Method and system for automatic generation of polishing trajectories for a table robot with 3D vision scanning

Abstract

The application relates to the field of furniture plate processing and robot intelligent polishing technology, and provides a 3D visual scanning table plate robot polishing track automatic generation method and system. The method performs positioning and clamping on a table plate to be polished, and obtains hole type information of the table plate; a three-dimensional line scanning camera is uniformly translated along the length direction of the table plate by a robot, three-dimensional coordinate data of an inner contour of a hole to be polished on the table plate is collected; the three-dimensional coordinate data is analyzed, offset calculation rules corresponding to hole types are matched, hole expansion offset, chamfer height offset and circular segment normal direction offset calculation are performed, tool wear compensation parameters are superimposed, and actual polishing path points are generated; a robot polishing track is generated according to the polishing path points, and the robot is controlled to perform polishing work along the track. Through the provided method, manual type demonstration is not needed, product change time is greatly shortened, and a multi-variety small-batch production mode is adapted.

Description

Technical Field

[0001] This application relates to the field of furniture board processing and robotic intelligent sanding technology, and in particular to a method and system for automatically generating the sanding trajectory of a tabletop robot using 3D visual scanning. Background Technology

[0002] With the development of the furniture board processing industry, the demand for hole grinding of quartz stone and slab countertops is constantly increasing. Currently, countertop hole grinding mainly relies on manual operation or traditional robot-taught grinding. Manual grinding is inefficient, produces inconsistent quality, is labor-intensive, and generates serious dust hazards.

[0003] Traditional robotic teaching grinding requires manual teaching of the trajectory for each hole size, which is time-consuming and cannot adapt to the production mode of multiple varieties and small batches. Moreover, it mostly uses two-dimensional vision, which can only obtain planar coordinates and lacks the ability to perceive the height direction. It cannot compensate for the uneven grinding depth caused by the bending deformation of the table. At the same time, it lacks an automatic detection and compensation mechanism for tool wear. The grinding quality continues to decline as the tool wears out, making it difficult to ensure product consistency and production efficiency. Summary of the Invention

[0004] This application provides a method and system for automatically generating the grinding trajectory of a tabletop robot using 3D vision scanning, aiming to solve the problems of low efficiency, unstable quality, high labor intensity, and serious dust hazards in the existing tabletop hole grinding mainly relying on manual operation or traditional robot teaching grinding.

[0005] In a first aspect, embodiments of this application provide a method for automatically generating a grinding trajectory for a 3D visual scanning tabletop robot, the method comprising: Positioning and clamping the table panel to be polished, and obtaining the hole type information of the table panel; the robot drives the 3D line scanning camera to move at a constant speed along the length of the table panel to collect the 3D coordinate data of the inner contour of the hole to be polished on the table panel. Analyze the three-dimensional coordinate data, match the offset calculation rules of the corresponding hole type, perform the calculation of hole enlargement offset, chamfer height offset and arc segment normal direction offset, superimpose tool wear compensation parameters, and generate the actual grinding path points; The robot's grinding trajectory is generated based on the grinding path points, and the robot is controlled to perform grinding operations along the trajectory. During the grinding process, the axial position of the tool is adjusted in real time according to the height coordinates of each path point to ensure that the grinding depth is consistent at all points.

[0006] In some embodiments, the step of positioning and clamping the table panel to be polished and obtaining the hole type information of the table panel includes: performing a barcode scanning operation when the table panel to be polished flows into the feeding station to obtain the hole type, board size and order data of the table panel; adjusting the width of the lateral clamping mechanism according to the board size, performing lateral clamping and vertical pressing on the table panel, so that the conveying mechanism descends and disengages from the table panel to complete the positioning and fixing.

[0007] In some embodiments, the step of having a robot drive a three-dimensional line scanning camera to move at a constant speed along the length of the table panel to collect three-dimensional coordinate data of the inner contour of the hole to be polished on the table panel includes: the robot driving the three-dimensional line scanning camera integrated at the end to move at a constant speed along the length of the table panel, triggering the camera to collect images when it reaches a designated position, identifying and extracting the three-dimensional coordinate data of the straight edges and arc angles of each segment of the hole to be polished, and completing the scanning of the inner contour of the hole.

[0008] In some embodiments, parsing the three-dimensional coordinate data and matching the offset calculation rules for the corresponding hole type includes: receiving the three-dimensional coordinate data obtained by scanning, identifying the type and number of holes to be ground, classifying and storing the coordinate points of each contour segment according to the hole type, and calling the trajectory offset calculation rules for the corresponding type.

[0009] In some embodiments, the calculation of hole enlargement offset, chamfer height offset, and arc segment normal direction offset includes: expanding the original hole contour outward according to a preset single-sided hole enlargement amount to generate a tool center plane trajectory; setting a chamfer height offset to control the tool's pressing depth; and performing equal-point offset calculations on the arc segment contour along the normal direction to maintain the arc segment trajectory smooth and continuous.

[0010] In some embodiments, the step of superimposing tool wear compensation parameters to generate actual grinding path points includes: obtaining the height compensation value and diameter compensation value obtained by tool detection, superimposing the tool wear compensation parameters onto the offset contour trajectory, and generating the actual grinding path points corresponding to the grinding process.

[0011] In some embodiments, generating a robot grinding trajectory based on grinding path points includes: matching tool parameters and feed rate parameters of the corresponding mesh size according to the grinding process requirements corresponding to the hole position, and generating a robot grinding trajectory containing multiple grinding processes in combination with the grinding path points.

[0012] In some embodiments, the control robot to perform grinding operations along a trajectory includes: controlling the robot to change the grinding tool for the corresponding process, performing grinding operations along the inner contour of the hole according to the generated grinding trajectory, automatically switching tools after each process is completed, and sequentially completing all grinding processes.

[0013] In some embodiments, the axial position of the tool is adjusted in real time according to the height coordinates of each path point during the grinding process to ensure that the grinding depth is consistent throughout. This includes: reading the height coordinate value corresponding to each grinding path point during the grinding process, adjusting the axial position of the tool in real time, compensating for the bending deformation of the table panel, verifying the rationality of the height coordinates, intercepting abnormal values, and ensuring that the grinding depth is consistent throughout.

[0014] Secondly, this application provides an automatic trajectory generation system for a 3D visual scanning tabletop robot grinding process, the system comprising: The positioning and clamping unit is used to perform positioning and clamping on the table panel to be polished and to obtain the hole type information of the table panel; the robot drives the three-dimensional line scanning camera to move at a constant speed along the length of the table panel to collect the three-dimensional coordinate data of the inner contour of the hole to be polished on the table panel. The data parsing unit is used to parse three-dimensional coordinate data, match the offset calculation rules of the corresponding hole type, perform hole enlargement offset, chamfer height offset and arc segment normal direction offset calculation, superimpose tool wear compensation parameters, and generate actual grinding path points. The trajectory generation unit is used to generate the robot's grinding trajectory based on the grinding path points, control the robot to perform grinding operations along the trajectory, and adjust the axial position of the tool in real time according to the height coordinates of each path point during the grinding process to ensure that the grinding depth is consistent at all points.

[0015] The method provided in this application has the following beneficial effects: 1. Automatic trajectory generation significantly improves changeover efficiency. By automatically acquiring the contour coordinates of hole positions through 3D vision scanning, the system automatically generates the grinding trajectory, eliminating the need for manual step-by-step teaching, greatly shortening product changeover time, and adapting to multi-variety, small-batch production modes.

[0016] 2. Real-time height tracking ensures consistent grinding quality. The tool's axial position is adjusted in real-time based on the height coordinates obtained from scanning, effectively compensating for table panel bending deformation and ensuring uniform chamfer depth and grinding effect throughout, thus improving product yield.

[0017] 3. Dynamic tool wear compensation extends tool life. Wear compensation values ​​are obtained through tool size detection and superimposed on the trajectory. Combined with a mechanism that dynamically adjusts the feed rate according to tool life, tool utilization is maximized while ensuring grinding quality.

[0018] 4. Multi-type hole position adaptive, strong system versatility. It can automatically adapt to various hole position types such as square, circular, elliptical and double hole combinations, and match the corresponding trajectory offset calculation rules. There is no need to develop separate programs for different hole types, reducing development and maintenance costs.

[0019] 5. High degree of automation, meeting mass production needs. The entire process achieves fully automated operation of barcode scanning and positioning, scanning and trajectory generation, and automatic grinding, which greatly reduces the intensity of manual labor and dust hazards, and the production capacity can meet the needs of large-scale industrial production.

[0020] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic flowchart illustrating the steps of an embodiment of the present application for an automatic generation method of grinding trajectory for a 3D visual scanning tabletop robot; Figure 2 This is a schematic diagram illustrating the principle of an automatic generation method for the grinding trajectory of a 3D vision scanning tabletop robot provided in an embodiment of this application; Figure 3 This is a schematic block diagram of a 3D vision scanning tabletop robot grinding trajectory automatic generation system provided in one embodiment of this application; Figure 4 This is a schematic block diagram of the structure of a computer device provided in an embodiment of this application.

[0023] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Detailed Implementation

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0026] It should be understood that, in order to clearly describe the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and the terms "first" and "second" are not necessarily different.

[0027] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0028] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0029] With the development of the furniture board processing industry, the demand for hole grinding of quartz stone and slab countertops is constantly increasing. Currently, countertop hole grinding mainly relies on manual operation or traditional robot-taught grinding. Manual grinding is inefficient, produces inconsistent quality, is labor-intensive, and generates serious dust hazards.

[0030] Traditional robotic teaching grinding requires manual teaching of the trajectory for each hole size, which is time-consuming and cannot adapt to the production mode of multiple varieties and small batches. Moreover, it mostly uses two-dimensional vision, which can only obtain planar coordinates and lacks the ability to perceive the height direction. It cannot compensate for the uneven grinding depth caused by the bending deformation of the table. At the same time, it lacks an automatic detection and compensation mechanism for tool wear. The grinding quality continues to decline as the tool wears out, making it difficult to ensure product consistency and production efficiency.

[0031] Please refer to Figure 1 This application provides a method for automatically generating the grinding trajectory of a tabletop robot using 3D visual scanning, applied to computer equipment. The computer equipment can be deployed on a single server or server cluster. It can also be deployed on handheld terminals, laptops, wearable devices, or robots, etc. It should be noted that every piece of information involved in the method provided in this application is extracted with the authorization of the relevant user and in accordance with relevant regulations, and will not infringe on user privacy.

[0032] The automatic generation method for robot grinding trajectory of tabletop based on 3D vision scanning provided in this embodiment relies on an automatic tabletop grinding system. The grinding system includes a feeding and conveying unit, a positioning and clamping unit, a 3D scanning unit, a robot grinding unit, a tool detection unit, an automatic tool magazine unit, and a control unit. Each unit operates collaboratively through an industrial bus and communication protocol.

