Robot flexible online labeling device and path planning method for curved surface workpiece

Abstract

The present application relates to a kind of robot flexible online label sticking device and path planning method for curved surface workpiece, including multi-axis robot arm, visual detection mechanism, label supply and peeling mechanism, workpiece conveying and positioning mechanism and control system.Multi-axis robot arm executes end installation with adaptive flexible label sticking head, it includes flexible adhesion layer, negative pressure adsorption cavity, deformation adjusting mechanism, and pressure perception and feedback mechanism being arranged between flexible adhesion layer and negative pressure adsorption cavity by outer to inner sequentially stacked arrangement.Label sticking before, deformation adjusting mechanism actively adjusts the initial curvature form of flexible adhesion layer, so that it is matched with the macroscopic curvature of the curved surface to be labeled, and passive adaptive deformation is generated by flexible adhesion layer along the curved surface shape when sticking, realize curved surface adaptive adhesion.Labeling path is automatically planned and generated by control system according to the workpiece curved surface model of three-dimensional reconstruction, without manual demonstration.

Description

Technical Field

[0001] This invention relates to the technical field of automated labeling equipment, and more specifically, to a robotic flexible online labeling device and path planning method for curved workpieces. Background Technology

[0002] Labeling equipment is an automated device that uses adhesives to attach paper, metal foil, or film labels to the surface of workpieces. It is widely used in packaging and labeling processes in industries such as food, pharmaceuticals, daily chemicals, automobile manufacturing, and electronics. As the manufacturing industry's requirements for product labeling continue to increase and product shapes become increasingly diverse, the technology of labeling equipment is also constantly evolving. From the initial manual feeding and foot pedal control, it has gradually developed into semi-automatic equipment controlled by motors and PLCs, and then into intelligent operating equipment with parameter programming and touch screen displays. In recent years, with the rapid development of machine vision, robotics, and artificial intelligence, labeling equipment is continuously progressing towards higher speeds, higher precision, and higher flexibility.

[0003] In existing technologies, typical automatic labeling equipment mainly includes core components such as a conveying system, a label supply system, a peeling mechanism, a labeling execution mechanism, and a control system. The conveying system typically uses a conveyor belt or conveyor chain to transport the workpiece to the labeling station; the label supply system includes components such as unwinding rollers, buffer rollers, and guide rollers to smoothly transport the roll of labels; the peeling mechanism uses a peeling plate to change the direction of the backing paper, thereby separating the label from the backing paper; the labeling execution mechanism affixes the peeled label to the surface of the workpiece, with roller-type and air-blowing types being the most common; the control system mostly uses a programmable logic controller (PLC) to achieve automated operation. Some high-end models have integrated a vision positioning system, which can automatically identify workpiece position deviations and adjust labeling parameters in real time.

[0004] However, existing labeling equipment still has the following significant shortcomings in practical applications: Firstly, existing fully automatic labeling machines are mainly suitable for relatively regular-shaped containers or flat workpieces such as cylinders, flat surfaces, and squares. Their labeling actuators are usually rigid labeling rollers or scrapers, with a fixed bonding surface shape. For workpieces with complex spatial curved surfaces, such as irregularly shaped shells with multiple curved transitions or automotive body panels, existing equipment struggles to achieve high-quality labeling. Although some existing solutions propose adaptive labeling heads for specific curved surfaces (such as arc surfaces), or use flexible pressure heads or floating attachment structures to allow for slight movement of the labeling contact surface, these solutions are mainly designed for single curvature or regular arc surfaces. Their flexible bonding layers can only passively deform with limited deformation range, lacking the ability to actively adjust the initial curvature of the bonding surface. Therefore, for complex free-form surfaces with varying curvature and irregular contours, rigid labeling heads cannot adapt to the surface shape, easily causing wrinkles, bubbles, or poor adhesion in the transition areas of the curved surface, making it difficult to guarantee labeling quality.

[0005] Secondly, the equipment lacks flexibility, and labeling path planning relies heavily on manual teaching, making it difficult to quickly switch between workpieces with different curved surfaces. Traditional labeling equipment is mostly dedicated equipment designed for specific product models. When dealing with workpieces of different shapes and sizes, it often requires extensive manual adjustments or even changes to tooling fixtures. The changeover and debugging process is highly dependent on the operator's experience and is time-consuming. With the introduction of industrial robots into the labeling field, some equipment has adopted robots as the main labeling performers, but the path planning of existing robot labeling systems still largely relies on manual teaching. The online teaching programming process is cumbersome, inefficient, and the accuracy depends entirely on the visual judgment of the teacher. For complex curved surface paths, online teaching programming is difficult to achieve satisfactory results. When the model of the workpiece to be labeled changes, even if there are only slight differences in the curved surface, a complete reprogramming process is required, which often takes several hours or even longer. This reliance on manual teaching severely restricts the response speed of the production line and the utilization rate of the equipment, making it difficult to meet the flexible production needs of modern manufacturing industries for multi-variety, small-batch, and rapid-response production.

