Polishing device and polishing method for repairing damage of curved surface component

Abstract

The invention discloses a grinding device and a grinding method for curved surface component damage repair, and belongs to the technical field of machining equipment. The polishing method for curved surface component damage repair comprises the following steps that S1, hand-eye calibration and tool calibration are conducted, specifically, a transformation matrix of a 3D visual system coordinate system relative to a mechanical arm tail end center coordinate system is obtained, and a transformation matrix of a tool coordinate system relative to the mechanical arm tail end center coordinate system is obtained; s2, setting a polishing initial state: setting the polishing initial state of the mechanical arm by utilizing a dragging teaching function; s3, polishing path planning, wherein a point cloud model of the surface of the workpiece is obtained, and polishing path point information is obtained; s4, a mechanical arm grinding program is generated, specifically, a mechanical arm grinding motion control program is generated; and S5, a polishing procedure. According to the grinding device and the grinding method for damage repair of the curved-surface component, constant-force grinding can be conducted on the irregular surface of the curved-surface component, and it is guaranteed that the surface of the curved-surface component is ground more evenly.

Description

Technical Field

[0001] The present invention relates to the technical field of machining equipment, and particularly relates to a grinding device and a grinding method for repairing damage to curved surface components. Background Art

[0002] Due to their excellent high specific strength, high specific stiffness, fatigue resistance and other characteristics, composite materials have been increasingly widely used in the aviation and aerospace fields, especially in key parts such as aircraft wings, fuselages and radomes. With the increase of these applications, the corresponding maintenance and repair problems have become increasingly prominent.

[0003] During the service life of an aircraft, components such as wings, fuselages and radomes made of composite materials may be damaged, and these damages will affect the structural integrity and flight performance of the aircraft. Therefore, timely repair is necessary. When repairing these components, the commonly used method is patch bonding repair. This process includes grinding and removing the damaged area to form a stepped or sloped cavity to be repaired, preparing a patch that matches the groove to be repaired at the same time, and then using equipment such as a hot patching instrument and auxiliary materials to paste the patch into the groove. After it cures, the repaired surface is further ground and polished.

[0004] In the work of damage repair, when the existing grinding equipment for damage repair faces a curved surface component with an irregular surface, there is a key problem: it is difficult to apply a constant grinding force to the surface of the curved surface component during the grinding process. This defect directly leads to subsequent problems. Due to the unstable grinding force, the grinding uniformity of the surface of the curved surface component is greatly reduced, and the grinding uniformity is closely related to the repair quality of the curved surface component, ultimately resulting in a serious impact on the repair quality of the curved surface component. For example, the robot flexible intelligent grinding system based on a six-dimensional force sensor disclosed in Chinese Patent Application No. CN201910524443.3 consists of a teaching pendant, a robot controller, a robot body, a six-dimensional force sensor, a polishing device, etc. Before carrying out the grinding operation, the operator can delimit the two-dimensional area of grinding by means of drag teaching, and then scan out the three-dimensional grinding path through slow simulation. During grinding, the force sensor monitors the relative position and contact force between the grinding device and the curved surface to be ground in real time, and the robot controller adjusts the posture of the grinding device according to the monitoring data to complete the operation. However, this system has obvious deficiencies: on the one hand, relying solely on drag teaching for path planning is very likely to cause the path accuracy to be difficult to reach the ideal level; on the other hand, when this system is applied to the field of composite material patch repair, in order to achieve material removal during the grinding of the cavity of the bonding surface, the force on the grinding contact surface will continuously increase with the increase of the grinding depth, and the requirement of constant force grinding cannot be achieved. Another example is the robot grinding control system and its operation method disclosed in Chinese Patent Application No. CN201711495033.8. This system is equipped with a line laser sensor for scanning and obtaining the geometric parameters of the workpiece to be ground, and is provided with a three-dimensional model construction unit, which is responsible for establishing the three-dimensional model of the workpiece. The grinding area generation unit extracts the three-dimensional model of the grinding area according to the CAD model, the trajectory generation unit formulates the grinding path of the grinding area and the grinding force of each path point, the simulation unit simulates the grinding process of the robot according to the grinding path and force in three-dimensional space, and there is also a trajectory optimization unit to optimize the grinding path and force, so that the robot controller controls the robot operation according to the optimized data. However, this system is mainly applicable to the surface deburring process of metal parts and is not suitable for patching and grinding irregular surfaces such as stepped or sloped surfaces on the workpiece surface. Summary of the Invention

[0005] The object of the present invention is to overcome the problems in the prior art and provide a grinding device and a grinding method for damage repair of curved surface components, which can perform constant force grinding on the irregular surface of the curved surface component and ensure that the surface of the curved surface component is ground more evenly.

