Workpiece handling method, device, control apparatus, system, and program product

Abstract

The application relates to the field of industrial automation, in particular to a workpiece turnover method, device, control equipment, system and program product. The method first performs three-dimensional imaging on a grabbing material frame, determines a target workpiece, a pose of the target workpiece and a grabbing point, and performs a turnover judgment. When turnover is required, the grabbing material frame is moved to a positioning point after being re-grabbed through a turnover table, a machining hole position is identified in combination with a two-dimensional image, a robot is controlled to align the workpiece with a positioning piece on the positioning machining table, and push assembly and machining are completed. After machining, the workpiece is sent into a placing material frame, and when the grabbing material frame is empty, a linkage transfer mechanism is used to complete adjacent material frame position compensation and grabbing and placing material frame switching, so that turnover of the next type of workpiece is performed. The grabbing success rate, positioning and machining accuracy and multi-workpiece continuous turnover efficiency are improved, and downtime and manual intervention are reduced.

Description

Technical Field

[0001] This application relates to the field of intelligent manufacturing technology, and in particular to a workpiece turnover method, apparatus, control equipment, system and program product. Background Technology

[0002] In industrial automated production, the gripping, alignment, and cyclic turnover of workpieces are the core links of precision manufacturing. In precision manufacturing fields such as automotive parts and precision hardware, there are stringent requirements for workpiece assembly accuracy and the ability of automated continuous operation of production lines.

[0003] Traditional automated production solutions are often designed for single workpieces, lacking compatibility with multiple types of workpieces, and rely on manual intervention for adjustments. This makes it difficult to achieve automated cyclical operations for multiple types of workpieces and cannot meet the needs of flexible production.

[0004] Therefore, there is an urgent need to provide an automated workpiece circulation solution that can adapt to multiple types of workpieces in order to improve the production line's flexible production capabilities and automated operation efficiency. Summary of the Invention

[0005] The workpiece turnover method, apparatus, control equipment, system, and program products provided in this application obtain the workpiece pose and gripping points through three-dimensional image recognition, and adjust the workpiece by flipping it as needed. Combined with two-dimensional images, the processing holes are precisely aligned and assembled. With the help of the double-sided material frame automatic rotation station, it can adapt to the automated processing of multiple types of workpieces. It gets rid of the shortcomings of traditional solutions that can only adapt to a single workpiece and rely on manual debugging, and realizes the fully automatic cyclic turnover of multiple types of workpieces, effectively improving the flexible production capacity and automated operation efficiency of the production line.

[0006] In a first aspect, embodiments of this application provide a workpiece turnover method, including:

[0007] After configuring the first material frame as a gripping material frame and the second material frame as a placing material frame, a 3D image of the gripping material frame is acquired. The first material frame and the second material frame are located at the gripping and placing stations on the left and right sides of the robot, respectively. The first material frame contains multiple workpieces, and the second material frame is an empty material frame.

[0008] Based on the 3D image, the target workpiece is determined from multiple workpieces stacked in the gripping frame, and the pose and gripping point of the target workpiece are determined.

[0009] Based on the pose and gripping points of the target workpiece, a flipping determination result is generated;

[0010] If the flipping determination result is the first result, then control the robot to grab the target workpiece and place the target workpiece on the flipping table on the side where the grabbing frame is located, and grab the target workpiece again. After grabbing the target workpiece again, control the robot to move along the preset path to the positioning point.

[0011] After the robot moves to the positioning point, a two-dimensional image of the robot is acquired, and the position of the machining hole of the target workpiece is determined based on the two-dimensional image;

[0012] Based on the position of the machining hole of the target workpiece, the robot is controlled to align the target workpiece with the target positioning part of the positioning machining table. After alignment, the robot is controlled to push the target workpiece to complete the assembly and cooperation with the target positioning part, and execute the preset machining process.

[0013] After the preset processing steps are completed, the robot is controlled to move the target workpiece to the material placement frame;

[0014] When the gripping frame is empty, control the movement of the transfer mechanism corresponding to the placement frame so that the third frame adjacent to the placement frame is in the gripping and placement position; switch the gripping frame to the third frame and switch the placement frame to the first frame. The workpieces stacked in the third frame and the first frame are different.

[0015] In one possible embodiment, a flipping determination result is generated based on the pose of the target workpiece and the gripping point, including:

[0016] The preset fixture model is called, and the gripping posture and contact area of ​​the robot after gripping the target workpiece are estimated based on the pose and gripping point of the target workpiece.

[0017] When the gripping posture interferes with the machining hole or the contact area is less than the preset area threshold, the flipping judgment result is determined as the first result.

[0018] In one possible embodiment, the method further includes:

[0019] If the flipping result is the second result, then control the robot to grab the target workpiece and move along the preset path.

[0020] In one possible embodiment, multiple material frames are distributed on the left and right sides of the robot, each material frame being used to stack workpieces of a corresponding type; the positioning processing table includes multiple positioning components, each positioning component corresponding to a type of workpiece. Based on the position of the machining hole of the target workpiece, the robot is controlled to align the target workpiece with the target positioning component of the positioning processing table, including:

[0021] Based on the type of the target workpiece, the target positioning component is determined from multiple positioning components;

[0022] Based on the position of the machining hole in the target workpiece, during or after the robot moves to the corresponding position of the target positioning part, the clamping posture of the grasped target workpiece is adjusted so that the machining hole of the target workpiece is aligned with the target positioning part.

[0023] In one possible embodiment, the method further includes:

[0024] The type of target workpiece is determined based on a two-dimensional image or the material frame identifier of the gripping frame.

[0025] In one possible embodiment, the method further includes:

[0026] Based on the type and / or geometric features of the target workpiece, the robot's gripper is switched to grasp the target workpiece using the switched gripper; the geometric features are extracted based on the 3D image.

[0027] In a second aspect, embodiments of this application provide a control device, including: a memory and a processor;

[0028] The memory stores instructions that the computer executes;

[0029] The processor executes computer execution instructions stored in memory, causing the processor to perform the methods described above.

[0030] Thirdly, this application provides a workpiece turnover system, including: a robot, a positioning processing table, a three-dimensional image acquisition device, a two-dimensional image acquisition device, a flipping table, a control unit, and two working modules respectively located on the left and right sides of the robot;

[0031] Each operating module includes a transfer mechanism and multiple material frames located above the transfer mechanism;

[0032] The material frames are configured to hold multiple workpieces, with at least some material frames holding different workpieces;

[0033] The transfer mechanism is used to move multiple material frames on it to switch the material frames located at the gripping and releasing station; the three-dimensional image acquisition device is used to acquire three-dimensional images of the material frames located at the gripping and releasing station.

[0034] The two-dimensional image acquisition device is set at the positioning point to acquire two-dimensional images of the robot after it has moved to the positioning point to grasp the workpiece.

[0035] The positioning processing table is located in the hollow area between the two working modules. Multiple positioning components are arranged on the positioning processing table and are configured to assemble and cooperate with the corresponding type of workpiece grasped by the robot to complete the workpiece processing.

[0036] The control unit is used to execute the methods provided above to control the robot to turn workpieces between the material frame, the flipping table and the positioning processing table, and to control the transfer mechanism to move multiple material frames on the corresponding side to switch the material frame located at the gripping and releasing station.

[0037] In one possible embodiment, the workpiece includes flat, right-angled, and irregularly shaped sheet metal parts;

[0038] Multiple positioning components include positioning components that respectively match the outer contour and machining holes of plate-shaped, right-angled, and irregular-shaped sheet metal parts.

[0039] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods provided above.

[0040] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the method described above.

[0041] The workpiece turnover method, apparatus, control equipment, system, and program products provided in this application, by acquiring three-dimensional images of workpieces in the gripping frame at the gripping station and determining the target workpiece, its pose, and gripping point based on the three-dimensional images, can accurately select the workpiece to be processed from multiple randomly stacked workpieces, thereby improving the gripping adaptability and gripping success rate of various types of workpieces; by generating a flipping determination result based on the pose and gripping point of the target workpiece, and controlling the robot to place the target workpiece on the flipping table and re-grab it when flipping is required, and then moving it to the positioning point, it can adjust the posture of workpieces with different initial postures, thereby providing... Subsequent positioning and processing provide a consistent assembly basis. By acquiring two-dimensional images and determining the position of the machining hole of the target workpiece after the robot moves to the positioning point, the robot is then controlled to align the target workpiece with the target positioning part of the positioning processing table, push and assemble it, and execute the preset processing steps. This can improve the workpiece positioning accuracy and the accuracy of processing fit, thereby ensuring processing quality and operational stability. After the preset processing steps are completed, the target workpiece is transferred to the placement frame, and the correspondence between the gripping frame and the placement frame is switched when the gripping frame is empty. This can achieve continuous turnover between different workpiece frames, thereby improving operational efficiency and production continuity. Attached Figure Description

[0042] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0043] Figure 1 A schematic flowchart illustrating the workpiece machining control method provided in an embodiment of this application;

[0044] Figure 2 This is a structural schematic diagram of the robot composite gripper provided in this application;

[0045] Figure 3a This is a schematic diagram of the packaging of the workpiece turnover system provided in this application;

[0046] Figure 3b for Figure 3a A schematic diagram of the internal layout of the workpiece turnover system shown.

