Large-curvature workpiece polishing track generation method, device, equipment and medium

Abstract

The application provides a large-curvature workpiece polishing track generation method, device, equipment and medium, the large-curvature workpiece polishing track generation method includes generating ideal polishing track according to the three-dimensional model of the workpiece to be polished;Collect multi-element environmental perception information of the workpiece being polished;According to the multi-element environmental perception information, the workpiece is positioned and the robot is positioned;According to the workpiece positioning result and the robot positioning result, the real-time relative pose of the workpiece and the robot is obtained, the ideal polishing track is corrected according to the real-time relative pose of the workpiece and the robot, and the actual polishing track is generated, the accurate relative position of the workpiece to be polished and the robot is obtained by collecting the multi-element environmental perception information of the workpiece being polished, no additional physical clamp is needed, the workpiece placement accuracy requirement is reduced, the operation efficiency is improved, and furthermore, through automatic generation and real-time correction of the polishing track, offline programming and manual teaching process are avoided, and the polishing efficiency and flexibility are further improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing technology, and in particular to a method, apparatus, equipment and medium for generating grinding trajectories for workpieces with large curvature. Background Technology

[0002] During manufacturing, workpieces often develop burrs, flash, and other excess material, affecting subsequent machining processes and the workpiece's appearance. Therefore, grinding is necessary. Existing automated grinding systems typically rely on high-precision custom fixtures for workpiece positioning, combined with offline programming or drag-and-drop teaching to generate the grinding trajectory. Robots are usually stationary or mounted on guide rails, completing the grinding task through multi-axis linkage. However, existing implementations have several problems: for example, workpiece positioning depends on specialized fixtures, requiring precise clamping, which is time-consuming, labor-intensive, and lacks flexibility. Furthermore, the grinding trajectory requires offline programming or manual drag-and-drop teaching. Different workpiece sizes require manual trajectory setting, resulting in significant human intervention, low processing efficiency, and particularly tedious manual planning for highly curved workpiece surfaces, which is prone to workpiece positioning errors and inaccurate grinding trajectory positions. Summary of the Invention

[0003] This invention provides a method, apparatus, equipment, and medium for generating grinding trajectories for workpieces with large curvature, which solves the defects of traditional grinding trajectory generation methods that rely on high-precision tooling fixtures, require manual setting of grinding trajectories, involve large amounts of manual intervention, have low processing efficiency, and result in inaccurate grinding trajectory positions for workpieces with large curvature.

[0004] This invention provides a method for generating grinding trajectories for workpieces with large curvature, comprising:

[0005] Generate an ideal grinding trajectory based on the 3D model of the workpiece to be ground;

[0006] Collect multi-dimensional environmental perception information of the workpiece being polished;

[0007] Workpiece and robot positioning are performed based on the aforementioned multi-dimensional environmental perception information;

[0008] The real-time relative pose of the workpiece and the robot is obtained based on the workpiece positioning result and the robot positioning result. The ideal grinding trajectory is corrected based on the real-time relative pose of the workpiece and the robot to generate the actual grinding trajectory.

[0009] According to the present invention, a method for generating a grinding trajectory for a workpiece with large curvature is provided, wherein generating an ideal grinding trajectory based on a three-dimensional model of the workpiece to be ground includes:

[0010] Point cloud sampling is performed on the surface to be polished based on the CAD model of the workpiece to be polished, and the point cloud sampling results are projected onto the XY plane.

[0011] The surface of the workpiece to be polished is analytically fitted using a quadratic parametric surface function. The outer contour of the workpiece after analytical fitting is extracted in the XY plane to obtain the outer contour of the workpiece point cloud on the XY plane.

[0012] The outer contour of the workpiece point cloud on the XY plane is offset, and the offset contour point cloud is back-projected to obtain the ideal grinding trajectory of the workpiece to be ground.

[0013] According to the present invention, a method for generating a grinding trajectory for a workpiece with large curvature is provided, wherein the acquisition of multi-dimensional environmental perception information of the workpiece being ground includes:

[0014] Image information of two-dimensional calibration objects in the environment where the workpiece is being polished is acquired by multiple 2D cameras, and point cloud information of three-dimensional calibration objects in a large-scale environment where the workpiece is being polished is acquired by multiple three-dimensional vision sensors.