[0006] The present invention provides a grinding device and a grinding method for damage repair of curved surface components, including the following steps: S1. Hand-eye calibration and tool calibration: Use the nine-point method to perform hand-eye calibration of the eye-in-hand for the 3D vision system to obtain the transformation matrix of the 3D vision system coordinate system relative to the end center coordinate system of the robotic arm. Use the six-point method to perform tool calibration for the strong grinding machine to obtain the transformation matrix of the tool coordinate system relative to the end center coordinate system of the robotic arm; S2. Grinding initial state setting: Use the drag-and-teach function to set the initial grinding state of the robotic arm; S3. Grinding path planning: Use the 3D vision system to scan the surface of the workpiece to obtain the point cloud model of the workpiece surface. Based on the obtained point cloud model and the shape of the concave pit to be ground on the workpiece, perform data processing to obtain the grinding path point information for the concave pit; S4. Generation of robotic arm grinding program: According to the grinding path point information obtained above, combined with the transformation matrix obtained from hand-eye calibration and the transformation matrix obtained from tool calibration, calculate the robotic arm grinding path coordinates to generate the robotic arm grinding motion control program; S5. Grinding process: Control the robotic arm to perform the grinding process according to the generated robotic arm grinding motion control program.

[0007] Preferably, after the grinding process is completed, detect the grinding effect of the workpiece. When the grinding effect does not meet the requirements, repeat steps S2 - S5, or when generating the grinding motion control program in step S4, set a grinding path plan with multiple loops to continuously grind the workpiece.

[0008] Preferably, when performing hand-eye calibration work, the specific steps are as follows: First, install the 3D vision system at the end of the robotic arm, control the robotic arm to drive the 3D vision system to the shooting position to ensure that all nine corner points in the calibration board of the nine-point method are within the field of view of the 3D vision system, control the 3D vision system to capture the image of the calibration board, record the pixel coordinates of the nine corner points in the calibration board, and then adjust the end of the robotic arm in turn so that the 3D vision system reaches the nine corner points in the calibration board, record the coordinates of the robotic arm of the nine corner points, and finally obtain the transformation matrix of the 3D vision system coordinate system relative to the end center coordinate system of the robotic arm.

[0009] Preferably, when performing tool calibration work, the specific steps are as follows: Select a fixed tip in the working space, use the drag-and-teach function of the robotic arm controller to manipulate the end of the robotic arm to coincide with the fixed tip in four different poses. Each time they coincide, collect the robotic arm pose data, record the four sets of pose data collected, which provides the basis for the calculation of tool calibration. Then control the end of the robotic arm to move a set distance along the positive x-axis and z-axis of its end center coordinate system in turn, record the robotic arm poses of the two points respectively, and finally run the script of the drag-and-teach function to obtain the transformation matrix of the tool coordinate system relative to the end center coordinate system of the robotic arm.

[0010] Preferably, the specific steps for obtaining accurate grinding path points are as follows: First, it is necessary to obtain the shape model of the pit on the workpiece to be ground through 3D modeling software, and convert the shape model into a 3D point cloud of the pit through the 3D model data conversion function of the 3D modeling software; then, in the point cloud processing software, perform point cloud registration on the 3D point cloud of the pit and the 3D point cloud of the workpiece surface captured by the 3D vision system. The principle of registration is to register the 3D point cloud of the pit at the position on the workpiece surface that needs to be ground. After completion of the registration, a 3D point cloud of the workpiece with the pit is obtained, and the 3D coordinates of the 3D point cloud at the pit part have also been converted to the 3D vision system coordinate system; then, in the point cloud processing software, separate the 3D point cloud at the pit by means of point cloud segmentation, and perform point cloud normal vector calculation and point cloud slicing processing on the separated 3D point cloud at the pit to obtain a point cloud cross-section line, and then perform point cloud downsampling and point cloud sorting on the obtained point cloud cross-section line; finally, plan the different point cloud cross-section lines after downsampling and sorting to obtain the grinding path point information for the pit.