[0006] Therefore, there is an urgent need to develop a flexible online labeling device and corresponding method that can adapt to workpieces with arbitrarily complex curved surfaces, possess active surface self-adaptation capabilities, and intelligent path planning functions. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a robot flexible online labeling device and path planning method for curved workpieces, in view of the above-mentioned defects of the prior art.

[0008] The technical solution adopted by this invention to solve its technical problem is as follows: On one hand, this invention provides a robotic flexible online labeling device for curved workpieces, comprising a multi-axis robotic arm, a vision inspection mechanism, a label supply and peeling mechanism, a workpiece conveying and positioning mechanism, and a control system; the vision inspection mechanism is disposed above the workpiece conveying and positioning mechanism, and is used to collect three-dimensional point cloud data and texture information of the workpiece surface before labeling, and to collect position images of the workpiece and label in real time during the labeling process; the label supply and peeling mechanism is disposed on one side of the multi-axis robotic arm, and is used to supply roll labels and peel the labels off the backing paper; the workpiece conveying and positioning mechanism is disposed in front of the multi-axis robotic arm, and is used to convey the workpiece to be labeled to the labeling station and position the workpiece; The multi-axis robotic arm is equipped with an adaptive flexible labeling head at its end effector for labeling curved workpieces. The adaptive flexible labeling head includes a flexible bonding layer, a negative pressure adsorption cavity, and a deformation adjustment mechanism stacked sequentially from the outside in, as well as a pressure sensing and feedback mechanism disposed between the flexible bonding layer and the negative pressure adsorption cavity. The outer surface of the flexible bonding layer serves as the bonding working surface. The negative pressure adsorption cavity is fitted to the inner side of the flexible bonding layer and provides negative pressure adsorption force. The deformation adjustment mechanism is disposed inside the negative pressure adsorption cavity and connected to the inner wall of the flexible bonding layer via multiple push rods passing through the negative pressure adsorption cavity, for adjusting the initial curvature of the flexible bonding layer. The multi-axis robotic arm, the adaptive flexible labeling head, the vision inspection mechanism, the label supply and peeling mechanism, and the workpiece conveying and positioning mechanism are all communicatively connected to the control system, which is used to receive data and control the coordinated operation of each mechanism.

[0009] The robotic flexible online labeling device of the present invention includes a flexible bonding layer that is a deformable sheet structure. In its natural state, the flexible bonding layer is a convex bowl-shaped arc and can undergo adhesive deformation with the surface undulations of the curved workpiece when subjected to external extrusion. Pressure sensors are embedded in the inner wall of the flexible bonding layer in a grid or concentric ring pattern to sense the contact pressure of each area of ​​the bonding surface.

[0010] The robotic flexible online labeling device of the present invention comprises a negative pressure adsorption chamber, which is a hollow shell concentrically arranged with the flexible bonding layer. The side of the chamber facing the flexible bonding layer is open, and the outline and size of this openness are consistent with the outline of the inner wall of the flexible bonding layer. The edge of the inner wall of the flexible bonding layer is fixed to the edge of the openness, making the flexible bonding layer the bottom closed wall of the negative pressure adsorption chamber. A vacuum pipe interface is provided on the side wall of the negative pressure adsorption chamber, which is connected to an external vacuum generator via an air pipe. The bonding surface of the flexible bonding layer is densely covered with adsorption holes penetrating the wall thickness of the flexible bonding layer, used to adsorb and fix the label onto the bonding surface by negative pressure after the vacuum generator is activated.

[0011] The robotic flexible online labeling device of the present invention has the following features: the diameter of the adsorption holes is 0.3mm to 0.8mm; the total area of ​​the adsorption holes on the bonding surface accounts for 5% to 15% of the total area of ​​the flexible bonding layer; and all adsorption holes are directly connected to the internal space of the negative pressure adsorption chamber.

[0012] The robotic flexible online labeling device of the present invention includes a deformation adjustment mechanism comprising a support plate and a plurality of electric push rods arranged in an array. The support plate is arranged parallel above the negative pressure adsorption cavity, and each electric push rod is fixedly installed on the side of the support plate facing the negative pressure adsorption cavity. A through hole is respectively opened at the top of the negative pressure adsorption cavity at a position corresponding to the output end of each electric push rod, and a guide sleeve with a sealing ring is installed in the through hole. The front end of each electric push rod passes through the corresponding guide sleeve and extends into the interior of the negative pressure adsorption cavity. The front end of each electric push rod is connected to a corresponding position on the inner wall surface of the flexible bonding layer through a freely rotatable connector. Each electric push rod is independently and controllably extended or retracted to push different areas of the flexible bonding layer along the normal direction of its arc surface before labeling, so as to adjust the initial arc shape of the flexible bonding layer to a shape close to the curvature of the surface to be labeled.