[0015] Feature extraction and feature registration are performed on the image information of the two-dimensional calibration object and the point cloud information of the three-dimensional calibration object to calibrate the three-dimensional vision sensor;

[0016] The point cloud information collected by the calibrated 3D vision sensor is converted to the same coordinate system, and the point cloud information in the same coordinate system is fused to obtain the multi-dimensional environmental perception information of the workpiece being polished.

[0017] According to the present invention, a method for generating grinding trajectories for workpieces with large curvature is provided, wherein the workpiece positioning based on the multivariate environmental perception information includes:

[0018] The multi-dimensional environmental perception information of the workpiece being polished is segmented using Euclidean clustering.

[0019] Extract the bounding box data corresponding to the segmented point cloud data to obtain the point cloud model of the workpiece to be polished;

[0020] The point cloud model of the workpiece to be polished is registered with the three-dimensional model of the workpiece to be polished to obtain the actual pose of the workpiece to be polished.

[0021] According to the present invention, a method for generating a grinding trajectory for a workpiece with large curvature, wherein the robot localization based on the multivariate environmental perception information includes:

[0022] Based on the actual pose of the workpiece to be polished and the bounding box data, the multi-dimensional environmental perception information of the workpiece to be polished is filtered and segmented to obtain robot point cloud blocks.

[0023] The robot model is coarsely located using the robot laser synchronous localization and mapping method to obtain the robot model;

[0024] The robot model is registered with the robot point cloud block using the iterative nearest point algorithm to obtain the robot's actual pose.

[0025] According to the present invention, a method for generating a grinding trajectory for a workpiece with large curvature includes correcting the ideal grinding trajectory based on the real-time relative pose of the workpiece and the robot, comprising:

[0026] The coordinates of the workpiece to be polished in the world coordinate system are obtained based on the real-time relative pose of the workpiece and the robot.

[0027] Based on the coordinates of the workpiece to be polished in the world coordinate system, the coordinate system of the ideal polishing trajectory is transformed to correct the ideal polishing trajectory and obtain the actual polishing trajectory.

[0028] According to the present invention, a method for generating grinding trajectories for workpieces with large curvature is provided, wherein the robot is a composite robot, and the composite robot includes an automated guided vehicle and a robotic arm.

[0029] The present invention also provides a device for generating grinding trajectories for workpieces with large curvature, comprising:

[0030] The generation module is used to generate an ideal grinding trajectory based on the three-dimensional model of the workpiece to be ground;

[0031] The data acquisition module is used to acquire multi-dimensional environmental perception information of the workpiece being polished.

[0032] The positioning module is used to perform workpiece positioning and robot positioning based on the multi-dimensional environmental perception information.

[0033] The correction module is used to obtain the real-time relative pose of the workpiece and the robot based on the workpiece positioning and the robot positioning, and to correct the ideal grinding trajectory based on the real-time relative pose of the workpiece and the robot to generate the actual grinding trajectory.

[0034] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for generating grinding trajectories for large curvature workpieces as described above.

[0035] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for generating grinding trajectories for large curvature workpieces as described above.

[0036] This invention provides a method, apparatus, equipment, and medium for generating grinding trajectories for workpieces with large curvature. It generates an ideal grinding trajectory based on a 3D model of the workpiece; collects multi-dimensional environmental perception information of the workpiece; performs workpiece and robot positioning based on the multi-dimensional environmental perception information; obtains the real-time relative pose of the workpiece and robot based on the workpiece and robot positioning results; and corrects the ideal grinding trajectory based on the real-time relative pose of the workpiece and robot to generate the actual grinding trajectory. By collecting multi-dimensional environmental perception information of the workpiece to obtain the accurate relative position of the workpiece and robot, it eliminates the need for additional physical fixtures, reduces the accuracy requirements for workpiece placement, improves operational efficiency, and is more suitable for grinding multiple types of small batches of parts. Furthermore, by automatically generating and correcting the grinding trajectory in real time, it avoids offline programming and manual teaching processes, further improving grinding efficiency and flexibility. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0038] Figure 1 This is one of the flowcharts illustrating the method for generating grinding trajectories for workpieces with large curvature provided by the present invention;

[0039] Figure 2 This is the second flowchart illustrating the method for generating grinding trajectories for workpieces with large curvature provided by the present invention.