[0033] Feeding and conveying unit: Utilizing a roller conveyor line driven by a servo motor, it transports the workpiece to be polished from the upstream station to the polishing station, and then delivers the polished workpiece to the downstream station. The conveyor line is equipped with multiple sets of photoelectric position detection sensors to detect the workpiece's position, control start / stop, and deceleration positioning.

[0034] Positioning and clamping unit: Includes a servo-driven long-side clamping mechanism, multiple sets of vertical lifting and pressing cylinders, and a lifting belt mechanism. The long-side clamping mechanism can adjust the clamping width according to the length of the workpiece to achieve lateral centering and positioning of the workpiece; the vertical lifting cylinders are used to lift and press the table panel from below; the belt mechanism can be raised and lowered as a whole, and after positioning, it descends and detaches from the bottom surface of the workpiece to avoid interference from the conveyor structure to the grinding process.

[0035] 3D Scanning Unit: Employs a 3D laser line scanner camera, integrated and mounted on the robot's end effector. The camera is equipped with an automatically opening and closing sealed protective cover. The cover opens during scanning operations and closes during polishing operations to isolate water mist and dust during the polishing process, protecting the precision of the optical components. The camera communicates with the robot control cabinet via the industrial Ethernet protocol to acquire 3D coordinate data of the table surface and the inner contours of holes.

[0036] Robotic Grinding Unit: Includes a multi-axis industrial robot, a constant force floating mechanism, a water-cooled electric spindle, and grinding tools. The constant force floating mechanism is connected between the robot's end flange and the electric spindle, providing axial floating buffer capability and allowing for multiple grinding pressure settings to ensure stable grinding contact force. The water-cooled electric spindle is used to clamp the grinding tools and provide rotational power, supports automatic tool changing, and is compatible with grinding tools of different grit sizes and lengths.

[0037] Tool Inspection Unit: Employs a high-resolution area array imaging camera, equipped with a backlight illumination module, and is fixedly positioned at the inspection station next to the tool magazine. After tool changing, the robot moves the tool to the inspection station, and the camera images and measures the rotating tool, detecting tool diameter, tool length, and surface defects such as chipping and pitting, with an inspection accuracy down to the micrometer level.

[0038] Automatic tool magazine unit: It is equipped with multiple tool stations and can store grinding tools of different grit and length. It is suitable for different thicknesses of table panels and different grinding processes. The tool position and life information are managed by the control unit.

[0039] Control Unit: This includes the robot control cabinet and the programmable logic controller (PLC), which communicate via an industrial bus protocol. The robot control cabinet controls the robot's movement, parses scan data, performs trajectory offset calculations, and generates grinding motion programs. The PLC controls the logical actions of the conveyor line, clamping mechanism, and tool magazine, and is responsible for tool life accumulation and management.

[0040] The provided method for automatically generating the grinding trajectory of a tabletop robot based on 3D vision scanning includes steps S101 to S103. Details are as follows: Step S101. Position and clamp the table panel to be polished to obtain the hole type information of the table panel; the robot drives the three-dimensional line scanning camera to move at a constant speed along the length of the table panel to collect the three-dimensional coordinate data of the inner contour of the hole to be polished on the table panel.

[0041] Specifically, when the grinding table panel is conveyed to the grinding station entrance by the feeding conveyor unit, the barcode scanning device reads the marking information on the surface of the table panel to obtain the corresponding hole type, board size parameters, processing requirements, and order association information. The hole types cover a variety of common types, including square holes, round holes, oval holes, and double-hole combinations, and the board size parameters include core processing parameters such as length, width, and thickness.

[0042] After acquiring the sheet information, the conveying unit continues to transport the worktable to the inside of the grinding table. The system calculates the stop position of the sheet based on its length information, ensuring that the center of the sheet is roughly aligned with the center of the grinding station. Once in position, the positioning and clamping unit executes clamping actions sequentially: First, the servo long-side clamping mechanism moves from both sides of the sheet towards the center along its length, with the jaws on both sides simultaneously contacting the side edges of the worktable, centering and positioning the sheet in the width direction. The clamping force is set to a reasonable range that ensures the sheet does not shift without causing edge chipping. Subsequently, multiple sets of vertical clamping cylinders distributed on the grinding table surface extend upwards simultaneously, with the flexible support pads at the top of the cylinders contacting the bottom surface of the worktable, lifting the sheet to a set height. After lifting to the designated position, the clamping mechanism on the table surface presses down from above, cooperating with the lifting cylinders to vertically clamp and fix the worktable. Finally, the conveyor belt mechanism descends as a whole, completely detaching from the bottom surface of the worktable, preventing belt vibration from affecting processing accuracy during grinding and providing space for grinding debris to fall. This completes the positioning and clamping of the tabletop.

[0043] After positioning is complete, the system sends a board positioning completion signal to the robot control cabinet. The robot moves the integrated 3D line scan camera at the end effector to the starting scanning position at one end of the platform, ensuring the camera's field of view covers the entire width of the board and the camera height is within the optimal working distance range. The robot sends a scan start signal to the 3D line scan camera, and the camera enters standby acquisition mode. Subsequently, the robot moves the camera to perform a uniform translational scan along the length of the platform at a set speed. The translational speed matches the camera's line frequency to ensure continuous and uninterrupted scan data.

[0044] During the scanning process, when the robot reaches the preset trigger position, it triggers the camera to perform image acquisition via a digital I / O signal. The camera acquires depth point cloud data of the table surface line by line through line laser projection and imaging reception. The scan covers the entire width of the table, ensuring that all holes to be polished are within the camera's field of view. When the robot reaches the scanning termination position at the other end of the table, it sends a scan end signal to the camera, and the camera stops acquiring data.

[0045] The camera's built-in processing unit preprocesses the acquired point cloud data: first, it performs point cloud denoising to remove abnormal noise points caused by reflections and dust; then, it uses an edge extraction algorithm to identify the inner contour boundary of the hole; next, it decomposes the contour boundary into straight edge segments and arc segments, extracting the three-dimensional coordinates of the sampling points on each segment; finally, it packages and outputs the results according to the agreed data format. The camera's planar position detection accuracy and height detection accuracy are both better than 0.3mm, and the total time for scanning and processing a single board is controlled within a few seconds, meeting the requirements of industrial production cycle time.

[0046] Step S102. Parse the three-dimensional coordinate data, match the offset calculation rules of the corresponding hole type, perform hole enlargement offset, chamfer height offset and arc segment normal direction offset calculation, superimpose tool wear compensation parameters, and generate actual grinding path points.

[0047] Specifically, the 3D line scan camera sends the processed contour data to the robot control cabinet in string format via a network communication protocol. The data string uses a delimiter-separated format, and the data segments include, in sequence: scan success flag, number of holes, hole type code, fraction of the long side, fraction of the wide side, number of arc division points, followed by multiple sets of 3D coordinate values. Each set of coordinates contains three values: X, Y, and Z, and the values ​​are rounded to two decimal places.

[0048] After receiving the data string, the robot control cabinet first reads the first scan success flag. If the flag indicates a scan failure, it sends a retransmission command to the camera and records the error code. If the retransmission still fails, it sends a scan failure signal to the main control system for manual intervention.

[0049] If the scan is successful, the system splits the string using delimiters, counts the total number of coordinate points, and converts the character data into a numerical 3D coordinate array, completing the parsing of the original contour data. After parsing, the system identifies the type of the current hole based on the hole type code and stores the contour coordinate points of different segments into corresponding position arrays. For square holes, the coordinates of the four straight edges and four rounded corners are stored in the long edge grinding point array, the wide edge grinding point array, and the rounded corner grinding point array, respectively. Each array adopts a two-dimensional structure and identifies the edge number and the point number, respectively. For workpieces with two holes, the coordinates of the two sets of holes are stored in separate array partitions, and a combination marker variable is set to distinguish the order of the holes. For circular holes, the coordinates of the entire arc contour are stored in the arc grinding point array segment by segment; For elliptical holes, the array structure is similar to that for square holes for storage, and the arc segments are described using elliptical parametric equations.

[0050] After data storage is complete, the system calls the trajectory offset calculation program corresponding to the hole type to perform offset calculations on the original contour coordinate points and generate the motion trajectory of the tool center. The offset calculation includes the following four parts: Hole Enlargement Offset Calculation: A total grinding offset in the planar direction is set to control the grinding amount during single-sided hole enlargement. The offset ranges from 1.2mm to 1.8mm, with the offsets in the X and Y directions remaining consistent. Using the original hole's inner contour as a reference, this offset distance is extended outwards along the contour's normal direction to obtain the planar motion trajectory of the grinding tool center, ensuring that the hole dimensions meet design requirements after grinding. The actual grinding amount is inversely proportional to the grinding feed rate; the system can compensate for differences in grinding amount caused by speed variations by adjusting the offset.

[0051] Chamfer height offset calculation: Set the chamfer height offset parameter in the Z-axis direction to control the tool's depth of cut, thereby controlling the size of the chamfer. Different Z-axis offset values ​​are used for rough grinding, fine grinding, and polishing. The depth of cut is smaller in rough grinding, while it is slightly larger in fine grinding and polishing. A uniform chamfer effect is achieved through multiple grinding passes.

[0052] Arc Segment Offset Calculation: For the arc corner portion of the hole, the trajectory is calculated using the equal-division point normal offset method. The arc segment is divided into a set number of points, and the arc normal direction at each point is calculated. The point is then offset outwards by a set distance along the normal direction to obtain the tool center trajectory point of the arc segment. The offset of each point is decomposed into X-direction offset components and Y-direction offset components, which are stored in the corresponding offset arrays to ensure a smooth and continuous arc segment trajectory without inflection points or abrupt changes. For circular holes, a uniform radial outward expansion offset method is used; for elliptical holes, the normal direction at each point is calculated based on the ellipse parametric equation, and the corresponding offset is performed.

[0053] Tool wear compensation superposition: The current tool height compensation value and diameter compensation value are obtained from the tool detection unit and superimposed on the above offset calculation results. Specifically, the height compensation value is superimposed on the Z-axis offset, and the diameter compensation value is superimposed on the planar hole reaming offset, realizing dynamic compensation for tool wear and ensuring that the grinding dimensions remain consistent after tool wear.

[0054] After completing all offset calculations, the actual grinding path point set for the corresponding grinding process is obtained. Each path point contains the three-dimensional coordinates of X, Y, and Z in the robot coordinate system.

[0055] Step S103. Generate the robot grinding trajectory based on the grinding path points, control the robot to perform grinding operations along the trajectory, and adjust the axial position of the tool in real time according to the height coordinates of each path point during the grinding process to ensure that the grinding depth is consistent at all points.