[0011] The present invention also provides a grinding device, including a strong grinder and a robotic arm, the strong grinder is connected to the end of the robotic arm, and is characterized by further including: A 3D vision system, provided on the robotic arm, the 3D vision system is used to scan the surface of the workpiece to obtain a point cloud model of the surface of the ground workpiece; A central controller, electrically connected to the 3D vision system, the central controller is used to perform data processing according to the obtained point cloud model and based on the shape of the pit to be ground on the workpiece, obtain the grinding path point information for the pit, and calculate the grinding path coordinates of the robotic arm according to the obtained grinding path point information for the pit, so as to generate a robotic arm grinding motion control program; A robotic arm controller, electrically connected to both the robotic arm and the central controller, the robotic arm controller has a drag teaching function, the robotic arm controller can determine the position information of the end of the robotic arm through the drag teaching function, and the robotic arm controller is used to control the action of the robotic arm according to the generated robotic arm grinding motion control program to grind the surface of the workpiece.

[0012] Preferably, an automatic tool changer is further provided on the strong grinder, a variety of specifications of diamond grinding tools are provided at the bottom of the automatic tool changer, the diamond grinding tools are provided with grinding wheels, and the automatic tool changer is used to replace the diamond grinding tools connected to the strong grinder.

[0013] Preferably, the strong grinder is electrically connected to a strong grinder driver, and the strong grinder driver is used to set the power and rotation speed of the strong grinder.

[0014] Preferably, it further includes an optical platform which is arranged below the 3D vision system. A workpiece fixture is provided on the optical platform for clamping the workpiece, and the level of the optical platform is adjustable.

[0015] Preferably, a dust collector is provided on the optical platform. The dust collector includes a dust suction hood and a dust suction main unit which are communicated with each other, and the dust suction hood is arranged above the workpiece fixture.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. For the grinding method for repairing damage of a curved surface member in the present invention, hand-eye calibration and tool calibration are performed through a robotic arm controller to obtain the transformation matrix of the 3D vision system coordinate system relative to the center coordinate system of the robotic arm end and the transformation matrix of the tool coordinate system relative to the center coordinate system of the robotic arm end. And a point cloud model of the surface of the workpiece to be ground is obtained. Then, the grinding path of the grinder is planned according to the point cloud model. Then, the workpiece is ground by the central controller according to the grinding path. It can realize the automation or semi-automation of the removal and grinding process involved in the repair process of the composite material curved surface member, and can perform constant-force grinding on the irregular surface of the curved surface member to ensure that the surface of the curved surface member is ground more evenly. And by using a robotic arm, the safety, cooperation and usability of the whole set of devices are greatly improved, the grinding and removal operation process involved in the composite material repair is efficiently realized, and the manual participation degree is reduced.

[0017] 2. The grinding tool of the present invention combines the function of robotic arm drag teaching to quickly realize the positioning of the initial grinding posture, uses the 3D vision system to quickly obtain the three-dimensional point cloud model of the surface of the grinding workpiece, performs data processing on the point cloud model based on the shape of the concave pit to be ground, and obtains accurate grinding path point information, meeting the requirements of fast and accurate repair and grinding of composite materials.

[0018] 3. The dust suction hood device in the present invention can efficiently absorb waste chips, dust, impurities, etc. generated during the grinding process, effectively preventing the escape and pollution of grinding impurities and dust.

[0019] 4. The whole set of grinding devices of the present invention integrates a dust suction device and a mobile platform, which is not only convenient for on-site deployment and adjustment during maintenance, but also reduces the occupational health hazards. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a schematic structural diagram of the present invention.

[0021] Figure 2 It is a schematic structural diagram at the grinder of the present invention.

[0022] Figure 3 It is a step diagram of the grinding method of the present invention.

[0023] Figure 4 Flow chart of the grinding path planning method of the present invention

[0024] Figure 5 Schematic diagram of the parallel point cloud slicing method of the present invention

[0025] Figure 6 Schematic diagram of the annular centering point cloud slicing method of the present invention

[0026] Explanation of reference numerals 1. Air source, 2. Optical platform, 3. Robotic arm controller, 4. Dust suction host, 5. Strong grinder driver, 6. Mouse, 7. Display screen, 8. Keyboard, 9. Host computer, 10. Workpiece, 11. Workpiece fixture, 12. Dust suction hood, 13. Strong grinder, 14. Tooling fixture, 15. 3D vision system, 16. Automatic tool changer, 17. Diamond grinding tool, 18. Robotic arm, 19. Air supply pipeline Detailed implementation manners

[0027] The following combines the attached Figures 1 - 6 , and describes in detail the specific implementation manners of the present invention. However, it should be understood that the protection scope of the present invention is not limited by the specific implementation manners. Based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the protection scope of the present invention