[0040] Figure 3(a) is a schematic diagram of the CAD model of the workpiece provided by the present invention;

[0041] Figure 3(b) is a schematic diagram of point cloud projection onto the XY plane provided by the present invention;

[0042] Figure 3(c) is a schematic diagram of the theoretical grinding trajectory provided by the present invention;

[0043] Figure 4 This is a schematic diagram of the relative pose relationship of the calibration objects in the system provided by the present invention;

[0044] Figure 5 This is the third flowchart of the method for generating grinding trajectories for large curvature workpieces provided by the present invention;

[0045] Figure 6 This is the fourth flowchart of the method for generating grinding trajectories for large curvature workpieces provided by the present invention;

[0046] Figure 7This is a schematic diagram of the structure of the grinding trajectory generation device for large curvature workpieces provided by the present invention;

[0047] Figure 8 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0049] Figure 1 The flowchart of the method for generating grinding trajectories for workpieces with large curvature provided in the embodiments of the present invention is as follows: Figure 1 As shown, the method for generating grinding trajectories for workpieces with large curvature provided in this embodiment of the invention includes:

[0050] Step 101: Generate an ideal grinding trajectory based on the 3D model of the workpiece to be ground;

[0051] Step 102: Collect multi-dimensional environmental perception information of the workpiece being polished;

[0052] Step 103: Perform workpiece and robot positioning based on multi-dimensional environmental perception information;

[0053] Step 104: Obtain the real-time relative pose of the workpiece and the robot based on the workpiece positioning and robot positioning, correct the ideal grinding trajectory based on the real-time relative pose of the workpiece and the robot, and generate the actual grinding trajectory.

[0054] Traditional methods for generating workpiece grinding trajectories typically rely on high-precision custom fixtures for workpiece positioning, combined with offline programming or drag-and-drop teaching to generate the grinding trajectory. Robots are usually stationary or mounted on guide rails, completing the grinding task through multi-axis linkage. However, existing implementations have many problems: for example, workpiece positioning depends on proprietary fixtures, the clamping process requires certain precision, the setup is time-consuming and labor-intensive, and lacks flexibility. Furthermore, the grinding trajectory needs to be set manually through offline programming or manual drag-and-drop teaching. For workpieces of different sizes, manual trajectory setting is required, resulting in significant human intervention, low processing efficiency, and particularly tedious manual planning for workpieces with large curvature surfaces, which is prone to workpiece positioning errors, leading to inaccurate grinding trajectory positions.

[0055] This invention provides a method for generating grinding trajectories for workpieces with large curvature. The method generates an ideal grinding trajectory based on a 3D model of the workpiece; collects multi-dimensional environmental perception information of the workpiece; performs workpiece and robot positioning based on the multi-dimensional environmental perception information; obtains the real-time relative pose of the workpiece and robot based on the workpiece and robot positioning; and corrects the ideal grinding trajectory based on the real-time relative pose of the workpiece and robot to generate the actual grinding trajectory. By collecting multi-dimensional environmental perception information of the workpiece to obtain the accurate relative position of the workpiece and robot, no additional physical fixtures are required, reducing the accuracy requirements for workpiece placement and improving operational efficiency. This method is more suitable for grinding multiple types of small batches of parts. Furthermore, by automatically generating and correcting the grinding trajectory in real time, offline programming and manual teaching processes are avoided, further improving grinding efficiency and flexibility.

[0056] Based on any of the above embodiments, generating an ideal grinding trajectory according to the three-dimensional model of the workpiece to be ground includes:

[0057] Step 201: Based on the CAD model of the workpiece to be polished, perform point cloud sampling on the surface to be polished, and project the point cloud sampling results onto the XY plane;

[0058] Step 202: Use a quadratic parametric surface function to perform analytical fitting on the surface of the workpiece to be polished, and extract the analytically fitted outer contour of the workpiece in the XY plane to obtain the workpiece point cloud outer contour on the XY plane.

[0059] Step 203: Offset the outer contour of the workpiece point cloud on the XY plane, and then back-project the offset contour point cloud to obtain the ideal grinding trajectory of the workpiece to be ground.

[0060] The theoretical grinding trajectory of the workpiece to be ground is automatically generated using a static trajectory generation algorithm, such as... Figure 2 As shown in Figure 3(a), point cloud sampling is first performed on the surface to be polished based on the CAD model of the workpiece (Figure 3(a)). The point cloud is then projected onto the XY plane (Figure 3(b)). A quadratic parametric surface function is used to analytically fit the surface to be polished. Then, the outer contour of the point cloud in the XY plane is obtained through contour extraction. An ordered outer contour is obtained through a point cloud sorting algorithm. The ordered outer contour point cloud is offset using a polygon shrinkage algorithm. The offset contour point cloud is then back-projected to obtain the polishing trajectory of the workpiece surface to be polished, thus guiding the polishing operation of the composite robot. The theoretical polishing trajectory is shown in Figure 3(c).