[0056] Specifically, firstly, the number and type of grinding processes are determined based on the type of holes and process requirements of the countertop. For example, square holes (like furnace holes) only require rough grinding to remove cutting burrs and marks; square holes (like basin holes) require three processes: rough grinding, fine grinding, and polishing, to gradually improve surface smoothness; round and oval holes also require these three processes. Each process corresponds to a different grit of grinding tool: low-grit tools for rough grinding, medium-grit tools for fine grinding, and high-grit tools for polishing.

[0057] The system matches corresponding tool parameters and feed rate parameters according to the process sequence, and generates a complete robot motion trajectory based on the grinding path points. Each process uses the same path point coordinates, but employs different feed rates and Z-axis offset parameters: the feed rate is lower in the rough grinding process to ensure grinding efficiency; the feed rate gradually increases in the fine grinding and polishing processes to ensure surface finish. The feed rate parameter is related to the remaining tool life. As the cumulative grinding length of the tool increases and the remaining tool life decreases, the feed rate gradually decreases according to a preset multi-segment decay curve, maximizing tool life while ensuring grinding quality.

[0058] Before the grinding operation begins, the robot picks up a grinding tool of the corresponding length from the automatic tool magazine based on the thickness parameters of the current sheet metal. This ensures that the tool's extension length is adapted to the sheet metal thickness, preventing the spindle from interfering with the sheet metal surface during grinding. After picking up the tool, the robot moves to the tool inspection station, where the tool inspection unit inspects the rotating tool, measuring its actual diameter and length, identifying any defects such as chipping or pitting on the tool surface, and updating the system parameters with the compensation values ​​obtained from the detection.

[0059] During the grinding process, the robot moves along the inner contour of the hole according to the generated trajectory. The constant force floating mechanism provides stable grinding pressure, ensuring a constant contact force between the tool and the workpiece surface. The robot reads the Z-axis coordinate value corresponding to each grinding path point in real time and dynamically adjusts the axial position of the tool to keep the tool at the set pressing depth. When the table panel has bending deformation or flatness deviation, the tool will rise and fall synchronously with the rise and fall of the panel height to ensure that the chamfer depth and fillet size of each side of the hole are consistent, avoiding quality problems such as excessive grinding on one side and insufficient grinding on the other side.

[0060] The system is equipped with a height value lower limit error prevention function, which verifies the rationality of the Z-axis coordinate of each path point in real time. When the coordinate value is detected to exceed the normal threshold range, the grinding action is stopped immediately and the system returns to a safe position to avoid the tool hitting the workpiece or equipment due to abnormal scanning data, thus ensuring production safety.

[0061] After a single grinding operation is completed, the robot automatically returns to the tool magazine to replace the tool for the next operation. After tool detection and compensation, it continues to execute the next grinding operation until all operations are completed. After all grinding is completed, the robot returns to the standby origin, the positioning and clamping unit releases the table panel, and the conveying unit sends the ground table panel out of the grinding station, while simultaneously feeding in the next piece of workpiece to be ground, thus starting the next work cycle.

[0062] For workpieces with dual holes, the system identifies the combination type and arrangement order of the holes during the scanning phase. It distinguishes the order of the holes by using marker variables, sets the start and end groups for cyclic grinding, and sequentially completes the grinding operation for all holes, achieving continuous multi-hole grinding in a single setup. The system supports skipping specific small holes, performing grinding only on designated main holes to meet diverse processing needs.

[0063] In some embodiments, the step of positioning and clamping the table panel to be polished and obtaining the hole type information of the table panel includes: performing a barcode scanning operation when the table panel to be polished flows into the feeding station to obtain the hole type, board size and order data of the table panel; adjusting the width of the lateral clamping mechanism according to the board size, performing lateral clamping and vertical pressing on the table panel, so that the conveying mechanism descends and disengages from the table panel to complete the positioning and fixing.

[0064] The worktable to be polished is placed on the feeding roller conveyor, which is driven by a servo motor and conveys forward at a set speed. When the worktable reaches the inlet barcode scanning station, a photoelectric sensor detects that the worktable has arrived and triggers the barcode scanner to read the barcode or QR code on the surface of the worktable. The identification information includes the product number of the worktable. Based on the product number, the system retrieves all processing information from the process database, including the corresponding hole type, worktable length, width, and thickness, number of holes, hole arrangement, and polishing process parameters. Hole types include various formats such as single square furnace holes, single square basin holes, single round holes, single oval holes, square hole + round hole combinations, and square hole + oval hole combinations; process parameters include preset values ​​for mesh count, feed rate, chamfer depth, and hole enlargement for each process.

[0065] After retrieving the information, the roller conveyor continues to transport the worktable panel into the grinding table. Multiple sets of photoelectric sensors installed at the grinding table entrance and inside the table monitor the position of the worktable in real time. When the front end of the worktable reaches the set stop position, the servo motor decelerates, and the worktable comes to a smooth stop. The stop position is dynamically adjusted according to the worktable length parameters to ensure that the deviation between the center of the worktable along its length and the center of the grinding table is within the allowable range. After stopping, the servo long-side clamping mechanisms on both sides of the grinding table are activated. The clamping jaws on both sides move synchronously towards the worktable under the drive of the servo motors. Flexible buffer pads are attached to the surface of the clamping jaws. They stop when they reach the set clamping force after contacting the side of the worktable, completing the centering and positioning of the worktable in the width direction. The centering accuracy can be controlled within 0.5mm.

[0066] After lateral clamping, multiple sets of lifting cylinders positioned below the grinding table surface move upwards synchronously. The non-metallic support pads at the top of the cylinders contact the bottom surface of the table panel, smoothly lifting the workpiece upwards and detaching its bottom surface from the roller conveyor surface. Simultaneously, corresponding downward pressing cylinders above the table extend downwards, their pressure plates contacting the upper surface of the workpiece, forming a vertical clamp with the lifting cylinders to completely fix the table panel in place. After clamping, the belt module of the roller conveyor retracts downwards, completely detaching from the bottom of the workpiece, eliminating interference from the conveyor structure and allowing space for grinding debris to fall and be collected. At this point, the table panel has completed six-degree-of-freedom positioning and clamping, providing a stable reference for subsequent scanning and grinding.

[0067] In some embodiments, the step of having a robot drive a three-dimensional line scanning camera to move at a constant speed along the length of the table panel to collect three-dimensional coordinate data of the inner contour of the hole to be polished on the table panel includes: the robot driving the three-dimensional line scanning camera integrated at the end to move at a constant speed along the length of the table panel, triggering the camera to collect images when it reaches a designated position, identifying and extracting the three-dimensional coordinate data of the straight edges and arc angles of each segment of the hole to be polished, and completing the scanning of the inner contour of the hole.

[0068] After positioning and clamping are completed, the robot control cabinet receives a scan start command from the host system. The robot moves the 3D laser line scanning camera integrated on the end flange to one end of the table panel along its length, adjusts the camera's attitude so that the laser line direction is parallel to the width direction of the table panel, and the camera's working distance is within the nominal optimal range. The robot sends a scan start command to the camera, and the camera completes initialization actions such as laser activation and sensor exposure parameter configuration, entering a ready-to-trigger state.

[0069] The robot then translates along the length of the platform at a constant linear velocity, matching the camera's line scanning frequency to ensure the point cloud spacing between adjacent scan lines meets accuracy requirements. During translation, when the robot reaches a preset IO trigger position, a hardware trigger signal is sent to the camera through the digital output port of the robot control cabinet, and the camera starts line-by-line scanning acquisition. Laser lines are projected onto the surface of the platform and the inner wall of the holes, forming bright contour lines. The camera's built-in imaging sensor receives the diffuse reflected light and calculates the three-dimensional coordinates of each scan point using the principle of triangulation.

[0070] The scanning range covers the entire width and length of the worktable, ensuring that all holes to be processed are completely scanned. When the robot reaches the scanning termination position at the other end of the worktable, it again notifies the camera to stop acquiring data via an IO signal and sends a scan end command to the camera. The width field of view of a single scan can reach hundreds of millimeters, and the length direction is unrestricted, adapting to different sizes of worktables ranging from 500mm to 3000mm in length.

[0071] The camera's built-in image processing unit preprocesses the raw point cloud data: First, a statistical filtering algorithm is used to remove outliers and eliminate abnormal points caused by dust, water stains, and surface reflections. Then, a reference plane on the upper surface of the platform is extracted through planar fitting, separating the platform area and the internal area of ​​the holes from the point cloud. Next, an edge extraction algorithm is used to identify the inner contour boundary of the holes, decomposing the closed contour into straight edge segments and arc transition segments. Finally, the contour points are resampled at equal intervals, outputting the three-dimensional coordinates of the uniformly distributed contour sampling points. The total acquisition and processing time is controlled within a few seconds, and the detection accuracy in both the planar and height directions is better than 0.3mm, accurately capturing the bending deformation of the board and the actual shape deviation of the holes.

[0072] In some embodiments, parsing the three-dimensional coordinate data and matching the offset calculation rules for the corresponding hole type includes: receiving the three-dimensional coordinate data obtained by scanning, identifying the type and number of holes to be ground, classifying and storing the coordinate points of each contour segment according to the hole type, and calling the trajectory offset calculation rules for the corresponding type.

[0073] After the 3D line scan camera completes contour extraction, it sends the processing results to the robot control cabinet via TCP / IP socket protocol in a predefined string format. The data string uses commas as delimiters, and the data segments are arranged as follows: the first bit is a scan success flag (1 for success, 0 for failure); the second bit is the number of holes; the third bit is the hole type code (1 for square, 2 for round, 3 for elliptical, 4 for square + round double hole, 5 for round + square double hole, etc.); the fourth bit is the number of equal divisions along the long side; the fifth bit is the number of equal divisions along the wide side; the sixth bit is the number of division points for the arc segment; starting from the seventh bit, every three consecutive values ​​form a group, corresponding to the X, Y, and Z coordinates of a contour sampling point, respectively. All values ​​are rounded to two decimal places, and there are no extra delimiters at the end of the string.

[0074] After receiving the complete string, the robot control cabinet first reads the first flag bit for status verification: if the flag bit is 0, an error code is recorded and a retransmission is requested from the camera. If multiple retransmissions fail, an alarm is triggered, prompting manual inspection of the workpiece and camera status; if the flag bit is 1, the data parsing process begins. The system calculates the total number of coordinate points by counting the number of commas in the string, verifies whether the data length matches the preset format, and if the verification passes, splits the string by commas, converts each character value to a floating-point value, and stores it in the global three-dimensional coordinate array.