[0028] As Figures 1 - 6 shown, a grinding device for repairing damage to a curved surface component provided by the present invention includes an optical platform 2, a workpiece fixture 11, a strong grinder driver 5, a strong grinder 13, and a robotic arm 18. The strong grinder 13 is connected to the end of the robotic arm 18, and further includes: a 3D vision system 15, a central controller, and a robotic arm controller 3. The 3D vision system 15 is arranged on the robotic arm 18. The 3D vision system 15 is used to scan the surface of the workpiece 10 to obtain a point cloud model of the surface of the grinding workpiece 10. The central controller is electrically connected to the 3D vision system 15. The central controller is used to perform data processing according to the obtained point cloud model and based on the shape of the concave pit to be ground on the workpiece 10, obtain the grinding path point information for the concave pit, and calculate the grinding path coordinates of the robotic arm 18 according to the obtained grinding path point information for the concave pit, so as to generate a grinding motion control program for the robotic arm 18. The robotic arm controller 3 is electrically connected to both the robotic arm 18 and the central controller. The robotic arm controller 3 has a drag and teach function. The robotic arm controller 3 can determine the position information of the end of the robotic arm 18 through the drag and teach function. The robotic arm controller 3 is used to control the action of the robotic arm 18 according to the generated grinding motion control program for the robotic arm 18 to grind the surface of the workpiece 10

[0029] The working principle of the above embodiments is briefly described as follows: The central controller includes a host computer 9 and a human-machine operation module composed of a display screen 7, a keyboard 8, a mouse 6, etc. The central controller is used to realize the integrated control of the whole grinding device by the operator. The 3D vision system 15 is a binocular structured light camera. The host computer 9 establishes a communication connection with the robotic arm controller 3, the binocular structured light camera, the strong grinder driver 5, and the dust collection host 4 and realizes the control function.

[0030] Handles are provided on the side of the optical platform 2, and casters and adjustable support feet are provided at the bottom to meet the needs of work position transformation; when the work position is transformed, the support feet are adjusted to retract the support feet, the casters are loosened, and the whole device is transferred through the handle; during the implementation of the operation, the support feet are adjusted to jack up the optical platform 2, and the level of the optical platform 2 is calibrated to ensure that the whole grinding device is stably placed.

[0031] The robotic arm 18 is a six-axis light robotic arm 18, which clamps the strong grinder 13 through a tooling fixture 14.

[0032] The 3D vision system 15 uses a binocular structured light camera.