[0047] Figure 4 For this application Figure 3a A schematic diagram of the positioning and processing table in the workpiece turnover system shown.

[0048] Figure 5 For this application Figure 3a The diagram shows the structural schematic of the transfer mechanism in the workpiece turnover system.

[0049] Figure 6 For this application Figure 3a The diagram shows the structure of the material frame in the workpiece turnover system.

[0050] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concepts of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0051] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0052] This technology relates to workpiece turnover, gripping, positioning, and assembly in industrial automation. It is suitable for production scenarios involving the sorting, flipping, positioning, processing, and turnover of small sheet metal parts or similar plate-shaped workpieces. A typical system usually includes a robot, gripping and placing stations, several material frames, a flipping table, a positioning processing table, and image acquisition devices to complete workpiece transfer and processing connections between different stations. These image acquisition devices are usually multiple, for example, composed of 3D image acquisition devices and 2D image acquisition devices.

[0053] On existing production lines, robots typically retrieve workpieces from a gripping frame according to a preset path, then transport them to a processing or assembly position, and finally place them back into a designated frame after processing. For stacked workpieces, an image acquisition device is often used to obtain images of the stacked workpieces, and gripping, transporting, and placing actions are performed based on the recognition results. The basic principle is to achieve cyclical operation using preset coordinates, predetermined cycle times, and fixed process flows.

[0054] However, when workpieces within the gripping frame are stacked and exhibit random postures, layer-to-layer obstruction, slight deformation, or inconsistent front and back states, relying solely on a fixed path or single-shot recognition result often fails to accurately determine the actual position and appropriate gripping point of the target workpiece. This can easily lead to unstable gripping, placement deviations, or even inaccurate alignment of subsequent machining holes and positioning components. Especially when different types of workpieces are rotated along the same line, if the changing of the gripping frame still relies on manual intervention or lacks an automatic switching mechanism linked to the gripping and releasing cycle, it will further impact continuous production efficiency.

[0055] The aforementioned defects not only reduce the success rate of grasping and assembly accuracy, but also cause the robot to frequently pause or repeatedly pick up and place items under abnormal working conditions, increasing cycle time fluctuations and the number of manual interventions, thereby affecting the overall processing stability and turnover efficiency. Therefore, how to achieve accurate grasping, flipping judgment, positioning assembly, and continuous turnover of target workpieces under conditions of workpiece stacking, posture changes, and multiple workpiece cyclic switching has become an urgent technical problem to be solved.

[0056] In view of this, a workpiece turnover method is provided. After configuring the first material frame as a gripping material frame and the second material frame as a placement material frame, a three-dimensional image of the gripping material frame is first acquired to determine the target workpiece, its pose, and gripping point from multiple workpieces, and a flipping determination result is generated accordingly. When it is determined that flipping is required, the robot is controlled to place the target workpiece on the flipping table on one side of the gripping material frame and re-grab it. Then, it moves along a preset path to the positioning point, where a two-dimensional image is acquired to determine the position of the processing hole of the target workpiece. The robot is then controlled to align the target workpiece with the target positioning part of the positioning processing table and complete the push assembly and preset processing steps. Finally, the target workpiece is transferred to the placement material frame.

[0057] To address the need for multi-frame turnover in continuous production, this method also utilizes gripping and placing stations set on the left and right sides of the robot and transfer mechanisms corresponding to the placement frames. When the gripping frame is empty, the adjacent third frame is brought into the gripping and placing station, and the gripping frame and the placement frame are switched, thereby achieving continuous connection of gripping, flipping, positioning processing and frame turnover in different workpiece switching scenarios.

[0058] Figure 1 This is a flowchart illustrating the workpiece machining control method provided in the embodiments of this application, as shown below. Figure 1 As shown, the workpiece machining control method includes:

[0059] S101: After configuring the gripping frame and placing the frame, acquire a 3D image of the gripping frame.

[0060] To facilitate the explanation of the cyclical turnover process of this application, the first material frame and the second material frame can be configured as a gripping material frame and a placing material frame, respectively.

[0061] The first and second material frames are located at the gripping and placing stations on the left and right sides of the robot, respectively. The first material frame contains multiple workpieces, while the second material frame is empty.

[0062] The material frames, including a first material frame and a second material frame, are used to hold multiple workpieces. At a certain moment, the first and second material frames are transferred by a transfer mechanism to the gripping and placing stations, for example, two aligned stations on the left and right sides of the robot. If the first material frame contains multiple workpieces and the second material frame is currently empty, then the first material frame is the gripping material frame, and the second material frame is the placing material frame. It's important to understand that the gripping and placing material frames are simply two different roles for the material frames. As the process progresses, the roles of the material frames change. For example, after all the workpieces in the first material frame have been gripped, its role can switch to a placing material frame, while other material frames on the opposite side that contain workpieces become gripping material frames. The gripping and placing stations are the working positions where the robot performs the picking and placing actions. They are usually arranged on the left and right sides of the robot, one gripping and placing station on each side, allowing the robot to complete gripping, transferring, and placing within a working radius.

[0063] Three-dimensional images are used to characterize the spatial distribution, stacking relationship, surface contour, and mutual occlusion of multiple workpieces within a gripping frame, such as the first frame. These images serve as the foundational data for subsequent workpiece identification, pose calculation, and gripping point determination. The three-dimensional images of the gripping frame can be acquired by a three-dimensional image acquisition device positioned above the corresponding gripping station.

[0064] In practice, the control system initializes the roles of the workpiece frames before the start of a work cycle. The first workpiece frame, located on one side of the robot (e.g., the left-hand gripping station), is configured as the gripping frame, and the second workpiece frame, located on the other side (e.g., the right-hand gripping station), is configured as the placement frame. Workpieces in the first workpiece frame are stacked, while the second workpiece frame is initially empty and is used to receive workpieces processed in the current cycle. When acquiring 3D images, the spatial information of the workpiece stack within the gripping frame can be obtained, and the positional relationship corresponding to the robot's actions can be established for subsequent identification and gripping control.

[0065] Based on the above analysis, it can be seen that by acquiring a 3D image of the grabbing frame after the frame role is configured, the current workpiece stacking state can be mapped to the robot control process. Subsequent steps no longer rely on fixed pick-up coordinates, thus adapting to working conditions with random postures, interlayer occlusion, and slight deformation.

[0066] S102: Based on the 3D image, determine the target workpiece from multiple workpieces stacked in the gripping frame, and determine the pose and gripping point of the target workpiece.

[0067] In this step, the target workpiece is the specific workpiece selected from multiple workpieces within the first material frame, which will be gripped and rotated by the robot in the current cycle. Pose describes the position and orientation of the target workpiece in space, and may include the three-dimensional coordinates of the workpiece reference point and its rotation information relative to the workstation coordinate system. The gripping point is the location where the robot's end effector clamps or adheres to the target workpiece; it can be a single point, two points, or the center of a group of contact areas, and its result directly determines the robot's pick-up action and end effector posture.

[0068] In practice, the control system analyzes the 3D image obtained in the previous step, identifies the workpieces suitable for gripping within the material frame, and determines the workpieces that meet the gripping requirements as the target workpieces. After the target workpiece is determined, its pose is further determined based on its spatial information in the 3D image. The determination of the gripping point is correlated with the pose result. The control system determines a position suitable for the robot to perform the gripping action within the accessible area of ​​the target workpiece, and writes the target workpiece pose parameters and gripping point parameters into the task data for this cycle, for use in flipping judgment and robot pick-up execution.

[0069] Based on the above analysis, this step can select the actual graspable target workpiece from random poses and partial occlusion environments by performing real-time identification and pose analysis of the stacked workpieces, and output grasping point parameters that can be directly used for robot control, thus forming the premise for subsequent flipping judgment and positioning assembly.

[0070] S103: Generate a flipping determination result based on the pose and gripping point of the target workpiece.

[0071] In this step, the flipping determination result is used to determine whether the target workpiece, after being gripped in its current position and gripping point, needs to be repositioned and gripped again via a flipping table before entering the positioning point and positioning machining table for hole positioning and machining. The flipping determination result is divided into a first result and a second result based on whether flipping is required. The first result indicates that flipping is required, meaning the process needs to proceed to the flipping table; the second result indicates that flipping is not required, meaning the process can directly proceed to the subsequent hole positioning process. Grip simulation can be performed based on the target workpiece's position and gripping point to determine if the gripping is secure or if it affects the machining operation for the positioning hole. The simulation results are then used to determine whether flipping is necessary, resulting in the flipping determination result. When the gripping state of the target workpiece is deemed unsuitable for subsequent positioning, identification, or assembly, the flipping determination result is set as the first result; when the gripping state of the target workpiece meets the requirements of the subsequent process, the flipping determination result is set as the second result. Therefore, the flipping determination result is used to determine whether the target workpiece needs to have its gripping state adjusted via a flipping table before entering the subsequent process.