[0075] After parsing, the system categorizes and stores the points in the coordinate array according to the hole type code and the number of holes, segmenting them by contour. For square holes, the points are stored sequentially in the order of long side, arc corner, wide side, and arc corner into the corresponding long side grinding point array, arc grinding point array, wide side grinding point array, and another arc grinding point array. Each array uses a two-dimensional structure, with the first dimension identifying the edge number and the front / back side, and the second dimension identifying the point number. For double-hole combination workpieces, the contour segments of the second group of holes are stored in the corresponding array partitions of the second group according to the hole arrangement order, and a combination marker variable is set to record the order of the holes. For circular holes, the entire arc is divided into four equal segments and stored in the arc grinding point array. For elliptical holes, the storage structure is the same as for square holes, storing the long straight side, short straight side, and transition arc separately. After categorization and storage, the system marks the corresponding trajectory offset algorithm type for each contour segment, preparing for subsequent offset calculations.

[0076] In some embodiments, the calculation of hole enlargement offset, chamfer height offset, and arc segment normal direction offset includes: expanding the original hole contour outward according to a preset single-sided hole enlargement amount to generate a tool center plane trajectory; setting a chamfer height offset to control the tool's pressing depth; and performing equal-point offset calculations on the arc segment contour along the normal direction to maintain the arc segment trajectory smooth and continuous.

[0077] After the contour data is classified and stored, the system calls the trajectory offset calculation subroutine for the corresponding hole type to perform offset calculations on the original contour coordinates, converting the inner wall contour of the hole into the center motion trajectory of the grinding tool. The offset calculation includes three core parts: hole enlargement offset, chamfer height offset, and arc segment normal offset.

[0078] The hole enlargement offset is used to control the amount of grinding on one side, achieving precise correction of hole dimensions. The system sets the total offset in the X and Y directions, with the offset values ​​in both directions being consistent, ranging from 1.2mm to 1.8mm, and a default value of 1.5mm. For straight-edge contours, the straight edge is directly translated outward along its normal direction by this offset distance to obtain the straight-edge trajectory at the tool center; the translation direction is perpendicular to the straight edge direction to ensure uniform grinding on one side. In actual production, the final grinding amount is related to the grinding feed rate, spindle speed, and tool grit. The faster the feed rate, the smaller the grinding amount per unit time. The system can fine-tune the hole enlargement offset according to the actual process parameters to compensate for dimensional deviations caused by the process parameters.

[0079] The chamfer height offset controls the grinding depth in the Z-axis direction, creating a chamfer effect on the hole edges. The system sets Z-axis offset parameters: a smaller Z-axis offset is used in rough grinding, typically set downwards by 1.2mm, to remove cutting burrs and marks; a slightly larger Z-axis offset is used in fine grinding and polishing, typically set downwards by 1.3mm, to gradually deepen the chamfer and improve surface finish. The Z-axis offset is directly superimposed on the Z-coordinate of each path point, controlling the tool's downward pressure depth.

[0080] The arc segment offset employs a normal direction equal division offset method to ensure a smooth and continuous trajectory at the arc transition. For the rounded corner transition segment of a square hole, the entire arc is evenly divided into 9 sampling points. For each sampling point, the normal direction of the arc at that point is calculated, extending outward along the arc radius. The offset is then performed outward along the normal direction by a predetermined hole enlargement amount, yielding the offset arc point coordinates. To improve computational efficiency, the system pre-calculates the X-direction and Y-direction offset component arrays corresponding to the 9 equally divided points. During calculation, the offset components are directly superimposed onto the original arc point coordinates, avoiding repeated calculations of trigonometric functions point by point. For circular holes, a uniform radial expansion method is used, with all points offset outward by the same distance along the radius. For elliptical holes, the normal vector of each point is calculated based on the ellipse parametric equation, and an equidistant offset is performed along the normal direction to ensure that the shape of the elliptical contour remains unchanged after offset.

[0081] After all offset calculations are completed, the initial tool center trajectory point is obtained. The trajectories are connected end to end to form a closed grinding path.

[0082] In some embodiments, the step of superimposing tool wear compensation parameters to generate actual grinding path points includes: obtaining the height compensation value and diameter compensation value obtained by tool detection, superimposing the tool wear compensation parameters onto the offset contour trajectory, and generating the actual grinding path points corresponding to the grinding process.

[0083] Based on the calculation of the basic trajectory offset, the system superimposes tool wear compensation parameters to generate the final actual grinding path points, thus eliminating the influence of tool wear on the grinding dimensions.

[0084] The tool compensation parameters are derived from real-time measurements by the tool detection unit. After each tool change, the robot moves the tool to a fixed tool detection station. This station is equipped with a high-resolution area array camera and a backlight source. The backlight source projects parallel light from behind the tool, creating a clear silhouette of the tool. During detection, the electric spindle rotates the tool at a constant speed, and the camera continuously acquires multiple frames of tool silhouette images. A sub-pixel edge extraction algorithm is used to measure the tool's diameter and effective length, and to identify defects such as chipping, dents, and runout on the cutting edge. The diameter detection accuracy can reach 0.02 mm, and the length detection accuracy is also at the micrometer level.

[0085] After measurement, the system calculates the difference between the actual and nominal diameter of the tool as the diameter compensation value, and the difference between the actual and nominal length as the height compensation value. Both compensation values ​​are stored in the tool parameter array, with the array index corresponding to the tool station number in the tool magazine. The diameter compensation value is used to correct the planar reaming amount; when the tool wear diameter decreases, the reaming offset is reduced accordingly to ensure the actual grinding dimensions remain unchanged. The height compensation value is used to correct the Z-axis downward pressure depth; when the tool wear length decreases, the Z-axis downward pressure is increased accordingly to ensure consistent chamfer depth.

[0086] During the trajectory calculation phase, the system reads the compensation value corresponding to the currently installed tool, adds the diameter compensation value to the hole reaming offset, and adds the height compensation value to the Z-axis chamfer offset, then recalculates the final coordinates of each path point. For multi-process grinding, the compensation parameters are updated every time the tool is changed, ensuring that the trajectory of each process adapts to the actual size of the current tool, achieving full-process dynamic closed-loop compensation for tool wear. After the compensation is superimposed, a set of actual grinding path points for the corresponding process is generated, with each point containing complete three-dimensional coordinates and process attributes.

[0087] In some embodiments, generating a robot grinding trajectory based on grinding path points includes: matching tool parameters and feed rate parameters of the corresponding mesh size according to the grinding process requirements corresponding to the hole position, and generating a robot grinding trajectory containing multiple grinding processes in combination with the grinding path points.

[0088] After generating the basic grinding path points, the system plans multiple grinding processes based on the hole type and process requirements, matches the corresponding tools and process parameters, and generates a complete robot grinding operation trajectory.

[0089] The system has built-in process templates for various hole types: square holes for furnaces only require one rough grinding process, using a low-grit rough grinding tool, mainly to remove marks and burrs left by waterjet cutting, and to meet basic chamfering requirements; square holes for basins require three processes, namely rough grinding, fine grinding, and polishing, using 50-grit, 180-grit, and 500-grit tools respectively, to gradually improve the smoothness of the hole edges and the roundness of the corners; round holes and elliptical holes also require three processes: rough grinding, fine grinding, and polishing, corresponding to different offset calculation subroutines.

[0090] Each process shares the same set of contour coordinates, but uses different feed rates and Z-axis offset parameters. The system is configured with multiple sets of speed parameters: the rough grinding process has two feed rates, with a higher rough grinding linear speed for straight edges and a slightly lower speed for curved edges to ensure uniform grinding at the rounded corners; the fine grinding process uses a medium feed rate to balance efficiency and surface quality; the polishing process uses a higher feed rate to achieve a high-gloss finish. The speed unit is uniformly set to millimeters per second and can be flexibly adjusted according to actual production needs.

[0091] The feed rate is linked to the remaining tool life. The system sets a cumulative grinding life for each tool, with a default total life of 800 meters of grinding path length. As the cumulative grinding length increases, the remaining tool life gradually decreases, and the feed rate decreases step by step according to a preset six-segment decay curve: the rated maximum speed is used when the tool is new, and the speed is reduced by one level for each certain percentage of tool life consumed. After tool wear, the speed is reduced to ensure grinding quality while maximizing the effective utilization of the tool. In addition, different tool lengths correspond to different base speeds. For longer tools, due to their larger overhang and slightly lower rigidity, the base speed is appropriately reduced proportionally to avoid machining vibration.

[0092] The system integrates the path points, feed rate, Z-axis offset, tool change action, and detection action of each process to generate a complete robot motion program, which includes spatial motion instructions, IO control instructions, and logical judgment instructions, and can be directly loaded into the robot for execution.

[0093] In some embodiments, the control robot to perform grinding operations along a trajectory includes: controlling the robot to change the grinding tool for the corresponding process, performing grinding operations along the inner contour of the hole according to the generated grinding trajectory, automatically switching tools after each process is completed, and sequentially completing all grinding processes.

[0094] The grinding operation is carried out in the preset sequence of procedures. The tool change and inspection are completed automatically before each procedure to ensure machining accuracy.

[0095] The automatic tool magazine has 9 tool stations, arranged in a matrix according to the sheet thickness and tool grit number. It covers three commonly used sheet thicknesses (15mm, 20mm, and 25mm) and three grit numbers (50, 180, and 500). Each sheet thickness corresponds to three tools with different grit numbers. The system automatically selects the corresponding length series of tools based on the thickness parameters of the current table panel, ensuring that the tool extension length is suitable for the sheet thickness and avoiding interference between the electric spindle flange and the sheet surface.

[0096] During process changeover, the robot carries the current tool to the automatic tool magazine, returns the tool to the corresponding station, and completes the tool unloading action. It then moves to the tool station for the next process, grabs a new tool, and locks it in place, completing the tool grabbing action. After tool change, the robot automatically moves to the tool inspection station to perform tool size detection and compensation value update. If severe chipping, diameter, or length exceeding the wear limit is detected, an alarm is immediately triggered, prompting the replacement of the tool to prevent workpiece scrap due to defective tools.

[0097] After tool inspection, the robot moves to a safe starting position above the hole and begins grinding according to the generated grinding trajectory. During grinding, the constant force floating mechanism adjusts the output pressure in real time to maintain a constant contact force between the tool and the workpiece. The system has three built-in constant force modes: coarse grinding uses a larger contact force to ensure grinding efficiency, fine grinding uses a medium contact force, and polishing uses a smaller contact force to ensure surface quality. The constant force floating mechanism has a 20mm axial floating stroke, which can absorb fluctuations in plate height and robot motion errors, avoiding chipping caused by hard contact of the tool.