[0061] Based on any of the above embodiments, multi-dimensional environmental perception information of the workpiece being polished is collected, including:

[0062] Step 301: Collect image information of two-dimensional calibration objects in the environment where the workpiece is located using multiple 2D cameras, and collect point cloud information of three-dimensional calibration objects in the environment where the workpiece is located using multiple three-dimensional vision sensors.

[0063] Step 302: Perform feature extraction and feature registration on the image information of the two-dimensional calibration object and the point cloud information of the three-dimensional calibration object to calibrate the three-dimensional vision sensor;

[0064] Step 303: Convert the point cloud information collected by the calibrated 3D vision sensor to the same coordinate system, and fuse the point cloud information in the same coordinate system to obtain multi-dimensional environmental perception information of the workpiece being polished.

[0065] In this embodiment of the invention, by fusing point cloud information collected by multiple three-dimensional vision sensors, a wide range of environmental perception information can be obtained, improving the accuracy and comprehensiveness of environmental perception.

[0066] In this embodiment of the invention, physical reflectors and other components are used to mark the coordinate information of the calibrated objects; a grinding scene map is established using a composite robot laser SLAM (Simultaneous Localization and Mapping); the pose relationship between the scene map and other sensor coordinate systems is established using the marking information from the reflectors; and the multi-sensory signals collected by multiple sensors are transformed into a unified coordinate system, ensuring multi-sensory perception of environmental information within a certain size range, and obtaining the relative pose relationships in the system, such as... Figure 4 As shown.

[0067] Based on any of the above embodiments, workpiece positioning based on multi-dimensional environmental perception information includes:

[0068] Step 401: Perform Euclidean clustering to segment the multi-dimensional environmental perception information of the workpiece being polished.

[0069] Step 402: Extract the bounding box data corresponding to the segmented point cloud data to obtain the point cloud model of the workpiece to be polished;

[0070] Step 403: Register the point cloud model of the workpiece to be polished with the three-dimensional model of the workpiece to be polished to obtain the actual pose of the workpiece to be polished.

[0071] In embodiments of the present invention, such as Figure 5As shown, the output point cloud of the 3D camera undergoes coordinate transformation based on the calibration results. Then, the output point cloud is uniformly sampled to obtain a fused point cloud. Through Z-axis filtering, Euclidean clustering segmentation, and point cloud recognition, real-time point cloud data of the fixture to be polished is acquired, and the workpiece bounding box data is extracted. The workpiece bounding box data is registered with the workpiece's CAD model to obtain the workpiece's actual pose. This achieves workpiece positioning without relying on physical fixtures, reducing hardware costs and improving the efficiency and flexibility of workpiece positioning.

[0072] Based on any of the above embodiments, robot localization based on multi-dimensional environmental perception information includes:

[0073] Step 501: Filter and segment the multi-dimensional environmental perception information of the workpiece to be polished based on the actual pose of the workpiece and the bounding box data to obtain robot point cloud blocks.

[0074] Step 502: Perform coarse localization of the robot model using the robot laser synchronous localization and mapping method to obtain the robot model;

[0075] Step 503: Use the iterative nearest point algorithm to register the robot model with the robot point cloud blocks to obtain the robot's actual pose.

[0076] In this embodiment of the invention, the fused point cloud is further filtered and segmented based on the actual pose of the workpiece and the bounding box data to obtain partial point cloud blocks of the robot. Coarse localization of the robot model is performed using the laser SLAM module of the composite robot, and then the robot model is registered with the acquired point cloud using an iterative nearest-point algorithm to obtain the robot's actual pose.

[0077] In traditional grinding methods, the workpiece size is generally small due to the limited workspace of the grinding robot, making it difficult to grind large parts. While the combination of a guide rail and robot or a robot and turntable can slightly expand the workspace, the flexibility remains insufficient. In this embodiment of the invention, the robot is a composite robot, comprising an automated guided vehicle (AGV) and a robotic arm. The AAV can move in any direction, allowing the robot to perform grinding operations quickly and efficiently along a grinding trajectory.