[0033] When the device is in use: Adjust the support feet to make the optical platform 2 level, and use the level to calibrate the level at multiple locations to ensure the stability of the entire device. Connect the external power supply and air supply lines, start the robot arm 18 and the binocular structured light camera through the human-machine operation module, and adjust the exposure, depth range and other parameters of the binocular structured light camera; according to the condition of the workpiece 10, set the working parameters of the strong grinder 13 through the strong grinder driver. Use the nine-point method to carry out eye-on-hand hand-eye calibration for the binocular structured light camera. Install the binocular structured light camera at the end of the robot arm 18, control the robot arm 18 to make the binocular structured light camera reach the appropriate position to capture the image of the calibration plate, and record the pixel coordinates of the nine corner points determined by the nine-point method; adjust the end of the robot arm 18 to reach the corner points of the calibration plate, and record the coordinates of the corner points in the central coordinate system of the end of the robot arm 18; execute the script to obtain the transformation matrix of the binocular structured light camera coordinate system relative to the central coordinate system of the end of the robot arm 18. Select a fixed tip in the workspace, use the drag-and-teach function of the robot controller 3, manipulate the end of the robot 18 to make it overlap with the fixed tip in four different postures, collect the posture data of the robot 18 each time it overlaps, record the four collected posture data, and provide a basis for the calculation of tool calibration, then control the end of the robot 18 to move the set distance in the positive direction of the x-axis and z-axis of the center coordinate system of its end in sequence, record the posture of the robot 18 at two points respectively, and finally run the script of the drag-and-teach function to run the script to obtain the conversion matrix of the tool coordinate system relative to the center coordinate system of the end of the robot 18. Install the workpiece 10 to be polished: the workpiece fixture 11 can be prepared by machining or 3D printing; after obtaining the workpiece fixture 11, place the workpiece 10 in a position convenient for polishing and fix it with the workpiece fixture 11, and connect it to the workpiece fixture 11 by threads according to the shape of the workpiece 10, and place the workpiece 10 in the middle of the optical platform 2. Initial grinding state setting: Use the drag-and-drop teaching function of the robot 18 to set the initial grinding state, which requires that the binocular structured light camera can capture the entire image of the workpiece 10 and that the robot 18 is not prone to motion singularities when executing the grinding program. Grinding path planning: Use the 3D vision system 15 to scan the surface of the workpiece 10 and obtain a point cloud model of the surface of the workpiece 10. In the central controller, data processing is performed based on the obtained point cloud model and the shape of the pit to be ground on the workpiece 10 to obtain accurate grinding path point information. The specific steps are as follows: Figure 4Shown as follows: First, it is necessary to obtain the shape model of the pit on the workpiece 10 to be polished through 3D modeling software, and convert it into a point cloud format file through the 3D model data conversion function of the 3D modeling software. Then, in the point cloud processing software, perform point cloud registration on the 3D point cloud of the pit and the 3D point cloud of the surface of the workpiece 10 captured by the 3D vision system 15. The principle of registration is to register the 3D point cloud of the pit at the position on the surface of the workpiece 10 that needs to be polished. After completing the registration, obtain the 3D point cloud of the workpiece 10 with the pit, and the 3D coordinates of the 3D point cloud at the pit part have also been converted to the coordinate system of the 3D vision system 15; then, in the point cloud processing software, separate the 3D point cloud at the pit through the method of point cloud segmentation, and perform point cloud normal vector calculation and point cloud slicing processing on the separated 3D point cloud at the pit to obtain the point cloud cross-section line. The slicing processing refers to slicing the point cloud using a parallel or annular centering type (such as Figure 5 and Figure 6 shown) cutting plane to obtain the point cloud cross-section line at the intersection of the cutting plane and the point cloud. The spacing or angle between each cutting plane during point cloud slicing is determined by the actual process parameters. Then, perform point cloud downsampling and point cloud sorting on the obtained point cloud cross-section line. Among them, the degree of point cloud downsampling is determined by the density of the final required path points. Finally, plan the different point cloud cross-section lines after downsampling and sorting using a reasonable polishing order to obtain the polishing path point information for the pits on the entire workpiece 10. Generation of the polishing program for the robotic arm 18: According to the polishing path point information in the binocular structured light camera coordinate system, convert the normal vector to Euler angles to process the path point information, and combine the hand-eye calibration matrix and the tool calibration matrix to calculate the polishing path pose of the robotic arm 18 and generate a polishing motion control program. Start the strong grinding machine 13, control the robotic arm 18 to execute the polishing process according to the generated program, and the program needs to be executed two or more times to eliminate the trajectory interval and burrs. After the polishing is completed, turn off the strong grinding machine 13 and the robotic arm 18, and clean the debris in the polishing area. Remove the workpiece 10 and detect the polishing effect visually or with professional equipment.

[0034] A polishing device for repairing damage to a curved surface component according to the present invention can realize the automation or semi-automation of the digging and polishing process involved in the repair of a composite curved surface component, and can perform constant-force polishing on the irregular surface of the curved surface component to ensure that the surface of the curved surface component is polished more evenly.

[0035] As a preferred solution, as shown in Figure 1 and Figure 2 shown, wherein, an automatic tool changer 16 is further provided on the strong grinding machine 13. A variety of specifications of diamond grinding tools 17 are provided at the bottom of the automatic tool changer 16. The diamond grinding tools 17 are equipped with grinding wheels. The automatic tool changer 16 is used to replace the diamond grinding tools 17 connected to the strong grinding machine 13.

[0036] By setting up an automatic tool changer 16, according to the structure of the surface of the workpiece 10 and the grinding requirements, the automatic tool changer 16 is controlled to act to change the tool of the strong grinder 13 and replace diamond grinding tools 17 of different specifications to meet the grinding requirements of different process parameters.

[0037] As a preferred solution, as Figure 1 and Figure 2 shown, wherein, the strong grinder 13 is electrically connected to a strong grinder driver 5, and the strong grinder driver 5 is used to set the power and rotation speed of the strong grinder 13. By setting the power and rotation speed of the strong grinder 13 through the strong grinder driver 5, the grinding parameters for the pit can be adjusted in a timely manner according to the pit shape, thereby further improving the grinding quality of the entire grinding device.