[0098] The robot operates continuously along a closed trajectory, completing one grinding cycle. It can repeat this cycle multiple times according to process requirements to ensure the required grinding amount is achieved. After a single process is completed, the robot raises the tool to a safe height, returns to the tool magazine to change to the next tool, and then executes subsequent processes sequentially until all processes are completed. For workpieces with two holes, after completing all processes for the first hole, the robot automatically moves to the second hole to continue grinding, eliminating the need for repeated clamping and enabling continuous multi-hole machining. The system supports selectively skipping specified holes; for example, the small hole in a faucet can be set to not be ground, allowing only furnace holes and basin holes to be machined, adapting to diverse product needs.

[0099] In some embodiments, the axial position of the tool is adjusted in real time according to the height coordinates of each path point during the grinding process to ensure that the grinding depth is consistent throughout. This includes: reading the height coordinate value corresponding to each grinding path point during the grinding process, adjusting the axial position of the tool in real time, compensating for the bending deformation of the table panel, verifying the rationality of the height coordinates, intercepting abnormal values, and ensuring that the grinding depth is consistent throughout.

[0100] During the grinding process, the system performs real-time adjustment of the tool's axial position based on the Z-axis coordinates of each path point obtained by scanning, achieving height-following grinding and compensating for the bending deformation of the table panel.

[0101] Because quartz stone and slab countertops may bend and deform during production, handling, and placement, and the height of different parts of the board may vary, if a fixed Z-axis height is used for grinding, the grinding amount on one side will be large and the grinding amount on the other side will be small, resulting in uneven chamfers and inconsistent radius corner sizes. In this embodiment, each path point of the grinding trajectory carries an actual measured Z-axis height value. When the robot moves to the corresponding point, the Z-axis position synchronously follows this height value, so that the tool always maintains a constant pressing depth relative to the board surface.

[0102] In practice, during trajectory interpolation, the robot performs linear interpolation on the Z-values ​​between adjacent path points to achieve a smooth transition in the height direction and avoid impact caused by abrupt changes in the Z-axis. When there is a height difference of about 1.5mm between the two sides of the table panel, the deviation of the chamfer depth at each location can be controlled within 0.3mm through real-time height tracking, and the size of the radius of each edge is basically consistent with the polishing effect, significantly improving the consistency of polishing quality.

[0103] Meanwhile, the system is equipped with a Z-value lower limit error prevention mechanism, which double-checks the rationality of the Z-axis coordinates before and during trajectory execution. Before execution, the system compares the Z-values ​​of all path points with the height of the table's reference plane. If the Z-value of a point is lower than the lower threshold or higher than the upper threshold, it is judged as abnormal data, automatically removed, and replaced by interpolation of adjacent points, or the grinding process is directly terminated. During execution, the system monitors the load current of the robot's Z-axis and the displacement of the constant force floating mechanism in real time. If a sudden change in load or the displacement exceeds the limit occurs, the feed is immediately stopped and the tool is lifted to prevent the tool from colliding with the workpiece or the table. Through the combination of height following and error prevention mechanisms, both the grinding quality of deformed plates and the safety of equipment operation are ensured.

[0104] In some embodiments, such as Figure 2 As shown, the grinding system in this embodiment consists of a feeding and conveying module, a grinding worktable, a six-axis industrial robot, a three-dimensional laser line scanning camera, a tool detection area array camera, a constant force floating mechanism, a water-cooled electric spindle, an automatic tool magazine, a PLC control unit, and a robot control cabinet. All units operate collaboratively through an industrial bus.

[0105] Six-axis industrial robot: maximum load not less than 200kg, maximum arm span not less than 2600mm, end flange for mounting grinding actuator, with continuous path interpolation function, can realize smooth spatial trajectory movement, and meet the full-area processing needs of large-size table panels.

[0106] 3D laser line scan camera: Integrated and mounted on the robot's end effector, equipped with an automatically opening and closing sealed protective cover. The cover automatically opens during camera operation and automatically closes during grinding operations, preventing water mist and dust from corroding the optical lens and maintaining long-term detection accuracy. The camera's maximum field of view is no less than 550mm, with no limit on the length direction. The detection accuracy in the planar direction is ±0.2mm, and the detection accuracy in the height direction is ±0.2mm. The single image acquisition time is less than 2 seconds, and the built-in algorithm processing time is approximately 2 seconds. The camera interacts with the robot control cabinet via TCP / IP socket protocol.

[0107] Tool inspection area array camera: with a resolution of 12 million pixels, using backlight illumination, a field of view of no more than 60mm, and a dimensional inspection accuracy of up to 0.02mm; the camera is fixedly installed at the inspection station next to the tool magazine and is used for online measurement of tool dimensions and defect identification after tool change.

[0108] Constant force floating mechanism: Connected between the robot end flange and the water-cooled electric spindle, the axial floating range is 20mm, and it supports three force control output conditions, corresponding to contact forces of 15N, 30N and 200N respectively, which can adapt to the pressure requirements of different processes such as rough grinding, fine grinding and polishing; the mechanism has a 0.5-second collision rapid retreat function, which can quickly retract in case of abnormal collision to protect the robot and the workpiece.

[0109] Water-cooled electric spindle: rated power 7.5kW, rated speed 12000r / min, supports automatic tool change; the heat generated during grinding is removed by water cooling circulation, maintaining a stable spindle operating temperature, suitable for long-term continuous grinding operations.

[0110] Automatic tool magazine: It has a total of 9 tool stations, which can accommodate grinding tools of three length specifications and three grit specifications, and are compatible with table panels of 15mm, 20mm and 25mm thickness, as well as three processes of 50 grit rough grinding, 180 grit fine grinding and 500 grit polishing, to meet the needs of all process processing.

[0111] PLC control unit: Communicates with the robot control cabinet via the ProfiNet industrial bus protocol, and is responsible for workpiece clamping logic control, conveyor start / stop control, and tool life accumulation management.

[0112] Robot control cabinet: Responsible for robot motion control, visual data analysis, trajectory offset calculation and polishing program execution, it is the core of the entire system's operation and control.

[0113] Grinding worktable: Equipped with a servo-driven long-side clamping mechanism, multiple sets of vertical lifting and pressing cylinders, and a liftable belt conveyor mechanism, it can achieve centering and secure clamping of table panels of different sizes.

[0114] The system can process tabletops with dimensions ranging from 500 to 3000 mm in length, 500 to 1200 mm in width, and 15 / 20 / 25 mm in thickness, with a single panel weight ranging from 20 to 180 kg. It can also process holes ranging from 290 to 900 mm × 300 to 500 mm, covering various hole types such as square holes, round holes, oval holes, and combinations of double holes.

[0115] The entire polishing process is divided into eight stages in sequence: board scanning and positioning, 3D contour scanning, data parsing and storage, trajectory offset calculation, height-following polishing, multi-pass processing, automatic tool change detection, and continuous polishing of dual holes. Specifically, it includes the following steps: Step 1: Panel scanning and positioning clamping: When the grinding table panel flows into the grinding station entrance along the feeding roller line, the scanning device installed at the entrance reads the identification code on the panel surface. The system retrieves the corresponding panel hole type, panel size parameters, processing requirements and order-related data from the production database.

[0116] The grinding table adjusts the stop position of the workpiece according to its length data, roughly aligning the center of the workpiece with the center of the worktable. After the workpiece is in place, the servo long-side clamping mechanism moves synchronously from both sides of the workpiece towards the center, clamping the workpiece laterally. The clamping force is controlled within a reasonable range to ensure that the workpiece does not shift or cause edge chipping. Subsequently, multiple sets of lifting cylinders on the worktable extend upwards simultaneously, with the flexible pads at the top of the cylinders contacting the bottom surface of the table panel, smoothly lifting the workpiece upwards. This, combined with the pressing structure above, completes the vertical fixation. After clamping, the belt conveyor mechanism inside the worktable descends as a whole, completely detaching from the bottom surface of the table panel, eliminating interference from the conveyor structure in the grinding process, and reserving space for grinding debris to fall, thus completing the positioning and clamping of the workpiece.

[0117] The sheet metal conveying and positioning cycle time for this step is approximately 20 seconds per sheet.

[0118] Step 2: 3D laser line scanning to acquire hole contours: After the platform is clamped and fixed, the PLC sends a positioning completion signal to the robot control cabinet, and the robot starts the scanning process. The robot moves the 3D laser line scanning camera at the end effector to the starting end in the length direction of the board, adjusts the camera posture so that the laser line direction is parallel to the width direction of the board, and the working distance is within the camera's optimal measurement range.

[0119] The robot first sends a start signal to the camera. After the camera completes laser activation and exposure parameter configuration, it enters standby acquisition mode. Then, the robot moves the camera along the length of the workpiece at a constant speed, matching the camera's line frequency to ensure continuous, uninterrupted scanning data. During this uniform movement, when the robot reaches a preset trigger position, a digital I / O signal triggers the camera to begin line-by-line acquisition. The camera uses laser triangulation to acquire depth data of the platform surface and the inner wall of the holes line by line. An internal algorithm identifies and extracts the X, Y, and Z coordinates of the four straight edges of the inner hole to be polished, as well as the X, Y, and Z coordinates of the four rounded corners. The scan covers the full width of the workpiece, and the length direction adaptively adjusts according to the workpiece size, eliminating the need for pre-setting hole positions.

[0120] After scanning is complete, the robot sends a stop signal to the camera, and the camera stops acquiring data. A single scan takes approximately 20 seconds per piece, allowing for the simultaneous acquisition of the contours of all holes on a single board.

[0121] Step 3: Contour Data Parsing and Coordinate Classification Storage: The 3D laser line scanning camera sends the scanning results in string format to the robot control cabinet via TCP / IP socket protocol. The string data uses commas as delimiters, and the format is defined as follows: the first 3 digits are the image capture success flag (1 for success, 0 for failure), the number of holes, and the hole type code (1 for square hole, 2 for round hole, 3 for elliptical hole); the next 3 digits are the number of equal divisions of the long side, the width side, and the number of arc division points; the subsequent 3 digits are grouped into sets of 3, corresponding to the x, y, and z coordinates of the next sampling point in the robot coordinate system. All values ​​are rounded to two decimal places, and there are no extra delimiters at the end of the string.

[0122] After receiving the complete string of data, the robot control cabinet first verifies the success flag at the beginning. If the flag indicates failure, a retransmission command is sent to the camera. If multiple retransmissions fail, a fault is reported for manual intervention. After successful verification, the system uses comma counting to count the total length and the total number of coordinate points. After verifying the data format match, the character data is converted to numerical data and stored in the global 3D coordinate array receive_pos.

[0123] The system then encodes the hole type and stores the coordinate points into corresponding grinding position arrays, categorizing them by contour segmentation. For square holes: Store the coordinates of the four straight edges and four arcs into the long edge grinding point array DM_L1, the arc grinding point array DM_C2, the wide edge grinding point array DM_W3, and the arc grinding point array DM_C4 respectively; if it is a double-hole workpiece, store the second set of hole positions into the arrays DM_L5, DM_C6, DM_W7, and DM_C8 respectively.