[0078] Based on any of the above embodiments, the ideal grinding trajectory is corrected according to the real-time relative pose of the workpiece and the robot, including:

[0079] Step 601: Obtain the coordinates of the workpiece to be ground in the world coordinate system based on the real-time relative pose of the workpiece and the robot;

[0080] Step 602: Based on the coordinates of the workpiece to be ground in the world coordinate system, perform pose transformation on the coordinate system of the ideal grinding trajectory to correct the ideal grinding trajectory and obtain the actual grinding trajectory.

[0081] After obtaining the actual grinding trajectory, the motion planning algorithm of the robotic arm can be invoked to execute the grinding task. The specific flowchart of the method for generating grinding trajectories for large curvature workpieces provided in this embodiment of the invention is as follows: Figure 6 As shown:

[0082] (1) Generate an ideal grinding trajectory based on the CAD model of the workpiece to be ground;

[0083] (2) Build a multi-sensor system based on the size range of the workpiece and complete the calibration;

[0084] (3) Workpiece positioning is achieved through point cloud acquisition, fusion, segmentation, recognition, and registration;

[0085] (4) Combine tooling pose information to perform point cloud segmentation and recognition of the composite robot, and correct the relative pose of the robot and the workpiece.

[0086] (5) The grinding trajectory is corrected according to the positioning and posture of the workpiece and the robot to obtain an accurate actual grinding trajectory to guide the robot to grind.

[0087] The present invention provides a method for generating grinding trajectories for workpieces with large curvature. This method generates a grinding trajectory in an ideal pose using the workpiece's CAD model. It then combines this with laser SLAM of a composite robot and multiple high-precision 3D cameras capturing environmental information of the grinding scene. Point cloud filtering, segmentation, and registration algorithms are used to accurately locate the workpiece and the composite robot. The resulting real-time relative pose of the workpiece and robot is used to correct the grinding trajectory, transforming the ideal grinding trajectory to the robot's base coordinate system to obtain the actual grinding trajectory and guide the robot in the grinding operation. This invention uses a multi-sensor system to obtain the accurate relative position of the tooling to be ground and the composite robot, eliminating the need for additional physical fixtures, reducing the precision requirements for workpiece placement, improving operational efficiency, and making it more suitable for grinding multiple types of small-batch parts. Furthermore, the automatic generation and real-time correction algorithms for grinding trajectories avoid offline programming and manual teaching processes, further improving the efficiency and flexibility of the grinding system. On the other hand, the robot adopts a composite robot form of AGV (Automated Guided Vehicle) + robotic arm, expanding the workspace of a single robot and enabling automated grinding of large-sized workpieces.

[0088] The grinding trajectory generation device for large curvature workpieces provided by the present invention is described below. The grinding trajectory generation device for large curvature workpieces described below can be referred to in correspondence with the grinding trajectory generation method for large curvature workpieces described above.

[0089] Figure 7 This is a schematic diagram of the grinding trajectory generation device for large curvature workpieces provided in an embodiment of the present invention, as shown below. Figure 7 As shown, the grinding trajectory generation device for large curvature workpieces provided in this embodiment of the invention includes:

[0090] The generation module 701 is used to generate an ideal grinding trajectory based on the three-dimensional model of the workpiece to be ground;

[0091] The acquisition module 702 is used to acquire multi-dimensional environmental perception information of the workpiece being polished;

[0092] The positioning module 703 is used for workpiece positioning and robot positioning based on multi-dimensional environmental perception information.

[0093] The correction module 704 is used to obtain the real-time relative pose of the workpiece and the robot based on the workpiece positioning and the robot positioning, correct the ideal grinding trajectory based on the real-time relative pose of the workpiece and the robot, and generate the actual grinding trajectory.

[0094] Based on any of the above embodiments, the generation module 701 is configured as follows:

[0095] Point cloud sampling is performed on the surface to be polished based on the CAD model of the workpiece to be polished, and the point cloud sampling results are projected onto the XY plane.

[0096] The surface of the workpiece to be polished is analytically fitted using a quadratic parametric surface function. The outer contour of the workpiece after analytical fitting is extracted in the XY plane to obtain the outer contour of the workpiece point cloud on the XY plane.

[0097] The outer contour of the workpiece point cloud on the XY plane is offset, and the offset contour point cloud is back-projected to obtain the ideal grinding trajectory of the workpiece to be ground.