[0038] As a preferred solution, as Figure 1 and Figure 2 shown, wherein, an optical platform 2 is further included, the optical platform 2 is arranged below the 3D vision system 15, a workpiece fixture 11 is arranged on the optical platform 2, and the workpiece fixture 11 is used to clamp the workpiece 10, and the level of the optical platform 2 is adjustable. By setting the workpiece fixture 11, when grinding the workpiece 10, first use the optical platform 2 to adjust the posture of the workpiece fixture 11, and then use the workpiece fixture 11 to clamp and fix the workpiece 10, so that the workpiece 10 is in a suitable grinding posture, and the workpiece fixture 11 can prevent the workpiece 10 from shaking during grinding, thereby ensuring the grinding quality of the entire grinding device.

[0039] As a preferred solution, as Figure 1 and Figure 2 shown, wherein, a dust collector is arranged on the optical platform 2, the dust collector includes a dust suction hood 12 and a dust suction main unit 4, the dust suction main unit 4 is communicated with the dust suction hood 12, and the dust suction hood 12 is arranged above the workpiece fixture 11. Fix the pipeline of the dust suction hood 12 on the side of the optical platform 2, the dust suction hood 12 is close to the surface of the workpiece 10, start the dust suction main unit 4, the dust suction main unit 4 is installed on the bottom support plate of the optical platform 2, and the dust suction hood 12 covers above the workpiece fixture 11, that is, above the grinding area of the workpiece 10, so that the composite material debris generated during the grinding process can be cleaned up in time to prevent dust from escaping and avoid hindering the normal movement of the strong grinder 13 due to debris accumulation.

[0040] As a preferred solution, as Figure 1As shown in the figure, the powerful grinding machine 13 is connected to a pneumatic control circuit, which includes a gas source 1 and a supply pipeline 19. One end of the supply pipeline 19 is connected to the gas source 1, and the other end is connected to the powerful grinding machine 13. An oil mist separator, a water droplet separator, a solenoid valve, a regulating valve, and a throttle valve are provided on the supply pipeline 19. The gas source 1 of this device is an air compressor. The pneumatic control circuit cooperates through its internal pneumatic components to clamp the diamond grinding tool 17 and automatically change the tool of the automatic tool changer 16.

[0041] The present invention also provides a method for grinding using a grinding device for repairing damage to curved surface components, including the following steps: S1. Hand-eye calibration and tool calibration: Use the nine-point method to carry out hand-eye calibration of the eye-in-hand for the 3D vision system 15 to obtain the transformation matrix of the 3D vision system 15 coordinate system relative to the end center coordinate system of the robotic arm 18. Use the six-point method to carry out tool calibration for the powerful grinding machine 13 to obtain the transformation matrix of the tool coordinate system relative to the end center coordinate system of the robotic arm 18; S2. Setting the initial grinding state: Use the drag and teach function to set the initial grinding state of the robotic arm 18; S3. Grinding path planning: Use the 3D vision system 15 to scan the surface of the workpiece 10 to obtain the point cloud model of the workpiece 10 surface. According to the obtained point cloud model and based on the shape of the concave pit to be ground on the workpiece 10, perform data processing to obtain the grinding path point information for the concave pit; S4. Generating the grinding program for the robotic arm 18: According to the obtained grinding path point information mentioned above, combined with the transformation matrix obtained from hand-eye calibration and the transformation matrix obtained from tool calibration, calculate the grinding path coordinates of the robotic arm 18 to generate the grinding motion control program for the robotic arm 18; S5. Grinding process: Control the robotic arm 18 to execute the grinding process according to the generated grinding motion control program for the robotic arm 18.

[0042] As a preferred solution, as Figure 3 shown, after the grinding process is completed, detect the grinding effect of the workpiece 10. When the grinding effect does not meet the requirements, repeat the steps S2 - S5, or when generating the grinding motion control program in step S4, set the grinding path planning to loop multiple times to continuously grind the workpiece 10.

[0043] As a preferred solution, as Figure 3As shown in the figure, when performing hand-eye calibration work, the specific steps are as follows: First, install the 3D vision system 15 at the end of the robotic arm 18, control the robotic arm 18 to drive the 3D vision system 15 to the shooting position to ensure that all nine corner points in the calibration board of the nine-point method are within the field of view of the 3D vision system 15, control the 3D vision system 15 to capture an image of the calibration board, record the pixel coordinates of the nine corner points in the calibration board, and then sequentially adjust the end of the robotic arm 18 to enable the 3D vision system 15 to reach the nine corner points in the calibration board, record the coordinates of the robotic arm 18 of the nine corner points, and finally obtain the transformation matrix of the 3D vision system 15 coordinate system relative to the center coordinate system of the end of the robotic arm 18.