[0124] Circular hole: Divide the entire arc into 4 segments and store the coordinate points in the arc grinding point array DM_C.

[0125] Elliptical holes: Similar to square holes, array structures are used to store the coordinates of the long straight side, the short straight side, and the transition arc. The arc segment is then offset using the elliptical parametric equation.

[0126] Step 4: Grinding Trajectory Offset Calculation: The system calls the corresponding trajectory offset calculation program based on the hole type: DM_fang_rot for square holes, DM_yuan_rot for round holes, and DM_tuo_yuan_rot for elliptical holes. This program performs offset calculations on the original contour coordinate points, converting the inner wall contour of the hole into the center motion trajectory of the grinding tool. The offset calculation includes the following four parts: Hole reaming offset calculation: Set offset_x and offset_y as the total grinding offset, keeping their values ​​consistent. The range is 1.2~1.8mm; in this embodiment, the default setting is 1.5mm, used to control the single-sided hole reaming grinding amount. Using the original hole inner contour as a reference, extend this offset distance outward along the contour normal direction to generate the center movement trajectory of the grinding tool. The actual grinding amount is inversely proportional to the grinding feed rate. Adjusting the offset can compensate for the difference in grinding amount caused by speed variations, ensuring that the final hole size meets design requirements.

[0127] Chamfer height offset calculation: Set offset_z as the chamfer height parameter. The default setting for rough grinding is -1.2, and the default setting for fine grinding and polishing is -1.3. This is used to control the depth of the tool's downward pressure in the Z-axis direction, so as to form a uniform chamfer effect at the edge of the hole.

[0128] Circular arc segment offset calculation: For the circular arc segment contour, the offset of 9 equally divided points is calculated using the trigonometric function method. The offset of each point is decomposed into an X-direction offset array offset_c_x{9} and a Y-direction offset array offset_c_y{9}. The offset direction of each arc point is along the arc normal direction at that point, ensuring that the grinding trajectory of the circular arc segment is smooth, continuous, and without inflection points or abrupt changes. Circular holes are calculated using an offset method that expands outward in all four directions around the entire perimeter. Elliptical holes are calculated based on the ellipse parametric equation to determine the normal direction of each point, and corresponding equidistant offsets are performed to ensure that the elliptical shape is not distorted after offset.

[0129] Tool wear compensation superposition: Read the current tool compensation parameter array DM_dao_bu fed back by the tool detection camera, where the first bit is the tool height compensation value and the second bit is the tool diameter compensation value. Superimpose the compensation values ​​onto the Z-axis offset and the planar hole expansion offset respectively to realize real-time dynamic compensation for tool wear and eliminate the influence of tool size changes on grinding quality.

[0130] Step 5: Z-axis height-following grinding: The robot generates a motion interpolation trajectory based on the calculated grinding path points. During the grinding process, the Z-axis coordinate value corresponding to each path point is read in real time, and the Z-axis axial position of the tool is dynamically adjusted. When the table panel has bending deformation, installation deformation, or flatness deviation, the tool rises and falls synchronously with the height fluctuation of the panel to ensure that the grinding chamfer depth of each side of the hole is consistent, avoiding quality defects such as excessive grinding on one side and insufficient grinding on the other side.

[0131] The system also has a Z-value lower limit error prevention function. Before the trajectory is executed, the Z-axis coordinates of all path points are checked for rationality, abnormal data exceeding the normal threshold are intercepted, noise points are removed and the data is completed by interpolation of adjacent points. During the grinding process, the Z-axis load and the displacement of the constant force floating mechanism are monitored in real time. If abnormal changes occur, the feed is stopped and the tool is lifted immediately to prevent abnormal scanning data from causing equipment collisions or workpiece damage.

[0132] Actual testing has verified that when there is a height difference of about 1.5mm between the two sides of the tabletop, automatic grinding along the Z-axis results in minimal difference between the size of the rounded corners and the surface grinding effect, and the chamfer depth deviation can be controlled within 0.3mm.

[0133] Step 6: Multi-pass grinding process execution: The system executes the corresponding grinding process according to the hole type and process requirements. Each process uses the same trajectory position point, but matches different feed speeds and Z-axis offset parameters: For square holes in furnace holes: only one rough grinding process is performed, using a 50-grit rough grinding tool and running the DM_fang_1 grinding program. This is mainly used to remove waterjet cutting marks and burrs, and to form a basic chamfer.

[0134] For square holes of the basin type: perform three processes in sequence: rough grinding, fine grinding and polishing, using 50-grit, 180-grit and 500-grit tools respectively, and running the corresponding DM_fang_1, DM_fang_2 and DM_fang_3 programs to gradually refine the surface texture and improve the smoothness of the hole edges.

[0135] Round holes / elliptical holes: Perform one coarse grinding, one fine grinding, and one polishing in sequence, corresponding to the DM_yuan_1 / DM_tuo_yuan_1, DM_yuan_2 / DM_tuo_yuan_2, and DM_yuan_3 / DM_tuo_yuan_3 programs respectively.

[0136] Feed rate parameters include rough grinding linear speed (including two speed segments: straight edge segment and arc segment), fine grinding linear speed, and polishing linear speed, all measured in millimeters per second. These speed parameters are related to the remaining tool life. As the remaining tool life decreases, the feed rate automatically decreases step-by-step according to a six-segment speed decay curve, maximizing tool life while ensuring grinding quality. For tools with lengths of 20mm and 25mm, the base feed rate is reduced by a preset ratio compared to a 15mm tool to avoid machining vibrations caused by the large overhang and low rigidity of longer tools.

[0137] In this embodiment, the rough grinding linear speed is greater than 0.7 m / min, with a peak value of 1 m / min; the fine grinding linear speed is greater than 1.4 m / min, with a peak value of 1.8 m / min; and the polishing linear speed is greater than 1.8 m / min, with a peak value of 2.5 m / min.

[0138] Step 7: Automatic tool change and tool detection and compensation: Based on the thickness data of the worktable, the robot automatically selects a grinding tool of the corresponding length from the tool magazine to ensure that the tool extension length is adapted to the thickness of the workpiece and to avoid interference of the spindle flange with the workpiece surface.

[0139] After each tool change, the robot moves the tool to the tool inspection station. A water-cooled electric spindle rotates the tool at a constant speed. The tool inspection camera, using backlighting, captures multiple frames of the rotating tool. A sub-pixel edge extraction algorithm detects the tool's diameter (range 15-17mm), effective length (range 60-100mm), and the presence of defects such as pits and chips on the cutting edge surface. The measured height and diameter compensation values ​​are fed back to the grinding system, updating the tool compensation parameters in real time. A single tool inspection cycle is approximately 8 seconds. If the detected tool wear exceeds the limit or has serious defects, the system immediately triggers an alarm, prohibiting the tool from being used in machining.

[0140] The tool life is managed by the PLC unit by accumulating the total grinding trajectory length of each board. The set total life of a single tool is 800 meters. When the cumulative grinding length of the tool reaches the set life, the system controls the robot to return to the standby origin and stop before the next tool grab, prompting the operator to replace the tool with a new one. Specifically, the 50-grit coarse grinding tool loses its grinding ability when its diameter wears down to 15.7mm, and the 180-grit fine grinding tool and the 500-grit polishing tool lose their polishing ability when their diameter wears down to 15.9mm. Corresponding wear limit thresholds are set.

[0141] Step 8: Continuous Grinding of Dual-Hole Combination Workpieces: When the scanning system identifies a dual-hole combination workpiece (such as a combination of square holes and round holes), it distinguishes the arrangement order of the two hole positions by using the receive_fangyuan and receive_yuanfang marker variables. It sets the dm_start and dm_end variables to control the starting and ending hole groups of the cyclic grinding, automatically completing all grinding processes for all hole positions in sequence, realizing continuous automatic grinding of multiple holes in a single clamping without the need for repeated positioning and clamping.

[0142] The system also supports hole selection, which can exclude holes that do not need to be polished, such as small holes for faucets, according to process requirements, and perform polishing operations only on specified holes such as stove holes and basin holes, adapting to diverse product processing needs.

[0143] After all the grinding processes are completed, the robot returns to the standby origin. The PLC controls the grinding worktable to release the clamping mechanism, and the belt conveyor mechanism rises to lift the table panel, sending the ground table panel out of the grinding station. At the same time, the next piece of workpiece to be ground is sent in, and the next work cycle begins.

[0144] The system described in this embodiment can complete the grinding of 50-60 individual panels in a single-shift 9.5-hour production mode, with an equipment utilization rate of approximately 90%, meeting the needs of large-scale industrial production. Compared to traditional manual teaching grinding methods, this solution eliminates the need for manual teaching of the trajectory for each product, reducing product changeover time from several hours to just seconds. It can efficiently adapt to flexible production modes with multiple varieties and small batches. At the same time, through technologies such as Z-axis height following and real-time tool compensation, it significantly improves the consistency of grinding quality and product qualification rate, effectively solving industry pain points such as uneven chamfering caused by panel deformation and quality degradation caused by tool wear.

[0145] In other alternative implementations, the same technical effect can be achieved through the following alternative forms: The 3D laser line scan camera can be replaced by a structured light 3D camera or a binocular stereo vision system, as long as the 3D coordinate data of the hole contour can be obtained, trajectory generation can be achieved. The TCP / IP socket communication protocol can be replaced with UDP, ModbusTCP or other industrial Ethernet protocols to achieve data transmission between the camera and the robot; The calculation of trigonometric function offset for circular arc segments can be replaced by Bézier curve fitting or B-spline curve fitting methods to achieve a smooth transition of the circular arc segment trajectory. The passive constant force floating mechanism can be replaced by a main power control scheme that combines a six-dimensional force sensor with a robot force control algorithm to achieve more precise constant force grinding; The PLC-based trajectory length accumulation lifespan management method can be replaced by an internal robot counter accumulation or a counting method based on the number of grinding cycles. A six-axis industrial robot can be replaced by a five-axis parallel robot or a SCARA robot, as long as the arm span and load meet the requirements of the grinding operation.

[0146] In some embodiments, such as Figure 2 As shown, the grinding equipment is arranged along its length as follows: a feeding roller section, a grinding station section, and a discharging roller section. The grinding station section contains a positioning and clamping mechanism, a robotic grinding unit, a tool magazine, and a tool detection unit. The double-hole combination panel to be ground is fed into the feeding roller section from the upstream station. The inlet barcode scanner reads the panel's markings, confirming that the panel is a combination of square and round holes, with a thickness of 20mm. The basin holes require three grinding processes, while the furnace holes require a rough grinding process.

