectral detection device (for detecting the distribution data of the sprayed material on the surface of the target workpiece).

[0091] The coverage determination module compares the actual spraying data with the 3D model to calculate the actual spraying coverage of the target workpiece surface and identify uncovered areas on the target workpiece surface that have not been sprayed. This module includes: a point cloud registration unit for registering the actual spraying data with the 3D model; a coverage calculation unit for calculating the ratio of the sprayed area to the total surface area of ​​the target workpiece based on the registration results, thus obtaining the actual spraying coverage; and an uncovered area marking unit for marking areas where no sprayed material was detected as uncovered areas.

[0092] The touch-up spraying path generation module is used to generate touch-up spraying paths based on uncovered areas. This module includes: a feature extraction unit, used to obtain the position coordinates and geometric features of the uncovered areas in the 3D model; and a path planning unit, used to generate touch-up spraying paths to cover the uncovered areas based on the position coordinates and geometric features of the uncovered areas, combined with the reachability space of the spraying robot and spraying parameters.

[0093] The control module is used to control the painting robot to perform painting operations along the initial painting path and to perform touch-up painting operations along the touch-up painting path.

[0094] Example 3 This embodiment provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method described in Embodiment 1. The computer-readable storage medium can be any tangible medium containing or storing a program, including but not limited to: USB flash drive, portable hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0095] Example 4 This embodiment provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements the method described in Embodiment 1. This computer program product can be downloaded and installed via a network, or distributed via physical media such as optical discs.

[0096] The workpiece full-coverage spraying path planning method and system based on model-driven and vision-guided methods provided by this invention can be widely applied in the following fields: In the field of anti-corrosion spraying of hydraulic metal structural components: it can be used for automated spraying operations such as sandblasting and rust removal, zinc (aluminum) spraying for corrosion protection and paint protection of hydraulic metal structural components such as large gates, hoists, and pressure steel pipes.

[0097] In the field of coating of structural components of construction machinery: it can be used for automated coating operations of structural components of construction machinery such as construction hoisting machinery, concrete machinery, and road machinery.

[0098] In the field of coating of building steel structures: it can be used for sandblasting and rust removal, powder coating, painting and other anti-corrosion operations of building steel structures.

[0099] Other large metal structural component painting applications: It can be used for automated painting operations of large metal structural components such as ship sections, aerospace components, and rail transit vehicles.

[0100] This invention can effectively improve the automation level and coating quality of large metal structural parts spraying operations, reduce manual labor intensity and skill dependence, and has significant industrial practical value and economic value.

[0101] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A workpiece full-coverage spraying path planning method based on model-driven and vision-guided methods, characterized in that, The workpiece full-coverage spraying path planning method includes the following steps: S1: Obtain the 3D model of the target workpiece; S2: Based on the three-dimensional model, an initial spraying path is generated to cover the surface of the target workpiece, and the initial spraying path avoids the target workpiece and surrounding obstacles; S3: Control the spraying robot to perform spraying operations on the target workpiece along the initial spraying path, and collect the actual spraying data on the surface of the target workpiece in real time; S4: Compare the actual spraying data with the three-dimensional model, calculate the actual spraying coverage of the target workpiece surface, and identify the uncovered areas on the target workpiece surface that were not sprayed. S5: Generate a respray path based on the uncovered area; S6: Control the painting robot to perform a touch-up spraying operation on the uncovered area along the touch-up spraying path.

2. The workpiece full-coverage spraying path planning method based on model-driven and vision-guided methods according to claim 1, characterized in that, The step S1 of obtaining the three-dimensional model of the target workpiece includes the following steps: S11. Import the CAD design model of the target workpiece into the system as the constructed three-dimensional model; And / or, the target workpiece is scanned by a three-dimensional vision scanning device to obtain the actual point cloud data of the target workpiece, and the actual point cloud data is filtered to remove noise and outliers; S12. Register and fuse the filtered point cloud data to generate a complete surface cloud model of the target workpiece. S13. Perform triangular meshing on the surface point cloud model to generate the three-dimensional model.