[0072] Based on the above analysis, it can be seen that by incorporating the pose and gripping point of the target workpiece into the flipping determination, it is possible to predetermine whether the pose or gripping state of the target workpiece needs to be adjusted via the flipping table before the gripping action is executed, so that the robot can determine the subsequent handling path.

[0073] S104: If the flipping determination result is the first result, then control the robot to grab the target workpiece and place the target workpiece on the flipping table on the side of the grabbing frame, and grab the target workpiece again. After grabbing the target workpiece again, control the robot to move along the preset path to the positioning point.

[0074] Among them, the robot is responsible for workpiece turnover. It can grasp and place workpieces through the robot's end effector, thereby realizing the cyclic turnover of workpieces between the material frame, the flipping table, and the positioning processing table.

[0075] When it is determined that flipping is necessary, i.e., when the flipping determination result is the first result, the robot needs to first place the gripped workpiece on the flipping table, which is located on the same side of the gripping frame as the robot. The flipping table is used to provide a transfer surface that can change the orientation of the workpiece or stabilize the re-gripping state of the workpiece.

[0076] The preset path is the calibrated motion trajectory of the robot from the flipping table or the gripping box to the positioning point, which is the reference position used for visual acquisition and pose verification. The positioning point can be the end point or an intermediate point of the preset path.

[0077] When the flipping result is the first result, the control system issues the first grasping command to the robot. The robot moves from the standby position to above the first material frame, adjusts its end effector posture according to the target workpiece's pose and the grasping point, descends along the approach direction, and completes adsorption or clamping. After picking up the workpiece, the robot first rises to a safe height, avoiding the edge of the material frame and surrounding equipment, and then moves along the configured path to the flipping table on the same side of the grasping frame. The flipping table surface can be equipped with limiting edges, positioning grooves, inclined guides, or friction support surfaces to keep the workpiece in a stable pose after the robot places it. After the robot places the target workpiece in the designated area of ​​the flipping table, it releases the clamping. The pose limiting structure of the flipping table makes the workpiece form a predetermined placement orientation, or the robot's wrist rotation action during placement makes the workpiece fall onto the table in an orientation different from the original grasping state. After placement, the robot approaches the target workpiece again according to the re-grasping reference position corresponding to the flipping table, and re-grabs it with a new grasping point or a new end effector posture.

[0078] Before the robot places the target workpiece it has grasped onto the flipping table, it can control the rotation angle of the robot's end effector based on the 3D image acquired after the robot grasps the target workpiece or based on the pose and grasping point of the target workpiece. After rotating the corresponding angle, the robot places the grasped workpiece onto the flipping table and waits for a short time, or controls the robot to return to the standby position and then controls the robot to grasp the target workpiece placed on the flipping table again.

[0079] After re-grabbing the target workpiece, the robot moves along a preset path to the positioning point. This preset path is defined in the control system by several key pose points, including at least the departure point from the flipping table, the transition obstacle avoidance point, and the arrival point at the positioning point. If the workpiece pose changes significantly after flipping, the control system can select the corresponding path branch based on the actual end effector posture after re-grabbing to ensure that the workpiece does not interfere with the camera bracket, the edge of the positioning processing table, or adjacent material frames during movement. Regardless of whether it passes the flipping table, the positioning point is set as a fixed visual reference position and matched with the field of view of the 2D image acquisition device.

[0080] Based on the above analysis, it is clear that the flipping table process is not simply about adding a transfer step, but rather about transforming a target workpiece that initially does not meet the requirements of subsequent processes into a holding state suitable for subsequent visual recognition and positioning assembly through a controlled placement and re-grip. In situations where the stacked workpieces have random poses, inconsistent front and back sides, or limited initial gripping points, this step directly determines whether subsequent recognition and assembly can proceed continuously, thus representing a key step in solving the corresponding technical problems. By setting up a flipping table on the side where the gripping frame is located and using a re-grip path, the robot can complete pose correction without returning to the manual workstation, ensuring a closed-loop production cycle within the automated process.

[0081] S105: After the robot moves to the positioning point, a two-dimensional image of the robot is acquired, and the position of the machining hole of the target workpiece is determined based on the two-dimensional image.

[0082] In this step, the two-dimensional image is used to reflect the planar visual information of the target workpiece at the positioning point, especially machined holes, edges, notches, or other planar identifiable features. The position of the machined hole is a core parameter required for subsequent alignment with the target positioning part on the positioning machining table. It includes at least the position of the center of the machined hole in the image coordinate system or the robot associated coordinate system, and if necessary, the hole axis direction, hole type, and relative relationship with the workpiece edge.

[0083] In practice, after the robot carries the target workpiece to the positioning point and pauses briefly to stabilize, the control system triggers the 2D image acquisition device to take a picture. The 2D image acquisition device can be fixedly installed below the positioning point or in a relatively fixed relationship with the robot's end effector. Before taking the picture, the workpiece surface can be illuminated with coaxial light, ring light, or lateral strip light through the light source assembly to enhance the grayscale contrast between the edge of the machined hole and the background of the board surface. For sheet metal parts with reflectivity, the exposure time, polarization conditions, and gain range can be set in the acquisition parameters to suppress the influence of bright areas on the identification of hole edges.

[0084] After acquiring two-dimensional images, the control system can first perform preprocessing such as distortion correction, grayscale normalization and field of view clipping, and then perform edge extraction, matching of round or irregular hole templates, contour closure judgment and center coordinate calculation to obtain the position of the machined hole of the target workpiece.

[0085] After the robot precisely moves and stops at the preset positioning point, completes posture calibration and position locking, a vision camera installed below the positioning point captures a high-definition two-dimensional image of the target workpiece. During the acquisition process, issues such as ambient light and shadow interference, image blurring, and distortion can be avoided by fixing exposure parameters, adjusting focus clarity, and enabling supplementary lighting and image stabilization, ensuring the integrity and accuracy of the original image data. After acquiring the two-dimensional image of the target workpiece grasped by the robot, preprocessing operations can be performed on the image, such as image grayscale conversion, noise filtering, edge enhancement, and distortion correction. This effectively removes invalid interference information such as noise, color variations, and light and shadow textures from the image, highlighting the workpiece contour and edge features of the machined holes, and improving the recognition of target features. Based on this, relying on pre-trained visual inspection algorithms and standard workpiece model templates, feature extraction and matching recognition are performed on the pre-processed images to accurately capture core feature information such as the circular outline, hole wall edge, and hole diameter boundary of the machined hole. Simultaneously, combining the reference coordinates of the robot's positioning points and the intrinsic and extrinsic parameter calibration data of the visual camera, a mapping transformation relationship between the image pixel coordinate system and the world physical coordinate system is established, accurately converting the pixel position information at the image level into the physical coordinate position in the actual working scene. To further ensure the accuracy of the machined hole position recognition, a sub-pixel edge fitting algorithm is used to refine the machined hole outline, compensating for the accuracy error of traditional pixel-level recognition. Simultaneously, verification and comparison are performed using the standard hole diameter parameters of the workpiece and the hole position layout reference to eliminate false holes, defective dents, and other misidentified targets. Finally, the center coordinates and hole position offset of the machined hole in the target workpiece are accurately located.

[0086] When the target workpiece contains multiple machined holes, the control system can first determine the candidate hole group based on the workpiece type template, then select the corresponding machined hole of the target workpiece from the candidate hole group according to the hole spacing and relative angle, and then determine the position of the machined hole of the target workpiece based on the theoretical parameters of the corresponding machined hole of the target workpiece and the acquired two-dimensional image.

[0087] Based on the calibration relationship between the 2D camera and the robot coordinate system, the image coordinates are converted into workstation plane coordinates to obtain the deviation of the center of the machined hole relative to the robot's current clamping posture. If the workpiece undergoes a slight rotation during movement, the system can re-estimate the workpiece plane rotation angle through the edge direction in the 2D image, so as to simultaneously obtain the positional and angular deviations of the machined hole on the target workpiece.

[0088] Based on the above analysis, this step performs a planar visual precision positioning at the positioning point, connecting the part-picking problem solved in the 3D gripping stage with the hole positions required in the processing and assembly stage. This allows the subsequent push assembly to be based on the real-time hole positions rather than fixed teaching points, thereby adapting to the cumulative errors caused by workpiece flipping, re-gripping, and handling.

[0089] S106: Based on the position of the machining hole of the target workpiece, control the robot to align the target workpiece with the target positioning part of the positioning machining table. After alignment, control the robot to push the target workpiece to complete the assembly and cooperation with the target positioning part, and execute the preset machining process.

[0090] In this step, the positioning machining table serves as a work platform for positioning, assembling, and subsequent processing of the workpiece. Multiple positioning components are installed on it, each corresponding to an assembly datum for different types of workpieces. The target positioning component is a specific component selected based on the target workpiece type and the location of the machining holes. Pushing is the mechanical action by which the robot advances the workpiece into position along the assembly direction after aligning the machining holes with the positioning components. The preset processing steps are the predetermined processing flows that need to be executed after the workpiece and target positioning components are assembled and mated. These may include pressing, riveting, punching, inspection, fastening, or other process actions corresponding to this positioning state.