[0098] Based on any of the above embodiments, the acquisition module 702 is configured as follows:

[0099] Image information of two-dimensional calibration objects in the environment where the workpiece is being polished is acquired by multiple 2D cameras, and point cloud information of three-dimensional calibration objects in a large-scale environment where the workpiece is being polished is acquired by multiple three-dimensional vision sensors.

[0100] Feature extraction and feature registration are performed on the image information of the two-dimensional calibration object and the point cloud information of the three-dimensional calibration object to calibrate the three-dimensional vision sensor;

[0101] The point cloud information collected by the calibrated 3D vision sensor is converted to the same coordinate system, and the point cloud information in the same coordinate system is fused to obtain the multi-dimensional environmental perception information of the workpiece being polished.

[0102] Based on any of the above embodiments, the positioning module 703 is configured as follows:

[0103] The multi-dimensional environmental perception information of the workpiece being polished is segmented using Euclidean clustering.

[0104] Extract the bounding box data corresponding to the segmented point cloud data to obtain the point cloud model of the workpiece to be polished;

[0105] The point cloud model of the workpiece to be polished is registered with the three-dimensional model of the workpiece to be polished to obtain the actual pose of the workpiece to be polished.

[0106] Based on any of the above embodiments, the positioning module 703 is further configured as follows:

[0107] Based on the actual pose of the workpiece to be polished and the bounding box data, the multi-dimensional environmental perception information of the workpiece to be polished is filtered and segmented to obtain robot point cloud blocks.

[0108] The robot model is coarsely located using the robot laser synchronous localization and mapping method to obtain the robot model;

[0109] The robot model is registered with the robot point cloud block using the iterative nearest point algorithm to obtain the robot's actual pose.

[0110] Based on any of the above embodiments, the correction module 704 is configured as follows:

[0111] The coordinates of the workpiece to be polished in the world coordinate system are obtained based on the real-time relative pose of the workpiece and the robot.

[0112] Based on the coordinates of the workpiece to be polished in the world coordinate system, the coordinate system of the ideal polishing trajectory is transformed to correct the ideal polishing trajectory and obtain the actual polishing trajectory.

[0113] This invention provides a grinding trajectory generation device for workpieces with large curvature. It generates an ideal grinding trajectory based on a 3D model of the workpiece; collects multi-dimensional environmental perception information of the workpiece; performs workpiece and robot positioning based on the multi-dimensional environmental perception information; obtains the real-time relative pose of the workpiece and robot based on the workpiece positioning and robot positioning results; and corrects the ideal grinding trajectory based on the real-time relative pose of the workpiece and robot to generate the actual grinding trajectory. By collecting multi-dimensional environmental perception information of the workpiece to obtain the accurate relative position of the workpiece and robot, it eliminates the need for additional physical fixtures, reduces the accuracy requirements for workpiece placement, improves operational efficiency, and is more suitable for grinding multiple types of small batches of parts. Furthermore, by automatically generating and correcting the grinding trajectory in real time, it avoids offline programming and manual teaching processes, further improving grinding efficiency and flexibility.

[0114] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 8 As shown, the electronic device may include a processor 810, a communication interface 820, a memory 830, and a communication bus 840. The processor 810, communication interface 820, and memory 830 communicate with each other via the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute a method for generating a grinding trajectory for a workpiece with large curvature. This method includes: generating an ideal grinding trajectory based on a three-dimensional model of the workpiece to be ground; acquiring multi-dimensional environmental perception information of the workpiece; performing workpiece positioning and robot positioning based on the multi-dimensional environmental perception information; obtaining the real-time relative pose of the workpiece and robot based on the workpiece positioning result and the robot positioning result; correcting the ideal grinding trajectory based on the real-time relative pose of the workpiece and robot; and generating an actual grinding trajectory.

[0115] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, 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 described in the various embodiments of the present 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.

[0116] On the other hand, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements a method for generating a grinding trajectory for a large curvature workpiece provided by the above methods. The method includes: generating an ideal grinding trajectory based on a three-dimensional model of the workpiece to be ground; acquiring multi-dimensional environmental perception information of the workpiece; performing workpiece positioning and robot positioning based on the multi-dimensional environmental perception information; obtaining the real-time relative pose of the workpiece and the robot based on the workpiece positioning result and the robot positioning result; correcting the ideal grinding trajectory based on the real-time relative pose of the workpiece and the robot; and generating an actual grinding trajectory.