3. The workpiece full-coverage spraying path planning method based on model-driven and vision-guided methods according to claim 2, characterized in that, When the 3D model is constructed through visual scanning, the following steps are included before step S2: S201. The target workpiece is located in actual position using a laser three-dimensional vision scanning device, and the deviation between the actual placement position of the target workpiece and the theoretical position of the three-dimensional model is obtained. S202. Perform position calibration compensation on the three-dimensional model based on the deviation; The step S2, which generates the initial spraying path, also includes the following steps: S21. Based on the three-dimensional model, the spraying path program for the target workpiece is automatically generated using offline programming software; or, the target workpiece is scanned by laser three-dimensional vision to form a point cloud model, and the spraying path program is automatically generated based on the point cloud model. S22. Identify obstacle information around the target workpiece, the obstacles including a bracket that carries the target workpiece and fixed equipment in the environment where the target workpiece is located; S23. Combine the acquired obstacle information and the spraying path program to generate the latest initial spraying path, so that the initial spraying path avoids the obstacle and prevents the spraying robot from interfering with or colliding with the obstacle; The automatic generation program for spraying paths from the point cloud model includes the following steps: S211. Scan the target workpiece using a laser three-dimensional vision scanning device to obtain the surface point cloud data of the target workpiece; S212. Preprocess the surface point cloud data to generate a three-dimensional point cloud model of the target workpiece; S213. Extract surface features from the three-dimensional point cloud model to identify the surface area to be coated on the target workpiece; S214. Based on the geometric features of the surface area to be sprayed and the kinematic parameters of the spraying robot, automatically generate a spraying path program that covers the surface area to be sprayed; Identifying obstacle information around the target workpiece includes the following steps: S221. Obtain a three-dimensional environment model of the environment in which the target workpiece is located; S222. Mark the position and outline information of the bracket and the fixing device in the three-dimensional environment model; S223. Use the position and outline information of the bracket and the fixing device as the obstacle information; The initial spraying path avoids the obstacle, which includes the following steps: S231. Calculate the envelope space of the spraying robot at each path point of the initial spraying path; S232. Determine whether the envelope space spatially interferes with any obstacle in the obstacle information; S233. If so, adjust the position or orientation of the path point until the envelope space does not interfere with any obstacles.

4. The workpiece full-coverage spraying path planning method based on model-driven and vision-guided methods according to claim 3, characterized in that, The following steps are included before step S3: S301. Based on the material of the target workpiece and the process requirements of the spraying operation, call the corresponding spraying process parameter set from the preset process parameter database. S302. The spraying process parameter set is sent to the spraying robot, wherein the spraying process parameter set includes at least one or more of spraying pressure, spraying flow rate, spraying speed and spraying distance; Step S3 also includes the following steps: S31. During the spraying operation performed by the spraying robot, the surface of the target workpiece is scanned in real time using a three-dimensional vision scanning device to obtain the actual point cloud data of the target workpiece surface; and / or, the distribution data of the spraying material on the surface of the target workpiece is detected in real time using a spectral detection device.

5. The workpiece full-coverage spraying path planning method based on model-driven and vision-guided methods according to claim 4, characterized in that, The step S4, which calculates the actual spray coverage and identifies uncovered areas, includes the following steps: S41. Register the actual spraying data with the three-dimensional model to determine the actual spraying state of each area on the surface of the target workpiece. S42. Based on the actual spraying state, calculate the ratio of the sprayed area to the total surface area of ​​the target workpiece to obtain the actual spraying coverage rate; S43. Mark the areas where no sprayed material was detected in the actual spraying state as uncovered areas; The point cloud registration of the actual spraying data and the 3D model includes the following steps: S411. Obtain the actual point cloud data from the actual spraying data; S412. Using the three-dimensional model as the target point cloud and the actual point cloud data as the source point cloud; S413. The source point cloud and the target point cloud are registered using the iterative nearest point algorithm to determine the spatial transformation matrix between the source point cloud and the target point cloud; S414. Based on the spatial transformation matrix, map the source point cloud to the coordinate system of the target point cloud; Obtaining the actual coating coverage includes the following steps: S421. Based on the actual spraying state, determine the sprayed areas that have been covered by the spraying material in each region of the target workpiece surface; S422. Calculate the sum of the areas of the already sprayed areas to obtain the total sprayed area; S423. Obtain the total surface area of ​​the target workpiece in the three-dimensional model; S424. Calculate the ratio of the total area already sprayed to the total surface area, and use it as the actual spray coverage rate; Step S4 also includes the following steps: S44. Highlight or color-mark the uncovered areas in the visualization interface to generate a coverage visualization report; S45. The visualization report is used to guide manual touch-up spraying or can be invoked in step S5. The coverage visualization report includes at least one of the following: A three-dimensional model view of the target workpiece, wherein the painted area is displayed in a first color and the uncovered area is displayed in a second color; The numerical value of the actual spray coverage; Location coordinates and area data of each uncovered area.