[0044] As a preferred solution, as Figure 3 shown in the figure, when performing tool calibration work, the specific steps are as follows: Select a fixed tip in the working space, use the drag and teach function of the robotic arm controller 3 to control the end of the robotic arm 18 to coincide with the fixed tip in four different poses, collect the pose data of the robotic arm 18 each time they coincide, record the pose data collected these four times to provide a basis for the calculation of tool calibration, then control the end of the robotic arm 18 to move a set distance in the positive directions of the x-axis and z-axis of its end center coordinate system in sequence, record the pose of the robotic arm 18 at the two points respectively, and finally run the script of the drag and teach function to obtain the transformation matrix of the tool coordinate system relative to the center coordinate system of the end of the robotic arm 18.

[0045] As a preferred solution, as Figure 4 shown in the figure, the specific steps to obtain accurate grinding path points are as follows: First, it is necessary to obtain the shape model of the concave pit on the workpiece 10 to be ground through 3D modeling software, and convert the shape model into a three-dimensional point cloud of the concave pit through the three-dimensional model data conversion function of the 3D modeling software; then perform point cloud registration on the three-dimensional point cloud of the concave pit and the three-dimensional point cloud of the surface of the workpiece 10 captured by the 3D vision system 15 in the point cloud processing software. The principle of registration is to register the three-dimensional point cloud of the concave pit at the position on the surface of the workpiece 10 that needs to be ground. After completion of registration, obtain the three-dimensional point cloud of the workpiece 10 with the concave pit, and the three-dimensional coordinates of the three-dimensional point cloud at the concave pit part have also been converted to the 3D vision system 15 coordinate system; then separate the three-dimensional point cloud at the concave pit in the point cloud processing software by means of point cloud segmentation, perform point cloud normal vector calculation and point cloud slicing processing on the separated three-dimensional point cloud at the concave pit to obtain a point cloud cross-section line, and then perform point cloud downsampling and point cloud sorting on the obtained point cloud cross-section line; finally, plan the different point cloud cross-section lines after downsampling and sorting to obtain the grinding path point information for the concave pit.

[0046] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention.

Claims

1. A grinding method for repairing damage to a curved surface component, characterized in that: The following steps are involved: S1. Hand-eye calibration and tool calibration: Use the nine-point method to perform eye-on-hand hand-eye calibration on the 3D vision system (15) to obtain the transformation matrix of the 3D vision system (15) coordinate system relative to the end center coordinate system of the robotic arm (18); use the six-point method to perform tool calibration on the high-performance grinder (13) to obtain the transformation matrix of the tool coordinate system relative to the end center coordinate system of the robotic arm (18); S2, setting the initial grinding state: using the drag-and-drop teaching function to set the initial grinding state of the robot arm (18); S3, grinding path planning: using a 3D vision system (15) to scan the surface of the workpiece (10) to obtain a point cloud model of the surface of the workpiece (10), and performing data processing based on the obtained point cloud model and the shape of the pit to be ground on the workpiece (10) to obtain grinding path point information for the pit; S4, generating a grinding program for the robot arm (18): according to the grinding path point information obtained above, combined with the conversion matrix obtained by the hand-eye calibration and the conversion matrix obtained by the tool calibration, the grinding path coordinates of the robot arm (18) are calculated to generate a grinding motion control program for the robot arm (18); S5, grinding process: controlling the robot arm (18) to perform the grinding process according to the generated robot arm (18) grinding motion control program.

2. The method for repairing damage to a curved surface component according to claim 1, characterized in that: After the grinding process is completed, the grinding effect of the workpiece (10) is detected. When the grinding effect does not meet the requirements, the steps S2-S5 are repeatedly executed, or when the grinding motion control program of step S4 is generated, a grinding path planning with multiple cycles is set to continuously grind the workpiece (10).

3. The method for repairing damage to a curved surface component according to claim 1, characterized in that: When performing the hand-eye calibration, the specific steps are as follows: first, the 3D vision system (15) is installed on the end of the robotic arm (18), the robotic arm (18) is controlled to drive the 3D vision system (15) to reach the shooting position to ensure that the nine corner points in the calibration plate of the nine-point method are all located within the field of view of the 3D vision system (15), the 3D vision system (15) is controlled to shoot the image of the calibration plate, the pixel coordinates of the nine corner points in the calibration plate are recorded, and then the end of the robotic arm (18) is adjusted in sequence to make the 3D vision system (15) reach the nine corner points in the calibration plate, the coordinates of the robotic arm (18) of the nine corner points are recorded, and finally, the transformation matrix of the coordinate system of the 3D vision system (15) relative to the central coordinate system of the end of the robotic arm (18) is obtained.