[0091] In practice, the control system determines the target positioning component corresponding to the current target workpiece based on the obtained machining hole position and the known arrangement of positioning components on the positioning machining table. It then calculates the deviation between the machining hole of the target workpiece and the target positioning component, converting this deviation into a clamping pose correction amount for the robot's end effector. Through position and angle compensation, the machining hole is aligned with the target positioning component. After the robot reaches the aligned position, it performs a pushing action in low-speed mode, causing the machining hole of the target workpiece to fit, conform to, or press against the target positioning component, completing the assembly. During the pushing process, the robot's axial force, displacement changes, or end effector vacuum state can be monitored to confirm whether the workpiece has reached the set assembly depth and mating state.

[0092] After assembly and mating are completed, the control system triggers a preset processing procedure. If the processing equipment is integrated on the positioning processing table, the positioning processing table performs the corresponding processing action while the robot maintains the workpiece mating state; if the processing action is completed collaboratively by the robot, the robot drives the workpiece or processing tool to complete the procedure according to the preset trajectory while maintaining the assembly state. For the case of co-production of different types of workpieces, this step works in conjunction with the aforementioned material frame switching logic, enabling the same robot to dock with different target positioning parts in different cycles without interrupting production.

[0093] Based on the above analysis, this step directly uses the machining hole positions obtained from the 2D image for alignment control, and completes the accurate alignment between the workpiece and the positioning machining table through clamping posture correction and pushing assembly. In this way, even if the workpiece experiences posture errors during subsequent gripping, flipping, and handling, these errors can be compensated for before entering the machining station, thus ensuring that the preset machining process is based on the actual hole alignment. This step, together with step S105, constitutes a key link in achieving machining accuracy and has a direct effect on solving the problem of inaccurate alignment between the machining hole and the positioning component.

[0094] S107: After the preset processing steps are completed, control the robot to move the target workpiece to the placement frame.

[0095] In this step, the placement frame is the frame used to receive the processed workpiece in the current cycle, and corresponds to the second frame in the initial state. Transfer refers to the process by which the robot moves the target workpiece, which has completed assembly and preset processing steps, from the positioning processing table to the designated position in the placement frame. This process is connected to the next gripping cycle.

[0096] In practice, after receiving a pre-set processing completion signal, the control system controls the robot to release its holding action associated with the positioning processing table and lift the target workpiece from the target positioning point. If the workpiece needs to be placed in a specific positioning posture after processing, the system performs a rotation correction on the end effector before leaving the positioning processing table to ensure the workpiece is aligned with the stacking direction in the placement box. The robot then moves along a pre-set transport path to the top of the placement box. This path also consists of a safe departure point, a transition point, and a placement arrival point, avoiding the gripping box, the flipping table, and processing equipment. Upon reaching the placement position, the robot lowers the workpiece to a specified height and performs a release action based on the current workpiece height or empty space information in the placement box. This process is repeated until the workpieces stored in the gripping box are processed through the signed process and placed in the placement box. The placement box gradually accumulates workpieces from an initial empty box until the gripping box is empty (0 workpieces).

[0097] To ensure the accuracy of the gripping and material box switching judgment in the next cycle, the control system updates the loading status of the gripping material box after each placement, such as the number of remaining workpieces, and can also update the loading status of the placement material box, such as the number of workpieces already placed. When the number of remaining workpieces is 0, the material box role is switched, thereby switching the gripping material box.

[0098] Furthermore, after the target workpiece is transferred to the placement frame, a vision or height sensor can be used to monitor the placement effect. This involves verifying the actual placement of the target workpiece after release to confirm that there is no tilting, misalignment, or jamming. If an anomaly is detected, the system can immediately trigger the robot to perform a fine-tuning placement or anomaly cleanup process. For plate-shaped workpieces that have already been processed, the robot maintains a set acceleration and end-effector posture during transfer to prevent collisions with surrounding structures due to inertial swaying. Thus, a single target workpiece completes a full turnaround from being picked up from the placement frame, flipped for inspection, positioned and assembled, processed, to being returned to the placement frame.

[0099] Based on the above analysis, this step reliably returns the processed parts to the placement frame and simultaneously updates the frame's status, providing a traceable inventory basis for subsequent frame switching and continuous turnover. This step, together with the preceding steps, forms a closed-loop operation, eliminating the need for manual workpiece transfer during transit.

[0100] S108: When the gripping frame is empty, control the movement of the transfer mechanism corresponding to the placement frame so that the third frame adjacent to the placement frame is in the gripping and placement position; switch the gripping frame to the third frame and switch the placement frame to the original gripping frame.

[0101] The third material frame contains different workpieces than the original grabbing material frame, which is stacked in the first material frame.

[0102] In this step, the transfer mechanism is used to perform position switching between material frames, thereby switching the material frame located at the gripping and placing station. The third material frame is the next material frame adjacent to the currently placed material frame, and the workpieces stacked in it are different from those in the first material frame, thus corresponding to the continuous switching production of different types of workpieces.

[0103] Multiple material frames can be arranged above the transfer mechanism, and these frames can be arranged in a row, for example, along the front-to-back direction of the robot. The gripping and placing station can be a station aligned with the robot.

[0104] An empty gripper frame indicates that all workpieces in the current gripper frame (the first frame) have been retrieved. At this point, the roles of gripper and placement frames need to be reset without stopping the entire production line. Specifically, the role of the first frame is switched to placement frame, and the third frame adjacent to the second frame is configured as a gripper frame. Through a similar process described above, the workpieces in the third frame are moved to the first frame after positioning and processing. This cycle can be repeated to achieve the cyclic processing and turnover of various types of workpieces.

[0105] In practice, the control system determines whether the gripping frame is empty based on the remaining number of workpieces in the first frame, the number of identifiable workpieces in the 3D image, or the empty frame detection signal at the bottom of the frame. When an empty frame is detected, the control system sends a control command to the transfer mechanism associated with the currently placed frame, thereby driving the transfer mechanism to move. The transfer mechanism can be a chain-type frame transfer mechanism, a roller conveyor mechanism, a pallet traversing mechanism, or a servo push mechanism. Its objective is to move the third frame adjacent to the currently placed frame to a gripping and placing station where the robot can perform gripping and placing operations. The control system can verify the position of the third frame based on station sensors, limit switches, or visual confirmation signals, or determine the third frame's position after the transfer mechanism provides a completion signal. Subsequently, the original gripping frame is released, the third frame is registered as the new gripping frame, and the original first frame is switched to the new placement frame for subsequently processed workpieces to be placed.

[0106] In this way, after the material frame of the previous type of workpiece is vacated and becomes an empty material frame, the empty material frame is immediately used as the placement material frame for the next type of workpiece, and the material frame containing the workpiece adjacent to the original placement material frame, such as the third material frame, takes over as the new gripping material frame. The entire process does not require manual movement of material frames or re-teaching of the path.

[0107] Based on the above analysis, this step extends the turnover of a single workpiece into a continuous turnover process of multiple workpieces through empty frame detection, linkage of the transfer mechanism, and switching of the workpiece role. It also allows different workpieces to be switched sequentially within the same production line. This step directly solves the problem of relying on manual intervention for the cyclic switching of multiple workpieces, enabling continuous connection of gripping, flipping, positioning processing, and workpiece turnover within the robot's cycle time.

[0108] The workpiece turnover method provided in this application acquires three-dimensional images of the workpieces in the gripping frame at the gripping station, and determines the target workpiece, its pose, and gripping point based on the three-dimensional images. This allows for accurate selection of the workpiece to be processed from multiple randomly stacked workpieces, thereby improving the adaptability to gripping various types of workpieces and the gripping success rate. By generating a flipping determination result based on the pose and gripping point of the target workpiece, and controlling the robot to place the target workpiece on the flipping table and re-grip it when flipping is required, and then moving it to the positioning point, the method can adjust the posture of workpieces with different initial postures, thereby improving the subsequent positioning and processing. It provides a consistent assembly basis; by acquiring two-dimensional images and determining the position of the machining hole of the target workpiece after the robot moves to the positioning point, and then controlling the robot to align the target workpiece with the target positioning part of the positioning machining table, push and assemble it, and execute the preset machining process, the workpiece positioning accuracy and machining fit accuracy can be improved, thereby ensuring machining quality and operation stability; by transferring the target workpiece to the placement frame after the preset machining process is completed, and switching the correspondence between the gripping frame and the placement frame when the gripping frame is empty, continuous turnover between different workpiece frames can be achieved, thereby improving operation efficiency and production continuity.

[0109] Based on the aforementioned embodiments, further, a flipping determination result is generated based on the pose and gripping point of the target workpiece, including: calling a preset fixture model, estimating the gripping posture and contact area after the robot grips the target workpiece based on the pose and gripping point of the target workpiece; and determining the flipping determination result as the first result when the gripping posture interferes with the machining hole or the contact area is less than a preset area threshold.