6. The workpiece full-coverage spraying path planning method based on model-driven and vision-guided methods according to claim 5, characterized in that, The step S5 of generating the supplementary spray path includes the following steps: S51. Obtain the position coordinates and geometric features of the uncovered area in the three-dimensional model; S52. Based on the location coordinates and geometric features of the uncovered area, and combined with the reachable space and spraying parameters of the spraying robot, generate a respraying path to cover the uncovered area; S53. The re-spraying path does not overlap with the initial spraying path, and the overlap area between the re-spraying path and the initial spraying path meets the preset overlap width requirement. Step S51 further includes the following steps: S511. Extract the surface mesh data corresponding to the uncovered area from the three-dimensional model; S512. Calculate the geometric center coordinates, boundary contour, and area of ​​the uncovered area based on the surface mesh data; S513. Determine the surface curvature change information of the uncovered area based on the normal distribution of the surface mesh data; Step S52 also includes the following steps: S521. Based on the geometric characteristics of the uncovered area, determine the coverage strategy of the respraying path. The coverage strategy includes a unidirectional spraying strategy, a reciprocating spraying strategy, or a spiral spraying strategy. S522. Generate an initial touch-up spraying path to cover the uncovered area based on the coverage strategy and the spraying parameters; S523. Perform smoothing optimization on the initial spray path to generate the spray path; The workpiece full-coverage spraying path planning method also includes the following steps: When the spraying operation is interrupted due to an abnormal situation, the current spraying progress information is saved, including the areas that have been sprayed and the areas that have not been sprayed. Once the abnormal situation is resolved, the spraying operation can continue from the interrupted position based on the spraying progress information, without the need for re-path planning or teaching. When the target workpiece needs to undergo multiple spraying processes, the workpiece full-coverage spraying path planning method further includes the following steps: The three-dimensional model is used as a reference model and shared and reused among the multiple spraying processes. When a deviation is detected between the actual position of the target workpiece and the reference model during any process, the deviation is recorded. When the reference model is called in subsequent processes, position compensation is performed based on the recorded deviation; The multiple spraying processes include at least one or two of sandblasting, zinc spraying and painting processes, and the spraying materials and process parameters corresponding to each spraying process are different; The workpiece full-coverage spraying path planning method also includes the following steps: The spraying path data, actual spraying coverage data, and uncovered area data generated in each of the aforementioned processes are associated with the baseline model to generate full-process spraying quality traceability data for the target workpiece. The target workpiece is a large metal structural component, including hydraulic metal structural components, construction machinery structural components, or building steel structural components.

7. A workpiece full-coverage spraying path planning system based on model-driven and vision-guided methods, characterized in that, The model-driven and vision-guided workpiece full-coverage spraying path planning system includes: The model acquisition module is used to acquire the three-dimensional model of the target workpiece; The path generation module is used to generate an initial spraying path covering the surface of the target workpiece based on the three-dimensional model, wherein the initial spraying path avoids the target workpiece and surrounding obstacles. The vision acquisition module is used to acquire the actual spraying data of the target workpiece surface in real time during the spraying operation performed by the spraying robot along the initial spraying path; The coverage determination module is used to compare the actual spraying data with the three-dimensional model, calculate the actual spraying coverage of the target workpiece surface, and identify the uncovered areas on the target workpiece surface that have not been sprayed. The respray path generation module is used to generate a respray path based on the uncovered area; The control module is used to control the painting robot to perform painting operations along the initial painting path and to perform touch-up painting operations along the touch-up painting path.

8. The workpiece full-coverage spraying path planning system based on model-driven and vision-guided methods according to claim 7, characterized in that, The coverage determination module includes: A point cloud registration unit is used to perform point cloud registration between the actual spraying data and the three-dimensional model. The coverage calculation unit is used to calculate the ratio of the sprayed area to the total surface area of ​​the target workpiece based on the registration result, so as to obtain the actual spraying coverage. The uncovered area marking unit is used to mark areas where no sprayed material was detected as uncovered areas; The visual acquisition module includes: At least one three-dimensional vision scanning device is disposed around the target workpiece to scan the surface of the target workpiece to obtain actual point cloud data; And / or, at least one spectral detection device for detecting the distribution data of the sprayed material on the surface of the target workpiece; The supplementary spraying path generation module includes: The feature extraction unit is used to obtain the position coordinates and geometric features of the uncovered area in the three-dimensional model; The path planning unit is used to generate a respray path to cover the uncovered area based on the location coordinates and geometric features of the uncovered area, combined with the reachable space and spraying parameters of the spraying robot.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the method as described in any one of claims 1 to 6.

Images