[0013] The robotic flexible online labeling device of the present invention includes a pressure sensing and feedback mechanism comprising a flexible circuit board, a signal processing chip set, and a metal shield. The flexible circuit board is bonded and fixed to the inner wall of the flexible bonding layer, and the pins of each pressure sensor are connected to corresponding solder points on the flexible circuit board via conductive lines. The signal processing chip set is soldered to the edge of the flexible circuit board, and each signal processing channel is connected to a corresponding sensor to convert the resistance signals collected by each sensor into voltage signals. The metal shield is placed over the signal processing chip set and electrically connected to the ground plane of the flexible circuit board. The output terminal of the flexible circuit board is connected to an analog-to-digital converter via a signal line, and the analog-to-digital converter converts the voltage signals of each channel into digital pressure values ​​and transmits them to the control system.

[0014] The robotic flexible online labeling device of the present invention includes a vision inspection mechanism comprising a gantry frame, a laser contour scanner, an industrial camera, and a grating projector. The laser contour scanner, the industrial camera, and the grating projector are arranged sequentially along the length of the crossbeam of the gantry frame, with their lenses all facing downwards. The laser contour scanner has a built-in laser emitter and a photosensitive receiver, used to emit a linear laser line onto the workpiece surface on the workpiece conveying and positioning mechanism and receive reflected light to obtain contour point cloud data of the workpiece surface. The grating projector projects an coded grating stripe pattern onto the workpiece surface to assist the laser contour scanner in obtaining depth information of the workpiece surface. A filter is installed at the front end of the lens of the industrial camera to filter out ambient light interference during the labeling process and acquire images of the workpiece surface in real time.

[0015] The robot flexible online labeling device of the present invention includes a control system comprising a main control module, a vision data processing module, and a bus communication module. The main control module includes an industrial control motherboard, a central processing unit, and memory. The main control module is electrically connected to the visual data processing module and the bus communication module, respectively, and is used to perform calculation tasks such as 3D point cloud reconstruction, surface model generation, and labeling path planning, and to issue control commands to each module. The visual data processing module includes an image acquisition card and a graphics processor. The input end of the image acquisition card is electrically connected to the image output end of the visual inspection mechanism, and the graphics processor is electrically connected to the output end of the image acquisition card. It is used to perform parallel accelerated processing on the acquired 3D point cloud data and image data, and transmit the processed data to the main control module. The bus communication module includes a bus controller and a bus transceiver. The bus controller is electrically connected to the main control module, and the bus transceiver is electrically connected to the bus controller and connected to the fieldbus network to realize data interaction and clock synchronization between the main control module and external devices.

[0016] The robotic flexible online labeling device of the present invention further includes a motion control module and a logic control module in the control system. The motion control module includes a servo control card and a pulse output interface. The servo control module is electrically connected to the servo drivers of each joint of the multi-axis robotic arm through the pulse output interface. It is used to receive path planning data sent by the main control module through the bus communication module and generate servo drive pulse signals for each joint to drive the robot to perform labeling motion. The logic control module includes a programmable logic device and an input / output expansion interface. The programmable logic device is electrically connected to the drive motor of the workpiece conveying and positioning mechanism, the vacuum pneumatic solenoid valve of the adaptive flexible labeling head, and each drive motor of the label supply and peeling mechanism through the input / output expansion interface, and is used to control the timing logic actions of workpiece conveying, workpiece positioning, label supply and label peeling.

[0017] On the other hand, the present invention also provides a robot flexible online labeling path planning method for curved workpieces, applied to a robot flexible online labeling device for curved workpieces as described above, wherein the method includes the following steps: Step 1: Scan the workpiece using the vision inspection mechanism to obtain the three-dimensional point cloud data and texture information of the workpiece surface, reconstruct the three-dimensional digital surface model of the workpiece, and extract the boundary features and curvature distribution information of the area to be labeled; generate workpiece pose information based on the scan data and feed it back to the workpiece conveying and positioning mechanism to position the workpiece. Step 2: Control the multi-axis robotic arm to move to the label supply and peeling mechanism, and pick up the peeled label through the adaptive flexible labeling head; Step 3: Based on the curvature distribution information of the area to be labeled extracted in Step 1, before the adaptive flexible labeling head contacts the workpiece, control the deformation adjustment mechanism to drive the flexible bonding layer to deform, so that the initial curvature shape of the flexible bonding layer matches the macroscopic curvature of the surface to be labeled. Step 4: Using the reference point of the picked-up label as a reference, plan the labeling motion path of the multi-axis robotic arm on the three-dimensional digital surface model, generate a path point sequence containing position coordinates and attitude angles, and perform interpolation fitting on the path point sequence to form a smooth and continuous robot motion trajectory. Step 5: During the labeling motion of the multi-axis robotic arm, the position and posture of the robotic arm end effector are corrected in real time by the vision detection mechanism; at the same time, the contact pressure data of each area of ​​the flexible bonding layer are acquired in real time by the pressure sensing and feedback mechanism, and the motion parameters of the multi-axis robotic arm are dynamically adjusted according to the contact pressure data. Step 6: After labeling is completed, the labeling quality is checked by the vision inspection mechanism. If the inspection fails, the current workpiece is marked as a defective product.