[0117] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and 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 modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0118] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0119] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for generating a polishing trajectory for a large curvature workpiece, the method comprising: The method comprises the following steps: ​ generating an ideal polishing track according to a three-dimensional model of a workpiece to be polished; collecting multi-environment perception information of the polishing workpiece; positioning the workpiece and a robot according to the multi-environment perception information; obtaining real-time relative poses of the workpiece and the robot according to the positioning results of the workpiece and the robot, correcting the ideal polishing track according to the real-time relative poses of the workpiece and the robot, and generating an actual polishing track; the step of correcting the ideal polishing track according to the real-time relative poses of the workpiece and the robot comprises: obtaining coordinates of the workpiece to be polished in a world coordinate system according to the real-time relative poses of the workpiece and the robot; performing pose transformation on a coordinate system in which the ideal polishing track is located according to the coordinates of the workpiece to be polished in the world coordinate system, so as to correct the ideal polishing track and obtain the actual polishing track.

2. The method of claim 1, wherein, the step of generating the ideal polishing track according to the three-dimensional model of the workpiece to be polished comprises: sampling point clouds on the surface of the workpiece to be polished according to a CAD model of the workpiece to be polished, and projecting the point cloud sampling result onto an XY plane; analytically fitting the surface of the workpiece to be polished by using a quadratic parametric surface function, extracting the outer contour of the workpiece point cloud in the XY plane after the analytical fitting, and obtaining the workpiece point cloud outer contour on the XY plane; offsetting the workpiece point cloud outer contour on the XY plane, and inversely projecting the offset contour point cloud to obtain the ideal polishing track of the workpiece to be polished.

3. The method of claim 1, wherein, the step of collecting the multi-environment perception information of the polishing workpiece comprises: collecting image information of two-dimensional calibration objects in the environment where the polishing workpiece is located by using multiple 2D cameras, and collecting three-dimensional calibration object point cloud information in the environment where the polishing workpiece is located by using multiple three-dimensional vision sensors; performing feature extraction and feature registration on the image information of the two-dimensional calibration objects and the three-dimensional calibration object point cloud information, so as to calibrate the three-dimensional vision sensors; converting the point cloud information collected by the calibrated three-dimensional vision sensors to the same coordinate system, and fusing the point cloud information in the same coordinate system to obtain the multi-environment perception information of the polishing workpiece.

4. The polishing path generation method for a large-curvature workpiece according to claim 1 or 3, characterized by, the step of positioning the workpiece according to the multi-environment perception information comprises: performing Euclidean clustering segmentation on the multi-environment perception information of the polishing workpiece; extracting bounding box data corresponding to the segmented point cloud data to obtain a point cloud model of the workpiece to be polished; registering the point cloud model of the workpiece to be polished with the three-dimensional model of the workpiece to be polished to obtain the actual pose of the workpiece to be polished.

5. The method of claim 4, wherein, the step of positioning the robot according to the multi-environment perception information comprises: filtering and segmenting the multi-environment perception information of the polishing workpiece according to the actual pose of the workpiece to be polished and the bounding box data to obtain a robot point cloud block; coarsely positioning a robot model according to a robot laser simultaneous localization and mapping method to obtain the robot model; registering the robot model with the robot point cloud block by using an iterative closest point algorithm to obtain the actual pose of the robot.

6. The method of claim 1, wherein, The robot is a composite robot, and the composite robot comprises an automatic guided vehicle and a mechanical arm.

7. A large curvature workpiece polishing trajectory generation device, characterized by, ​ The generating module is configured to generate an ideal polishing track according to a three-dimensional model of a workpiece to be polished. The collecting module is configured to collect multi-element environment perception information of the workpiece to be polished. The positioning module is configured to perform workpiece positioning and robot positioning according to the multi-element environment perception information. The correcting module is configured to obtain real-time relative poses of the workpiece and the robot according to the workpiece positioning and the robot positioning, correct the ideal polishing track according to the real-time relative poses of the workpiece and the robot, and generate an actual polishing track. The correcting the ideal polishing track according to the real-time relative poses of the workpiece and the robot includes: obtaining coordinates of the workpiece to be polished in a world coordinate system according to the real-time relative poses of the workpiece and the robot. The ideal polishing track is corrected according to the coordinates of the workpiece to be polished in the world coordinate system to obtain the actual polishing track.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the large-curvature workpiece polishing track generation method according to any one of claims 1 to 6. 9.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the large-curvature workpiece polishing track generation method according to any one of claims 1 to 6.

Images