4. The method for repairing damage to a curved surface component according to claim 1, characterized in that: When performing tool calibration, the specific steps are as follows: select a fixed tip in the workspace, use the drag teaching function of the robot arm controller (3), control the end of the robot arm (18) to make it overlap with the fixed tip in four different postures, collect the posture data of the robot arm (18) each time it overlaps, record the four collected posture data, and provide a basis for the calculation of tool calibration, then control the end of the robot arm (18) to move a set distance in sequence along the positive direction of the x-axis and z-axis of the center coordinate system of its end, record the posture of the robot arm (18) at two points respectively, and finally run the script of the drag teaching function to obtain the transformation matrix of the tool coordinate system relative to the center coordinate system of the end of the robot arm (18).

5. The method for repairing damage to a curved surface component according to claim 4, characterized in that: The specific steps for obtaining accurate grinding path points are as follows: first, a shape model of the pit on the workpiece (10) to be ground is obtained through a three-dimensional modeling software, and the shape model is converted into a three-dimensional point cloud of the pit through the three-dimensional model data conversion function of the three-dimensional modeling software; then, in the point cloud processing software, the three-dimensional point cloud of the pit and the three-dimensional point cloud of the surface of the workpiece (10) photographed by the 3D vision system (15) are point cloud registered. The principle of registration is to register the three-dimensional point cloud of the pit at the position on the surface of the workpiece (10) that needs to be polished. After completing the registration, a point cloud with a pit is obtained. The three-dimensional point cloud of the workpiece (10) at the pit is obtained, and the three-dimensional coordinates of the three-dimensional point cloud at the pit have also been converted to the coordinate system of the 3D vision system (15); then, the three-dimensional point cloud at the pit is separated by point cloud segmentation in the point cloud processing software, and the point cloud normal vector calculation and point cloud slicing processing are performed on the separated three-dimensional point cloud at the pit to obtain the point cloud section line, and then the obtained point cloud section line is point cloud down-sampled and point cloud sorted; finally, the different point cloud section lines after down-sampling and sorting are planned to obtain the grinding path point information for the pit.

6. A grinding device, comprising a grinding machine (13) and a mechanical arm (18), wherein the grinding machine (13) is connected to the end of the mechanical arm (18), characterized in that: Also includes: A 3D vision system (15) is disposed on the robotic arm (18), and the 3D vision system (15) is used to scan the surface of the workpiece (10) to obtain a point cloud model of the surface of the polished workpiece (10); A central controller is electrically connected to the 3D vision system (15), and is used to perform data processing based on the acquired point cloud model and the shape of the pit to be polished on the workpiece (10), obtain polishing path point information for the pit, and calculate the polishing path coordinates of the robot arm (18) based on the acquired polishing path point information for the pit, thereby generating a polishing motion control program for the robot arm (18); A robot arm controller (3) is electrically connected to the robot arm (18) and the central controller. The robot arm controller (3) has a dragging teaching function. The robot arm controller (3) can determine the point information of the end of the robot arm (18) through the dragging teaching function. The robot arm controller (3) is used to control the movement of the robot arm (18) according to the generated robot arm (18) grinding motion control program to grind the surface of the workpiece (10).

7. The grinding device according to claim 6, characterized in that: The high-performance grinding machine (13) is also provided with an automatic tool changing device (16), and a diamond grinding tool (17) of various specifications is provided at the bottom of the automatic tool changing device (16). The diamond grinding tool (17) has a grinding wheel, and the automatic tool changing device (16) is used to replace the diamond grinding tool (17) connected to the high-performance grinding machine (13).

8. The grinding device according to claim 6, characterized in that: The high-speed mill (13) is electrically connected to a high-speed mill driver (5), and the high-speed mill driver (5) is used to set the power and rotation speed of the high-speed mill (13).

9. The grinding device according to claim 6, characterized in that: It also includes an optical platform (2), wherein the optical platform (2) is arranged below the 3D vision system (15), a workpiece clamp (11) is arranged on the optical platform (2), the workpiece clamp (11) is used to clamp the workpiece (10), and the horizontality of the optical platform (2) is adjustable.

10. The grinding device according to claim 9, characterized in that: A vacuum cleaner is provided on the optical platform (2), the vacuum cleaner comprising a vacuum hood (12) and a vacuum main unit (4), the vacuum main unit (4) being connected to the vacuum hood (12), and the vacuum hood (12) being arranged above the workpiece fixture (11).

Images