[0110] In practical implementation, after the control system calls the fixture model based on the target workpiece's pose and gripping point, it can perform collision verification on the robot's end effector's spatial pose after gripping, and simultaneously calculate the contact area between the end effector and the target workpiece surface, and calculate the area of ​​the contact area to obtain the contact area. When the simulation results show that any part of the end effector coincides with, obstructs, or enters the preset safety gap range of the machining hole on the target workpiece, it is determined that the gripping posture interferes with the machining hole; when the simulation results show that the contact area between the end effector and the target workpiece is less than the preset area threshold, it indicates that the current gripping method is insufficient to form a stable grip. At this time, the output flipping determination result is the first result.

[0111] If the gripping posture does not interfere with the machining hole and the contact area is greater than or equal to the preset area threshold, the flipping determination result is the second result, that is, there is no need to flip the workpiece, and the robot can be controlled to directly move the gripped target workpiece to the preset positioning point.

[0112] By modeling and estimating the clamping posture and contact area before gripping, and using the interference of the machining hole and the contact area threshold as the basis for flipping judgment, the flipping judgment can be consistent with the actual gripping feasibility. This allows subsequent gripping, positioning, and hole assembly to be performed under the condition of meeting the space avoidance conditions, thereby improving the accuracy of posture judgment and gripping stability during workpiece turnover.

[0113] In one possible implementation, the method further includes:

[0114] If the flipping result is the second result, then control the robot to grab the target workpiece and move along the preset path.

[0115] In this embodiment, the second result indicates that the target workpiece can proceed to the subsequent process without being flipped. Upon receiving this flipping determination result, the control system does not trigger any actions related to the flipping table. Instead, it maintains the current gripping state of the target workpiece and drives the robot to continue running according to the trajectory data of a preset path stored in the motion controller. The preset path can be a fixed movement path for the robot to reach the subsequent positioning point after gripping the target workpiece, or it can be a motion trajectory generated based on the pose of the target workpiece and the gripping point. The trajectory can be implemented using joint interpolation or Cartesian space interpolation to ensure smooth movement of the robot's end effector during transport.

[0116] In practical implementation, after the robot grasps the target workpiece, the control system outputs corresponding pose-maintaining commands based on the grasping posture of the target workpiece. This ensures that the end effector gripper or adsorption component maintains its gripping relationship with the workpiece during movement, preventing any additional changes in the workpiece's pose. The starting point of the preset path can be set as the current position after grasping, and the ending point can be set as the subsequent positioning point or placement position. The coordinates, velocity, and acceleration parameters of each inflection point in the path are pre-written into the control program, allowing the robot to directly enter the next workstation without needing to flip the workpiece.

[0117] By moving directly along a preset path when the determination result is the second result, the robot can quickly send the target workpiece that meets the conditions to the subsequent processing position and maintain grasping stability before entering the subsequent workstation.

[0118] By adopting the above method, when the target workpiece does not need to be flipped, the system can directly drive the robot to execute the preset path movement based on the second result, reducing unnecessary flipping and picking actions, so that the handling process after grasping is naturally connected with the subsequent positioning, processing or placement links, thereby making the workpiece turnover process more continuous and stable.

[0119] In one possible implementation, multiple material frames are distributed on the left and right sides of the robot, each material frame being used to stack workpieces of a corresponding type; the positioning processing table includes multiple positioning components, each positioning component corresponding to a type of workpiece. Based on the position of the machining hole of the target workpiece, the robot is controlled to align the target workpiece with the target positioning component of the positioning processing table, including: determining the target positioning component from the multiple positioning components based on the type of the target workpiece; and adjusting the clamping posture of the grasped target workpiece during or after the robot moves to the corresponding position of the target positioning component, based on the position of the machining hole of the target workpiece, so that the machining hole of the target workpiece is aligned with the target positioning component.

[0120] In this solution, multiple material frames are positioned on the left and right sides of the robot. These frames can be constructed from welded steel plates or aluminum profiles. Each frame holds the same type of plate-shaped workpiece, facilitating the robot's handling and turnover based on workpiece type. The positioning processing table is a rigid platform mounted on the frame, equipped with multiple positioning components. These components can consist of positioning pins, guide blocks, or limit blocks, corresponding to the machining hole positions of different workpiece types. The positioning components are fixed to the table surface with screws or positioning seats for easy replacement and calibration. The target positioning component is determined based on the target workpiece type. After receiving workpiece type information from the vision recognition module, the control system selects the target positioning component matching that type from a preset mapping relationship and then generates alignment instructions based on the center coordinates of the machining hole.

[0121] As the robot approaches the positioning machining table carrying the target workpiece, the control system further adjusts the gripping posture of the robot's end effector based on the position of the machining hole. The gripper can be a pneumatic parallel gripper or an electric gripper, and its gripping surface can be equipped with elastic anti-slip pads to keep the workpiece stable during adjustment.

[0122] This structure enables the control system to first lock the corresponding target positioning part according to the workpiece type, and then finely adjust the pose of the target workpiece according to the actual position of the machining hole, so that the center of the machining hole is consistent with the axis or reference surface of the positioning part. Then the robot can push the target workpiece to complete the assembly and enter the preset processing procedure.

[0123] Through the aforementioned settings, a collaborative control relationship is formed between the selection of the target positioning component and the alignment of the machining hole. The workpiece type is used to determine the reference target positioning component, and the machining hole position is used to complete the end pose compensation, so that different types of workpieces can be accurately positioned on the same machining table, improving the consistency of assembly and positioning stability, and reducing subsequent machining deviations.

[0124] In one possible implementation, the method further includes:

[0125] The type of target workpiece is determined based on a two-dimensional image or the material frame identifier of the gripping frame.

[0126] The two-dimensional image is the image captured by a two-dimensional image acquisition device at the positioning point. The material frame marking of the gripping frame can be identified by the three-dimensional image captured by the three-dimensional image acquisition device above the gripping frame, or by the barcode scanning device mounted on the robot or deployed above the gripping frame.

[0127] When determining the type of a target workpiece based on a two-dimensional image, the control system can compare the contour boundary, opening shape, bending shape, and local identification features of the target workpiece extracted from the two-dimensional image with pre-stored image templates, feature libraries, or classification models, thereby identifying the type of the target workpiece.

[0128] In another implementation, each material frame is marked with a unique identifier corresponding to the type of workpiece it contains. The control system uses a barcode reader, visual recognition camera, or RFID reader to scan the barcode and obtain the identifier information. Based on this identifier, the system queries a pre-defined mapping table to determine the type of the target workpiece. This method is suitable for scenarios where workpieces have similar shapes but the source of the material frames is clearly distinguishable. After outputting the target workpiece type, the control system can directly call the positioning and assembly logic matching that type to complete subsequent alignment control.

[0129] During operation, the type recognition result is combined with the machined hole position recognition result of the target workpiece, enabling the control system to quickly determine the target positioning part from multiple positioning parts and adjust the robot's gripping posture and alignment accordingly. After determining the type using two-dimensional image recognition or material frame identification, the system can maintain the continuity of the recognition link during workpiece or material frame switching, reducing manual intervention and enabling different types of workpieces to correspond to different gripping, positioning, and assembly control parameters.

[0130] By adopting the above method, the system can directly identify the workpiece type when there is sufficient two-dimensional image information, and can also quickly determine the workpiece type when the material frame is clearly marked, thereby providing a basis for the selection of subsequent target positioning parts, and ensuring consistency and continuity in the turnover, positioning and assembly control of different types of workpieces.

[0131] In one possible implementation, the method further includes:

[0132] Based on the type and / or geometric features of the target workpiece, the robot's gripper is switched to grasp the target workpiece using the switched gripper; the geometric features are extracted based on the 3D image.

[0133] In practice, the robot's gripper is a composite gripper, which includes multiple interchangeable grippers. The gripping end structure, opening size, contact surface shape, or adsorption layout of the different grippers are different from each other, so as to adapt to workpieces of different sizes, thicknesses, or outer contours.

[0134] When the control system determines that the current fixture does not match the target workpiece based on the type of the target workpiece or the geometric features extracted from the 3D image, it drives the robot to the fixture change position to complete the fixture switch. After the switch is completed, the new fixture is used to perform the gripping action. Fixture change can be achieved through a quick-change connection structure, which may include a mounting base, locking components, and positioning components. The robot end effector and the fixture body maintain connection stability through mechanical positioning and locking, facilitating rapid switching between different fixtures.

[0135] During operation, the 3D image acquisition device first performs 3D acquisition of the gripping frame, and extracts the outer contour, corner shape, or concave and convex features of the target workpiece from the 3D image. Then, combined with the type information, it determines the appropriate fixture mode for the workpiece, controls the robot to switch to the corresponding fixture, and completes the gripping. In this way, the robot can select a matching gripping method based on the workpiece type and spatial shape, and link the gripping action with the workpiece recognition result.

[0136] Figure 2 This is a structural schematic diagram of the robot composite gripper provided in this application, as shown below. Figure 2 As shown, the robot's end effector is equipped with a composite gripper, which includes a switching cylinder, a suction end, and a gripping end. The switching cylinder is connected to both the suction end and the gripping end for driving them to extend alternately to the work position. The suction end adopts a negative pressure adsorption structure, which is suitable for adsorbing and picking up flat sheet metal workpieces. The gripping end is a claw-type structure, which is used to clamp irregularly shaped and bent sheet metal workpieces.