[0147] Based on the parameters of a plate length of 3000mm and a width of 1200mm, the system adjusts the stop position of the grinding station and the opening of the long-side clamping mechanism. After the plate enters the grinding station, it comes to a smooth stop. The long-side clamping mechanism clamps the plate from both sides, and the lifting cylinder and the pressing cylinder work together to complete the vertical fixation. The belt mechanism descends and disengages from the bottom of the plate. After positioning, the six-axis robot drives the end-effector 3D line scanning camera to move laterally at a constant speed from one end of the plate, covering the entire width of the plate in a single scan, and sequentially acquiring the 3D data of the inner contours of the square basin hole and the circular furnace hole.

[0148] The camera packages and sends the scanned data to the robot control cabinet. The data includes the number of holes (2), the hole type code (4, square + round combination), and the complete outline coordinates of the two holes. After parsing the data, the system stores the four sides and four rounded corners of the square holes in the first grinding array, and the entire arc of the round holes in the second grinding array. At the same time, the hole order is marked as square holes first, round holes second.

[0149] The system selects 20mm long cutting tools based on a 20mm plate thickness and plans the grinding sequence: first, complete the rough grinding, fine grinding, and polishing of the square basin hole, then complete the rough grinding of the circular furnace hole. The robot first picks up a 50-grit rough grinding tool from the tool magazine, obtains compensation values ​​through tool detection, and adds them to the trajectory offset calculation for the square hole to generate the rough grinding path points; then, it performs rough grinding of the square hole, with the Z-axis following the changes in plate height in real time during the grinding process. After rough grinding is completed, it switches to a 180-grit fine grinding tool, performs fine grinding after detection and compensation; then, it switches to a 500-grit polishing tool and performs polishing.

[0150] After all three processes for the square basin hole are completed, the robot changes to a 50-grit coarse grinding tool, calls the circular hole trajectory offset subroutine to generate the coarse grinding trajectory for the circular hole, and performs the coarse grinding operation for the circular furnace hole. After all holes are ground, the robot returns to the standby origin, the clamping mechanism releases, the belt mechanism rises to support the plate, the roller conveyor starts to send the ground plate to the discharge roller section, and at the same time the next plate to be ground is fed in from the feeding end, starting the next work cycle.

[0151] The entire process requires no manual teaching of the trajectory or manual adjustment of parameters. Different hole types and sizes of plates can be automatically identified and processed. Changeover time is only a few seconds of scanning, which greatly improves the efficiency of multi-variety, small-batch production. A single shift of 9.5 hours can complete the processing of 50 to 60 products, and the equipment utilization rate can reach over 90%, meeting the needs of industrial mass production.

[0152] In some embodiments, while acquiring the point cloud of hole contours, the system simultaneously collects global point cloud data of the upper surface of the platform, selecting the four corner points and multiple reference edges of the platform as feature point sets. The measured point cloud is iteratively registered with the standard flat plate model to solve for the overall bending deformation, warping deformation, and torsional deformation parameters of the plate, and a surface deformation model of the plate is constructed.

[0153] When generating the grinding trajectory, instead of relying solely on the Z-value of a single point for discrete tracking, a continuous surface fitting is performed on the entire grinding path based on a surface deformation model to calculate the theoretical board height at any position on the trajectory. During the robot's movement, the Z-axis position is smoothly adjusted in real time according to the fitted surface. Compared to discrete point linear interpolation, the trajectory has higher smoothness, and the Z-axis movement has no abrupt changes, further improving the uniformity of the chamfer and the surface finish after grinding.

[0154] Meanwhile, the system can automatically adjust the grinding feed speed according to the amount of deformation: it automatically reduces the feed speed in areas with large deformation curvature to ensure uniform grinding amount; and it maintains high-speed feed in flat areas to balance processing efficiency and processing quality. For plates with deformation exceeding the threshold, the system can determine them as unqualified products in advance, skip the grinding process directly and divert them to avoid ineffective processing.

[0155] In some embodiments, this embodiment is an additional intelligent operation and maintenance embodiment, which realizes accurate prediction of tool life and optimization of tool change scheduling through historical data modeling.

[0156] The system accumulates the total grinding trajectory length of each tool in real time, and records process parameters such as machining hole type, process type, feed rate, and grinding pressure for each tool. Based on a large amount of historical tool wear data, a wear prediction model is constructed. By inputting features such as cumulative machining length, process parameters, and plate material, the model predicts the current actual diameter, edge wear, and remaining effective life of the tool.

[0157] When the remaining tool life is predicted to be insufficient to complete all grinding operations on the current workpiece, the system will automatically schedule tool changes before processing to avoid tool failure during processing and resulting in workpiece scrap. When the remaining tool life is sufficient to complete multiple workpieces, the system will rationally arrange continuous processing of workpieces of the same type to reduce the number of tool changes and improve the overall uptime.

[0158] Meanwhile, the system can compare the actual life curves of tools of the same specifications to identify tools with abnormal wear and provide early warnings of tool quality problems or abnormal process parameters. The life prediction model can be continuously iterated and optimized as production data accumulates, and the prediction accuracy gradually improves, increasing the average tool utilization rate by more than 10% while ensuring grinding quality.

[0159] In some embodiments, after all grinding processes are completed, the robot drives a 3D line scan camera to perform a secondary scan of the holes, measuring quality parameters such as the chamfer dimensions, fillet radii, and surface roughness characteristics after grinding. The measured quality parameters are compared with standard process requirements to calculate dimensional deviations and surface quality deviations.

[0160] If the deviation is within the allowable range, the workpiece is deemed qualified and discharged normally; if the deviation exceeds the threshold but is still within the correctable range, the system automatically calculates the compensation amount, adjusts the hole expansion offset, Z-axis offset and feed speed parameters, and automatically executes a grinding process to correct the quality deviation; if the deviation is too large to be corrected, it is deemed unqualified, triggers an alarm and is diverted for processing.

[0161] Simultaneously, the system feeds back deviation data to the process parameter library, self-learning and correcting the default offset parameters for similar hole positions and tools from the same batch, gradually reducing system deviation and achieving continuous iterative optimization of grinding quality. Through online detection and closed-loop correction, the first-pass yield rate can be increased to over 99%, significantly reducing manual rework workload.

[0162] In some embodiments, when a workpiece contains two or more holes, the system obtains the position coordinates of all holes and the machining process requirements, and uses a genetic algorithm for global machining path planning. The planning objective is to minimize the total machining stroke and the number of tool changes. Taking into account constraints such as hole machining sequence, tool change sequence, and robot obstacle avoidance, the optimal machining sequence is solved.

[0163] For example, for workpieces with both square and round holes, the system compares the total time consumed by various sequences such as "three processes for square hole → rough grinding of round hole" versus "rough grinding of square hole → rough grinding of round hole → fine grinding of square hole → polishing of square hole," and selects the solution with the shortest total travel and the fewest tool changes. For complex workpieces with more holes, this can significantly reduce the robot's idle travel and the number of tool changes, shortening the single workpiece machining cycle time by 10% to 15%.

[0164] Meanwhile, the system automatically embeds tool detection actions into the path planning, completing tool detection during gaps without affecting the main process flow, thus avoiding additional processing cycles. The global path planning results directly generate the robot motion program, eliminating the need for manual arrangement of processing sequences and adapting to the flexible production needs of multiple varieties and multi-hole types.

[0165] Please see Figure 3 As shown, Figure 3This is a schematic diagram of the structure of a 3D vision scanning automatic trajectory generation system 200 for tabletop robot grinding, provided in an embodiment of this application. This 3D vision scanning automatic trajectory generation system 200 is used to execute the steps of the 3D vision scanning automatic trajectory generation method for tabletop robot grinding shown in the above embodiments. The 3D vision scanning automatic trajectory generation system 200 can be a single server or a server cluster, or it can be a terminal, such as a handheld terminal, a laptop computer, a wearable device, or a robot.

[0166] like Figure 3 As shown, the 3D vision scanning-based automatic trajectory generation system 200 for tabletop robot grinding includes: The positioning and clamping unit 201 is used to perform positioning and clamping on the table panel to be polished and to obtain the hole type information of the table panel; the robot drives the three-dimensional line scanning camera to move at a constant speed along the length of the table panel to collect the three-dimensional coordinate data of the inner contour of the hole to be polished on the table panel. Data parsing unit 202 is used to parse three-dimensional coordinate data, match the offset calculation rules of the corresponding hole position type, perform hole enlargement offset, chamfer height offset and arc segment normal direction offset calculation, superimpose tool wear compensation parameters, and generate actual grinding path points; The trajectory generation unit 203 is used to generate the robot's grinding trajectory based on the grinding path points, control the robot to perform grinding operations along the trajectory, and adjust the axial position of the tool in real time according to the height coordinates of each path point during the grinding process to ensure that the grinding depth is consistent at all points.

[0167] In some embodiments, the step of positioning and clamping the table panel to be polished and obtaining the hole type information of the table panel includes: performing a barcode scanning operation when the table panel to be polished flows into the feeding station to obtain the hole type, board size and order data of the table panel; adjusting the width of the lateral clamping mechanism according to the board size, performing lateral clamping and vertical pressing on the table panel, so that the conveying mechanism descends and disengages from the table panel to complete the positioning and fixing.

[0168] In some embodiments, the step of having a robot drive a three-dimensional line scanning camera to move at a constant speed along the length of the table panel to collect three-dimensional coordinate data of the inner contour of the hole to be polished on the table panel includes: the robot driving the three-dimensional line scanning camera integrated at the end to move at a constant speed along the length of the table panel, triggering the camera to collect images when it reaches a designated position, identifying and extracting the three-dimensional coordinate data of the straight edges and arc angles of each segment of the hole to be polished, and completing the scanning of the inner contour of the hole.

[0169] In some embodiments, parsing the three-dimensional coordinate data and matching the offset calculation rules for the corresponding hole type includes: receiving the three-dimensional coordinate data obtained by scanning, identifying the type and number of holes to be ground, classifying and storing the coordinate points of each contour segment according to the hole type, and calling the trajectory offset calculation rules for the corresponding type.

[0170] In some embodiments, the calculation of hole enlargement offset, chamfer height offset, and arc segment normal direction offset includes: expanding the original hole contour outward according to a preset single-sided hole enlargement amount to generate a tool center plane trajectory; setting a chamfer height offset to control the tool's pressing depth; and performing equal-point offset calculations on the arc segment contour along the normal direction to maintain the arc segment trajectory smooth and continuous.

[0171] In some embodiments, the step of superimposing tool wear compensation parameters to generate actual grinding path points includes: obtaining the height compensation value and diameter compensation value obtained by tool detection, superimposing the tool wear compensation parameters onto the offset contour trajectory, and generating the actual grinding path points corresponding to the grinding process.

[0172] In some embodiments, generating a robot grinding trajectory based on grinding path points includes: matching tool parameters and feed rate parameters of the corresponding mesh size according to the grinding process requirements corresponding to the hole position, and generating a robot grinding trajectory containing multiple grinding processes in combination with the grinding path points.