[0137] By adopting the above method, the robot can automatically select appropriate fixtures for workpieces of different types or geometric shapes, reduce gripping offset, slippage or gripping failure caused by mismatch between fixture and workpiece, and enable stacked workpieces to be gripped stably in complex poses, thereby improving the adaptability and continuity of workpiece turnover.

[0138] This application also provides a control device, including: a memory and a processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory, causing the processor to perform the method provided above.

[0139] By storing the workpiece turnover method in memory as computer-executed instructions and having it called and executed by the processor, the control equipment can uniformly complete the following tasks: 3D image acquisition of the gripping frame, determination of the target workpiece pose and gripping point, flipping judgment, placement of the flipping table and re-gripping, 2D image positioning, alignment, pushing and assembly, and control of processing and placing the gripping frame for transfer. This integrates visual recognition, motion control, and station switching. The equipment can adaptively handle the randomness of the posture of stacked workpieces, occlusion, and differences between the front and back sides, thereby improving gripping stability, positioning accuracy, and processing and assembly consistency. Furthermore, combined with the control logic for switching empty workpiece frames and the entry of adjacent workpiece frames into the gripping and placing station, manual intervention and downtime are reduced, improving the stability of the production cycle and the overall automation efficiency in scenarios with continuous turnover of multiple workpiece frames.

[0140] In the specific implementation process, the processor executes computer execution instructions stored in the memory, causing the processor to perform the method provided in the above embodiments.

[0141] The specific implementation process of the processor can be found in the above method embodiments, and its implementation principle and technical effect are similar, so it will not be repeated here.

[0142] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0143] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0144] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0145] Figure 3a This is a schematic diagram of the packaging of the workpiece turnover system provided in this application. Figure 3b for Figure 3a The diagram shows the internal layout of the workpiece turnover system, combined with... Figure 3a and Figure 3b It can be seen that the workpiece turnover system includes a robot, a positioning processing table, a 3D image acquisition device, a 2D image acquisition device (not shown in the figure, located at the positioning point), a flipping table, a control unit, and two working modules located on the left and right sides of the robot; each working module includes a transfer mechanism and multiple material frames set above the transfer mechanism. Figure 3b This corresponds to situations where the contents need to be flipped over.

[0146] See also Figure 3a In the workpiece turnover system, each component is enclosed within a sealed cabinet. The cabinet's side walls are made of transparent panels, allowing for direct observation of the robot, material frames, and positioning processing station's operational status without disassembling the cabinet. A human-machine interface display is mounted on the front of the cabinet, connected to the control unit, enabling operators to view the equipment's operating status and switch workpiece processing models. A 3D image acquisition device is installed on the top of the cabinet, with its field of view covering the entire material frame area on both sides, allowing for real-time acquisition of the 3D shape data of the scattered workpieces within the current gripping and placing station.

[0147] The material frames are configured to hold multiple workpieces, with at least some frames holding different workpieces. A transfer mechanism is used to move multiple material frames on it to switch the material frames located at the gripping and placing station. A 3D image acquisition device is used to acquire 3D images of the material frames located at the gripping and placing station. A 2D image acquisition device is set at the positioning point to acquire 2D images of the workpiece after the robot grips the workpiece and moves to the positioning point. A positioning processing table is located in the hollow area between the two work modules. Multiple positioning components are arranged on the positioning processing table and configured to assemble and cooperate with the corresponding type of workpiece gripped by the robot to complete the workpiece processing. A control unit is used to execute the above methods to control the robot to turn the workpieces between the material frames, the flipping table, and the positioning processing table, and to control the transfer mechanism to move multiple material frames on the corresponding side to switch the material frames located at the gripping and placing station.

[0148] A flipping table is positioned between the corresponding side transfer mechanism and the robot. At least one flipping table can be configured on each side to allow the robot to perform the flipping operation of the workpiece being gripped. The structures of each flipping table can be identical, differing only in their deployment locations. This allows the robot to select the nearest flipping table for flipping, or to adapt different flipping tables to different types of workpieces, thus selecting the flipping table that matches the type of workpiece being gripped when flipping is required.

[0149] A robot is an automated actuator used to perform gripping, transporting, flipping, and placing actions between a material frame, a turning table, and a positioning processing table. It undertakes the active operation of workpiece transfer in the overall system and uses an end effector to clamp, suction, or hold workpieces to adapt to different workpiece gripping methods. The robot is preferably installed in the center of the work area, with its working radius covering the left and right work modules, the turning table, and the positioning processing table located in the hollow area, so as to complete multi-station connections without changing the main body's installation position. In one possible embodiment, the robot can be a six-axis industrial robot, a SCARA (Selective Compliance Assembly Robot Arm), or a multi-joint robotic arm, or it can be replaced by a gantry transport mechanism, a parallel transport mechanism, or other automated execution equipment suitable for repetitive positioning. Its body typically consists of a base, articulated arm segments, and an end flange, on which vacuum suction cups, mechanical grippers, magnetic suction devices, or composite grippers can be installed. It should be understood that the above examples are for illustrative purposes only and are not limiting. The robot's size and working range should match the width of the material frame, the size of the flipping table, and the outer boundary of the positioning processing table. Its arm span is usually greater than the maximum distance from the single-sided working module to the center of the hollow area to ensure that the end effector can cover the gripping point, the flipping table receiving position, and the positioning component assembly position when in the position.

[0150] A positioning and machining table is a platform structure located in the hollow area between two work modules, used to support workpiece positioning, assembly, and subsequent processing. Its function is to provide an assembly reference for the workpiece after it has been grasped by the robot and to form a positioning and assembly fit with the corresponding type of workpiece. The positioning and machining table corresponds to the robot's workspace, the positioning points of the 2D image acquisition device, and the re-grabbing path after flipping. It is typically fixedly installed on a ground foundation, frame beam, or independent support frame to ensure stability during operation. In one possible embodiment, the positioning and machining table can be a plate-type table, a frame-type table, or a jig table, or it can be made as a cast iron platform, a welded steel structure table, or an aluminum profile assembly table. Multiple positioning components arranged on its surface can be designed to achieve positioning and assembly fit. The materials of the positioning components can be hardened steel, alloy steel, stainless steel, or wear-resistant engineering materials to maintain dimensional stability during repeated assembly. The planar dimensions of the positioning and machining table are usually larger than the maximum external dimensions of a single workpiece. The arrangement of the positioning components is adapted to the corresponding type of workpiece, enabling the workpiece to be assembled and positioned after it is in place, and providing a reference surface for subsequent processing steps.

[0151] Figure 4 For this application Figure 3a The diagram shows the structure of the positioning and machining stage in the workpiece turnover system. Figure 4 Taking a positioning machining table that includes three types of positioning machining parts as an example, such as Figure 4 As shown, the positioning processing table consists of a base plate, supporting columns, and a positioning panel. Three sets of independent positioning processing components are arranged on the positioning panel, corresponding to flat sheet metal parts, right-angle sheet metal parts, and irregularly shaped sheet metal parts, respectively. Each positioning component includes a positioning pin and a limiting block. The positioning pin is used to insert into the corresponding workpiece's machining hole to achieve hole positioning, and the limiting block conforms to the workpiece's side contour to complete shape limitation. Flat sheet metal parts are adapted to the central rectangular positioning component, right-angle sheet metal parts are adapted to the left-side bent positioning component, and irregularly shaped sheet metal parts are adapted to the right-side arc-shaped contour positioning component, enabling precise assembly and positioning of three different types of sheet metal parts on the same positioning processing table.

[0152] A 3D image acquisition device is a visual inspection unit used to perform 3D imaging of the material frame within the gripping station and output spatial pose information. It is used to identify the contour, surface height difference, and relative pose of the target workpiece under conditions of workpiece stacking, random poses, or occlusion between upper and lower layers, thereby generating executable gripping points and gripping postures for the robot. This device is typically installed above or to the side of the gripping station, ensuring its field of view covers the upper surface of the material frame and the stacked area. It communicates with the control unit to output depth maps, point cloud maps, or 3D reconstruction results. In one possible embodiment, the 3D image acquisition device can employ a structured light camera, a binocular vision camera, a laser contour scanning device, or a Time-of-Flight (ToF) depth camera. Alternatively, a monocular camera combined with calibration algorithms can be used to generate 3D perception results. Its installation position should avoid the robot's motion trajectory and the material frame switching stroke, while ensuring a complete imaging range of the inner boundary of the material frame, the upper surface of the workpiece, and the grippable edges. The 3D measurement accuracy should be adapted to the gripping height range above the material frame, so that the control unit can determine the workpiece stacking level, the target workpiece posture, and the approach direction of the end effector based on the imaging data.

[0153] A two-dimensional image acquisition device is a vision unit installed at a positioning point to acquire planar images and perform position correction after the robot has carried the workpiece to its position. Its function is to obtain the planar position deviation of the workpiece to compensate for the cumulative errors in the robot's handling, flipping, or assembly processes. This device is typically positioned within the visible area of ​​the robot after it reaches the positioning point, with its optical axis aligned with the plane of the workpiece held by the robot's end effector, so that imaging can be completed when the workpiece is docked or suspended at the positioning point. In one possible embodiment, the two-dimensional image acquisition device can be an industrial area array camera, a CCD (Charge-Coupled Device) camera, or a CMOS (Complementary Metal-Oxide-Semiconductor) camera, and can be used in conjunction with a ring light source, a bar light source, or a coaxial light source to form stable illumination. Its installation method can be fixed bracket mounting, beam suspension mounting, or integrated mounting with the positioning point. The visual field should cover the imaging area of ​​the workpiece at the positioning point, and the image resolution is set according to the positioning accuracy requirements so that the control unit can calculate the offset, rotation angle, and alignment relationship between the workpiece and the positioning processing table using the planar image.

[0154] A flipping table is an intermediate process platform located within the robot's reach, used for flipping, changing the surface, or repositioning workpieces. Its function is to provide a transitional position where the workpiece's front and back surfaces do not match the processing requirements, allowing for repeated receiving, flipping, and re-gripping. The flipping table is typically located in the continuous working area between the gripping station and the positioning processing table, and its height, boundaries, and receiving posture are adapted to the robot's end effector so that the workpiece can be placed stably and re-gripped. In one possible embodiment, the flipping table can be a rotary flipping table, a lifting flipping table, a table with a flipping clamping mechanism, or a transitional support with a guide surface. It can also employ a structure with side guards, flipping grooves, or rotatable pallets. Its materials can be welded steel plates, aluminum alloy frame parts or high-strength engineering plastic parts. The flipping support surface can be equipped with wear-resistant pads, rubber buffer layers or low-friction coatings to meet the requirements that the workpiece surface is not easily damaged. The table size of the flipping table should be larger than the external size of a single workpiece and leave enough edge margin for the robot to grasp and place, so that the workpiece can maintain a gripping posture after flipping and enter the subsequent positioning process.

[0155] The control unit is the core of the system, responsible for the unified scheduling and status assessment of the robot, vision device, flipping table, positioning processing table, and transfer mechanism. It generates gripping, flipping, positioning, placement, and material frame switching commands based on 3D images, 2D images, and equipment status information, and interlocks the action sequence of each actuator. The control unit can be housed in an electrical control cabinet, industrial computer, or edge control cabinet, and connected to the robot controller, vision acquisition device, transfer mechanism driver, and flipping table driver via wired communication or an industrial network. In one possible embodiment, the control unit can be implemented using a PLC (Programmable Logic Controller), industrial computer, embedded controller, or distributed control system, internally storing workpiece type recognition rules, pose solving programs, path planning programs, and workstation switching logic. Structurally, the control unit can also be linked with safety sensors, limit switches, position detection devices, and abnormal alarm modules to interrupt actions and replan when the material frame is not in place, workpiece recognition fails, or positioning deviation exceeds limits.

[0156] The two work modules are turnover units located on the left and right sides of the robot, respectively responsible for the loading and switching of material frames. Their function is to provide independently switchable material frame sets for different workpiece types or different process stages, enabling the robot to perform gripping, placement, and cyclical transfer between the left and right sides. The two work modules are typically arranged symmetrically or approximately symmetrically around the robot, forming a continuous work path together with the spatial positions of the gripping and placing stations, the material receiving position of the flipping table, and the hollow area of ​​the positioning processing table. In one possible embodiment, the work modules can be arranged in a left-right double-sided gantry configuration, a ground rail configuration, or a vertical support frame configuration. Their structural dimensions should be coordinated with the robot's arm span, material frame switching stroke, and workpiece transport trajectory, ensuring that any material frame in either work module, after switching to the gripping and placing station, does not interfere with the robot body or surrounding workstations. Furthermore, the left and right work modules can be configured with the same or different types of material frames to meet the needs of multi-product co-line turnover.

[0157] The transfer mechanism is a mechanical transmission unit installed within the working module, used to drive multiple material frames on it to translate along a predetermined direction and complete the switching of gripping and releasing positions. Its function is to move different material frames sequentially to a position that the robot can grasp under the command of the control unit, thereby realizing automatic frame changing without manual handling. The transfer mechanism is usually installed on the bottom guide surface, side support, or load-bearing beam of the working module, and is connected to the support part, slider, or positioning seat at the bottom of the material frame to ensure stability and repeatability during translation. In one possible embodiment, the transfer mechanism can be a guide rail slide, belt slide, lead screw slide, linear motor drive mechanism, or rack and pinion drive mechanism. Its guiding direction is generally parallel to the robot's working surface. The stroke length is determined based on the number of material frames on the same side, the width of a single material frame, and the center position of the gripping station. Multiple material frames can be arranged side by side along a straight line, or they can be arranged in a stepped manner through layered brackets or segmented slides. After the material frame is switched, the upper surface, edge reference, and gripping area of ​​the target material frame should be accurately placed within the robot's gripping range and kept consistent with the field of view of the three-dimensional image acquisition device.

[0158] Figure 5 For this application Figure 3a The schematic diagram of the transfer mechanism in the workpiece turnover system is shown below. Figure 5As shown, the transfer mechanism is the same as the transfer mechanism in the previous work module. The transfer mechanism includes a transfer module, a servo motor, and a bellows cover. The servo motor is connected to the transfer module and is used to drive the transfer module to perform horizontal linear translation. The transfer module carries at least two material frames. By translating the module, different material frames are switched to the corresponding gripping and placing positions of the robot. The bellows cover is arranged along the movement stroke of the transfer module and completely covers the internal transmission guide rail of the module. It is used to prevent workpiece debris and dust from entering the guide rail pair and to achieve dust protection for the transmission structure. Different material frames can be loaded with flat, right-angled, and irregularly shaped sheet metal parts respectively. When the material frame of one gripping and placing position is finished, the servo motor drives the transfer module to translate and switch the adjacent material frames with different workpieces to the working position to realize continuous turnover processing of multiple types of workpieces.

[0159] A material frame is a container used to stack multiple workpieces and move along the transfer mechanism as the workstation switches. Its function is to orderly stack, temporarily store, and transfer workpieces to be grabbed, flipped, or placed. The material frame is set above the transfer mechanism and forms a detachable or fixed connection with it so that it can be moved horizontally to the grab and place station as a whole during the switching process. In one possible embodiment, the material frame can be a metal material frame, a plastic material frame, or a composite material material frame, and can also be made into a rectangular frame type, a pallet type, or a structure with side wall enclosures. The specific materials can be stamped steel plate welded frame, aluminum alloy profile frame, or high-strength engineering plastic frame. The material frame can be equipped with partition ribs, positioning blocks, buffer pads or anti-slip pads to adapt to the stacking state of different types of workpieces. At least some material frames can hold different workpieces, so the external dimensions, holding depth and stacking spacing of the material frame can be set according to the corresponding workpiece specifications. The planar dimensions of the material frame are usually slightly larger than the maximum outline of the workpiece in order to leave room for gripping and visual recognition. At the same time, its height should take into account the number of workpieces stacked and the robot gripping depth requirements, so that the upper layer of the target workpiece can be clearly identified by the three-dimensional image acquisition device.

[0160] Figure 6 For this application Figure 3a The schematic diagram of the material frame in the workpiece turnover system shown is as follows: Figure 6 As shown, the material frame is a square frame structure with its side walls sloping inwards. Figure 6 Taking right-angled sheet metal parts as an example, the inclined sidewalls can prevent right-angled, flat, and irregularly shaped sheet metal parts from getting stuck at the corners of the inner wall of the material frame. The workpieces can slide smoothly down the inclined sidewalls to the bottom of the material frame, greatly improving the clearing rate of the frame after the robot picks up the material. The material frame is placed on the transfer module. The same transfer module can carry at least two material frames. Different types of sheet metal workpieces can be stacked in different material frames. With the help of the servo motor to drive the transfer module to move horizontally, the automatic switching of the material frames at the gripping and placing station can be realized.

[0161] Based on the above analysis, the workpiece turnover system provided in this application, during operation, first drives the corresponding side transfer mechanism to switch the material frame containing the workpiece to the gripping and placing station. After the three-dimensional image acquisition device performs spatial imaging of the material frame, the control unit determines the pose, grippable edge, and gripping point of the target workpiece based on depth information and stacking relationship, and guides the robot to the corresponding gripping posture. After the robot completes gripping, the control unit determines whether flipping is required based on the workpiece type and current facing state. When flipping is required, the robot places the target workpiece on the flipping table, and after the flipping table completes flipping, posture adjustment, or reorientation, it grips again and moves along the preset path to the positioning point. When the robot reaches the positioning point, the two-dimensional image acquisition device takes a picture of the workpiece plane. The control unit calculates the positional deviation based on the image information and compensates for the robot's end-effector posture to align the workpiece with the target positioning part on the positioning processing table. Subsequently, the robot pushes the workpiece to the positioning processing table to complete assembly and execute the preset processing steps. After processing or assembly, the robot sends the target workpiece to the work module where the placement frame is located, completing placement and turnover. When the workpiece in the gripping frame is consumed to an empty state, the control unit further drives the adjacent frame to enter the gripping and placing station through the transfer mechanism, and simultaneously completes the switching between the gripping frame and the placement frame, so that the left and right work modules can alternately supply and receive materials according to the rhythm. The system determines the spatial pose of the stacked workpieces through 3D vision, corrects the positioning deviation through 2D vision, completes the workpiece posture transformation through the flipping table, realizes the assembly and cooperation between the workpiece and the processing datum through the positioning parts on the positioning processing table, and completes the automatic switching of multiple frames through the transfer mechanism. A continuous closed-loop turnover link is formed between the workstations, so that the system can adapt to the working conditions of random workpiece posture, inconsistent front and back states, and co-line switching of multiple types of frames, and makes gripping, flipping, positioning and processing actions continuously connected under unified control.