[0173] In some embodiments, the control robot to perform grinding operations along a trajectory includes: controlling the robot to change the grinding tool for the corresponding process, performing grinding operations along the inner contour of the hole according to the generated grinding trajectory, automatically switching tools after each process is completed, and sequentially completing all grinding processes.

[0174] In some embodiments, the axial position of the tool is adjusted in real time according to the height coordinates of each path point during the grinding process to ensure that the grinding depth is consistent throughout. This includes: reading the height coordinate value corresponding to each grinding path point during the grinding process, adjusting the axial position of the tool in real time, compensating for the bending deformation of the table panel, verifying the rationality of the height coordinates, intercepting abnormal values, and ensuring that the grinding depth is consistent throughout.

[0175] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the 3D vision scanning tabletop robot grinding trajectory automatic generation system and each module described above can be referred to the corresponding content in the various embodiments of the 3D vision scanning tabletop robot grinding trajectory automatic generation method, and will not be repeated here.

[0176] The aforementioned method for automatically generating the grinding trajectory of a 3D vision-scanning tabletop robot can be implemented as a computer program, which can, for example, Figure 3It runs on the system shown.

[0177] Please see Figure 4 , Figure 4 This is a schematic block diagram of the structure of a computer device provided in an embodiment of this application. The computer device includes a processor, a memory, and a network interface connected via a device bus, wherein the memory may include a storage medium and internal memory.

[0178] The storage medium can store operating devices and computer programs. The computer program includes program instructions that, when executed, cause the processor to perform any method for automatically generating the grinding trajectory of a 3D vision-scanning tabletop robot.

[0179] The processor provides computing and control capabilities, supporting the operation of the entire computer device.

[0180] The internal memory provides an environment for the execution of computer programs in non-volatile storage media. When the computer program is executed by the processor, it enables the processor to execute any method for automatically generating the grinding trajectory of a 3D vision scanning tabletop robot.

[0181] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the terminal to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0182] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0183] In one embodiment, the processor is configured to run a computer program stored in memory to perform the following steps: Positioning and clamping the table panel to be polished, and obtaining the hole type information of the table panel; the robot drives the 3D line scanning camera to move at a constant speed along the length of the table panel to collect the 3D coordinate data of the inner contour of the hole to be polished on the table panel. Analyze the three-dimensional coordinate data, match the offset calculation rules of the corresponding hole type, perform the calculation of hole enlargement offset, chamfer height offset and arc segment normal direction offset, superimpose tool wear compensation parameters, and generate the actual grinding path points; The robot's grinding trajectory is generated based on the grinding path points, and the robot is controlled to perform grinding operations along the trajectory. During the grinding process, the axial position of the tool is adjusted in real time according to the height coordinates of each path point to ensure that the grinding depth is consistent at all points.

[0184] In some embodiments, the step of positioning and clamping the table panel to be polished and obtaining the hole type information of the table panel includes: performing a barcode scanning operation when the table panel to be polished flows into the feeding station to obtain the hole type, board size and order data of the table panel; adjusting the width of the lateral clamping mechanism according to the board size, performing lateral clamping and vertical pressing on the table panel, so that the conveying mechanism descends and disengages from the table panel to complete the positioning and fixing.

[0185] In some embodiments, the step of having a robot drive a three-dimensional line scanning camera to move at a constant speed along the length of the table panel to collect three-dimensional coordinate data of the inner contour of the hole to be polished on the table panel includes: the robot driving the three-dimensional line scanning camera integrated at the end to move at a constant speed along the length of the table panel, triggering the camera to collect images when it reaches a designated position, identifying and extracting the three-dimensional coordinate data of the straight edges and arc angles of each segment of the hole to be polished, and completing the scanning of the inner contour of the hole.

[0186] In some embodiments, parsing the three-dimensional coordinate data and matching the offset calculation rules for the corresponding hole type includes: receiving the three-dimensional coordinate data obtained by scanning, identifying the type and number of holes to be ground, classifying and storing the coordinate points of each contour segment according to the hole type, and calling the trajectory offset calculation rules for the corresponding type.

[0187] In some embodiments, the calculation of hole enlargement offset, chamfer height offset, and arc segment normal direction offset includes: expanding the original hole contour outward according to a preset single-sided hole enlargement amount to generate a tool center plane trajectory; setting a chamfer height offset to control the tool's pressing depth; and performing equal-point offset calculations on the arc segment contour along the normal direction to maintain the arc segment trajectory smooth and continuous.

[0188] In some embodiments, the step of superimposing tool wear compensation parameters to generate actual grinding path points includes: obtaining the height compensation value and diameter compensation value obtained by tool detection, superimposing the tool wear compensation parameters onto the offset contour trajectory, and generating the actual grinding path points corresponding to the grinding process.

[0189] In some embodiments, generating a robot grinding trajectory based on grinding path points includes: matching tool parameters and feed rate parameters of the corresponding mesh size according to the grinding process requirements corresponding to the hole position, and generating a robot grinding trajectory containing multiple grinding processes in combination with the grinding path points.

[0190] In some embodiments, the control robot to perform grinding operations along a trajectory includes: controlling the robot to change the grinding tool for the corresponding process, performing grinding operations along the inner contour of the hole according to the generated grinding trajectory, automatically switching tools after each process is completed, and sequentially completing all grinding processes.

[0191] In some embodiments, the axial position of the tool is adjusted in real time according to the height coordinates of each path point during the grinding process to ensure that the grinding depth is consistent throughout. This includes: reading the height coordinate value corresponding to each grinding path point during the grinding process, adjusting the axial position of the tool in real time, compensating for the bending deformation of the table panel, verifying the rationality of the height coordinates, intercepting abnormal values, and ensuring that the grinding depth is consistent throughout.

[0192] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to implement the steps of the automatic generation method for the grinding trajectory of a tabletop robot using 3D visual scanning as provided in any embodiment of this application.

[0193] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0194] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for automatically generating the grinding trajectory of a tabletop robot using 3D vision scanning, characterized in that, include: Position and clamp the grinding table panel to obtain the hole type information of the table panel; The robot drives the 3D line scanning camera to move at a constant speed along the length of the table panel to collect the 3D coordinate data of the inner contour of the hole to be polished on the table panel. Analyze the three-dimensional coordinate data, match the offset calculation rules of the corresponding hole type, perform the calculation of hole enlargement offset, chamfer height offset and arc segment normal direction offset, superimpose tool wear compensation parameters, and generate the actual grinding path points; The robot's grinding trajectory is generated based on the grinding path points, and the robot is controlled to perform grinding operations along the trajectory. During the grinding process, the axial position of the tool is adjusted in real time according to the height coordinates of each path point to ensure that the grinding depth is consistent at all points.

2. The method according to claim 1, characterized in that, The process of positioning and clamping the table panel to be polished, and obtaining the hole type information of the table panel, includes: When the grinding table panel flows into the feeding station, a barcode scanning operation is performed to obtain the hole type, panel size and order data of the table panel; Adjust the width of the lateral clamping mechanism according to the size of the plate, perform lateral clamping and vertical pressing on the table panel, so that the conveying mechanism descends and disengages from the table panel to complete the positioning and fixing.

3. The method according to claim 1, characterized in that, The process involves a robot-driven 3D line scanning camera moving at a constant speed along the length of the worktable to collect 3D coordinate data of the inner contour of the hole to be ground on the worktable, including: The robot drives the integrated 3D line scanning camera at the end to move at a constant speed along the length of the table panel. When it reaches the designated position, it triggers the camera to acquire images, identify and extract the 3D coordinate data of each straight edge and arc angle of the hole to be polished, and complete the scanning of the inner contour of the hole.

4. The method according to claim 1, characterized in that, The analytical three-dimensional coordinate data is matched with the offset calculation rules corresponding to the hole position type, including: The system receives the 3D coordinate data obtained from the scan, identifies the type and number of holes to be ground, stores the coordinate points of each contour segment according to the hole type, and calls the trajectory offset calculation rules of the corresponding type.

5. The method according to claim 4, characterized in that, The calculation of hole enlargement offset, chamfer height offset, and arc segment normal direction offset includes: The original hole contour is expanded outward according to the preset single-sided hole expansion amount to generate the tool center plane trajectory; Set the chamfer height offset to control the depth of tool press-down; Perform equal-point offset calculations on the arc segment profile along the normal direction to maintain the smooth and continuous trajectory of the arc segment.

6. The method according to claim 5, characterized in that, The superimposed tool wear compensation parameters generate the actual grinding path points, including: The height compensation value and diameter compensation value obtained from tool detection are acquired, and the tool wear compensation parameters are superimposed on the offset contour trajectory to generate the actual grinding path points corresponding to the grinding process.

7. The method according to claim 1, characterized in that, The step of generating the robot grinding trajectory based on the grinding path points includes: Based on the grinding process requirements corresponding to the hole positions, the corresponding tool parameters and feed rate parameters are matched, and a robot grinding trajectory containing multiple grinding processes is generated by combining the grinding path points.

8. The method according to claim 1, characterized in that, The controlled robot performs a polishing operation along a trajectory, including: The robot is controlled to change the grinding tools for the corresponding process and perform grinding operations along the inner contour of the hole according to the generated grinding trajectory. The tool is automatically switched after each process is completed, and all grinding processes are completed in sequence.

9. The method according to claim 8, characterized in that, During the grinding process, the axial position of the tool is adjusted in real time based on the height coordinates of each path point to ensure consistent grinding depth throughout, including: During the grinding process, the height coordinate value corresponding to each grinding path point is read, the axial position of the tool is adjusted in real time, the bending deformation of the table panel is compensated, the rationality of the height coordinate is checked, abnormal values ​​are intercepted, and the grinding depth is ensured to be consistent in all places.

10. A 3D vision scanning-based automatic trajectory generation system for tabletop robot grinding, used to implement the method as described in any one of claims 1-9, characterized in that, include: The positioning and clamping unit is used to perform positioning and clamping on the table panel to be polished and to obtain the hole type information of the table panel; The robot drives the 3D line scanning camera to move at a constant speed along the length of the table panel to collect the 3D coordinate data of the inner contour of the hole to be polished on the table panel. The data parsing unit is used to parse three-dimensional coordinate data, match the offset calculation rules of the corresponding hole type, perform hole enlargement offset, chamfer height offset and arc segment normal direction offset calculation, superimpose tool wear compensation parameters, and generate actual grinding path points. The trajectory generation unit is used to generate the robot's grinding trajectory based on the grinding path points, control the robot to perform grinding operations along the trajectory, and adjust the axial position of the tool in real time according to the height coordinates of each path point during the grinding process to ensure that the grinding depth is consistent at all points.

Images