[0162] In one possible implementation, the workpiece includes flat, right-angled, and irregularly shaped sheet metal parts; multiple positioning elements include positioning elements that respectively match the outer contour and machining holes of the flat, right-angled, and irregularly shaped sheet metal parts.

[0163] In this application, the workpiece type is defined as a structural classification of objects to be turned over, positioned, and processed. Flat workpieces refer to sheet metal parts with relatively regular outlines, basically flat surfaces, and uniform thickness. Right-angled workpieces refer to sheet metal parts with bent edges, corner edges, or right-angled outlines such as L-shapes and T-shapes. Irregularly shaped sheet metal parts refer to sheet metal parts with irregular or non-standardized outer contours, notches, edges, or hole arrangements. The corresponding positioning components are mating reference structures set on the positioning processing table. The positioning components match the outlines and processing holes of the sheet metal parts, right-angled parts, and irregularly shaped sheet metal parts, respectively. They are used to limit the outline, guide the edges, and align the holes of the workpiece after the robot delivers the target workpiece to the positioning point, so that different types of workpieces can complete positioning assembly and subsequent processing within the same workpiece turnover system. For example, flat workpieces can be positioned by planar support positioning parts corresponding to their planar boundaries; right-angled workpieces can be positioned by corner limiting positioning parts adapted to the shape of bent edges or corners; irregularly shaped sheet metal parts can be positioned by irregularly shaped contour positioning parts that match their irregular outer contours and pin-type positioning parts corresponding to the machining holes. The positioning parts can be set on the surface of the positioning processing table in the form of cylindrical pins, tapered pins, limiting blocks, side blocks, elastic pressure blocks or combination positioning pins, and their working end faces, guide slopes and positioning base surfaces are all adapted to different workpiece types.

[0164] When the system starts, the control unit calls the corresponding positioning component combination according to the type information of the workpiece to be processed. During the process of the robot moving the workpiece from the gripping station to the positioning processing table, the contour limiting component on the table first constrains the shape of the workpiece, and then the processing holes on the workpiece are gradually fitted into the corresponding positioning pins or tapered pins, thereby completing the posture correction and benchmark establishment. For flat workpieces, after the robot places them stably on the corresponding planar support positioning component, the lateral stop blocks restrict their in-plane displacement, and the positioning pins are inserted into the preset holes to complete the assembly benchmark determination. For right-angled workpieces, the bent edge or corner of the workpiece first contacts the corner limiting component, and the guide ramp makes the workpiece synchronously abut in the horizontal and vertical directions, and then the hole positioning component completes the final locking. For irregularly shaped sheet metal parts, the contour-customized positioning block guides the irregular edges segment by segment, and cooperates with one or more positioning pins to engage with the processing holes to achieve stable positioning under irregular shape conditions. Based on the above-mentioned cooperation relationship, it can be seen that different types of workpieces can obtain positioning constraints that match their shape and hole features in a unified workpiece turnover system, thereby reducing assembly errors caused by contour differences, bending shapes or hole offsets, reducing the number of times the robot repeatedly adjusts and manually corrects, and thus improving the compatibility of turnover of multi-category sheet metal parts, positioning repeatability accuracy and alignment stability of subsequent processing.

[0165] The positioning and processing table is equipped with multiple sets of independent positioning components. These components are adapted to three types of sheet metal workpieces: flat, right-angled, and irregular shapes. Each set of positioning components matches the outer contour and machining hole of the corresponding workpiece, enabling multiple workpieces of different specifications to share the same set of positioning fixtures to complete assembly and positioning.

[0166] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods provided above.

[0167] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the method provided above.

[0168] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0169] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0170] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0171] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0172] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0173] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0174] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0175] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for workpiece turnover, characterized in that, include: After configuring the first material frame as a gripping material frame and the second material frame as a placing material frame, a three-dimensional image of the gripping material frame is acquired; the first material frame and the second material frame are respectively located at the gripping and placing stations on the left and right sides of the robot, the first material frame contains multiple workpieces, and the second material frame is an empty material frame; Based on the three-dimensional image, a target workpiece is determined from multiple workpieces stacked in the gripping frame, and the pose and gripping point of the target workpiece are determined. Based on the pose and gripping point of the target workpiece, a flipping determination result is generated; If the flipping determination result is the first result, then control the robot to grab the target workpiece and place the target workpiece on the flipping table on the side of the grabbing frame, and grab the target workpiece again. After grabbing the target workpiece again, control the robot to move along the preset path to the positioning point. After the robot moves to the positioning point, a two-dimensional image of the robot is acquired, and the position of the machining hole of the target workpiece is determined based on the two-dimensional image; Based on the position of the machining hole of the target workpiece, the robot is controlled to align the target workpiece with the target positioning part of the positioning processing table. After alignment, the robot is controlled to push the target workpiece to complete the assembly and cooperation with the target positioning part and execute the preset processing steps. After the preset processing steps are completed, the robot is controlled to transfer the target workpiece to the placement frame; When the gripping frame is empty, the transfer mechanism corresponding to the placement frame is controlled to move so that the third frame adjacent to the placement frame is in the gripping and placing position; the gripping frame is switched to the third frame, and the placement frame is switched to the first frame, wherein the workpieces stacked in the third frame are different from those in the first frame.

2. The method according to claim 1, characterized in that, The step of generating a flipping determination result based on the pose and gripping point of the target workpiece includes: By calling a preset fixture model, and based on the pose and gripping point of the target workpiece, the gripping posture and contact area of ​​the robot after gripping the target workpiece are estimated. When the gripping posture interferes with the processing hole or the contact area is less than a preset area threshold, the flipping determination result is determined to be the first result.

3. The method according to claim 1, characterized in that, The method further includes: If the flipping determination result is the second result, then control the robot to grab the target workpiece and move along the preset path.

4. The method according to claim 1, characterized in that, The robot has multiple material frames distributed on its left and right sides, each frame for stacking workpieces of a corresponding type; the positioning processing table includes multiple positioning components, each corresponding to a type of workpiece. The step of controlling the robot to align the target workpiece with the target positioning component of the positioning processing table based on the position of the machining hole of the target workpiece includes: Based on the type of the target workpiece, the target positioning element is determined from the plurality of positioning elements; Based on the position of the machining hole of the target workpiece, during or after the robot moves to the corresponding position of the target positioning component, the clamping posture of the grasped target workpiece is adjusted so that the machining hole of the target workpiece is aligned with the target positioning component.

5. The method according to claim 4, characterized in that, The method further includes: The type of the target workpiece is determined based on the two-dimensional image or the material frame identifier of the gripping frame.

6. The method according to any one of claims 1-5, characterized in that, The method further includes: Based on the type and / or geometric features of the target workpiece, the robot's gripper is switched to grasp the target workpiece using the switched gripper; the geometric features are extracted based on the three-dimensional image.

7. A control device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-6.

8. A workpiece turnover system, characterized in that, include: The robot, positioning processing table, three-dimensional image acquisition device, two-dimensional image acquisition device, flipping table, control unit, and two working modules respectively located on the left and right sides of the robot; Each of the aforementioned operating modules includes a transfer mechanism and multiple material frames disposed above the transfer mechanism; The material frames are configured to hold multiple workpieces, with at least some material frames holding different workpieces; The transfer mechanism is used to drive multiple material frames on it to translate, so as to switch the material frames located at the gripping and releasing station; the three-dimensional image acquisition device is used to acquire three-dimensional images of the material frames located at the gripping and releasing station. The two-dimensional image acquisition device is set at the positioning point and is used to acquire a two-dimensional image of the robot after the robot moves to the positioning point to grasp the workpiece. The positioning processing table is located in the hollow area between the two working modules. The positioning processing table is equipped with multiple positioning components, which are configured to assemble and cooperate with the corresponding type of workpiece grasped by the robot to complete the workpiece processing. The control unit is used to execute the method according to any one of claims 1-6, to control the robot to rotate workpieces between the material frame, the flipping table and the positioning processing table, and to control the transfer mechanism to drive multiple material frames on the corresponding side to move, so as to switch the material frame located at the gripping and placing station.

9. The system according to claim 8, characterized in that, The workpieces include flat, right-angled, and irregularly shaped sheet metal parts; The plurality of positioning components include positioning components that respectively match the outer contour and machining holes of plate-shaped, right-angled, and irregularly shaped sheet metal parts.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method described in any one of claims 1-6.

Images