[0018] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention comprises an adaptive flexible labeling head consisting of a flexible bonding layer, a negative pressure adsorption cavity, and a deformation adjustment mechanism stacked sequentially from the outside in. The flexible bonding layer adopts a deformable bowl-shaped arc-shaped sheet structure, which can passively adapt to the shape of the curved surface when in contact with the workpiece. The deformation adjustment mechanism has multiple electric push rods arranged in an array and independently controlled. Before labeling, it can actively push different areas of the flexible bonding layer to shift according to the curvature distribution information of the area to be labeled, adjusting its initial arc shape to be close to the curvature shape of the surface to be labeled. Through the dual surface adaptive mechanism of "active pre-deformation + passive adaptive deformation", the bonding surface can achieve tight adhesion with any complex free-form surface, fundamentally solving the problems of wrinkles, bubbles, and poor adhesion in the transition area of ​​curved surfaces caused by existing rigid labeling heads or single flexible labeling heads, and significantly improving the labeling quality of curved surfaces.

[0019] 2. This invention acquires 3D point cloud data and texture information of the workpiece surface through a vision inspection mechanism. The control system automatically reconstructs the 3D digital surface model of the workpiece and extracts the boundary features and curvature distribution information of the area to be labeled. It then automatically generates a path point sequence containing position coordinates and attitude angles, and interpolates and fits the path point sequence to form a smooth and continuous robot motion trajectory. The entire process requires no manual teaching. When the workpiece model changes, the system only needs to select the corresponding workpiece model in the control system or rescan the workpiece; the system can then automatically complete the path planning. The changeover time is reduced from several hours in existing technologies to several minutes, significantly improving the flexibility of the equipment and the rapid response capability of the production line.

[0020] 3. This invention forms a complete closed-loop control chain through the coordinated operation of a vision inspection mechanism, a pressure sensing and feedback mechanism, and a control system. Before labeling, the vision inspection mechanism collects 3D point cloud data for path planning; during labeling, the vision inspection mechanism acquires real-time images of the workpiece surface and corrects the end-effector pose, while the pressure sensors of the pressure sensing and feedback mechanism monitor the contact pressure distribution in various areas of the flexible bonding layer in real time, and the control system dynamically adjusts the motion parameters of the robotic arm based on the pressure data; after labeling, the vision inspection mechanism inspects the labeling quality. Through closed-loop control throughout the entire process, the labeling accuracy can reach ±0.2mm, far superior to the ±0.5mm of traditional equipment, ensuring the consistency of labeling quality. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be further described below in conjunction with the accompanying drawings and embodiments. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort: Figure 1 This is a schematic diagram of the structure of a robotic flexible online labeling device for curved workpieces according to Embodiment 1 of the present invention.

[0022] Figure 2 yes Figure 1 A schematic diagram of the structure of the adaptive flexible labeling head 6.

[0023] Figure 3 yes Figure 2 A schematic diagram of the inner surface of the flexible bonding layer 61.

[0024] Figure 4 yes Figure 3 A schematic diagram of the pressure sensing and feedback mechanism 64.

[0025] Figure 5 yes Figure 1 A schematic diagram of the structure of the visual inspection mechanism 2.

[0026] Figure 6 yes Figure 1 The circuit principle block diagram of the control system 5. Detailed Implementation

[0027] The terms "first," "second," "third," and "fourth," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0028] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0029] "Multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0030] Furthermore, the terms indicating orientation, such as "up, down, front, back, left, right, upper end, lower end, longitudinal," etc., are all based on the posture and position of the device or equipment described in this solution during normal use.

[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, a clear and complete description will be provided below in conjunction with the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the protection scope of the present invention.

[0032] Example 1: like Figure 1-6 As shown, the present invention provides a robotic flexible online labeling device for curved workpieces, including a multi-axis robotic arm 1, a vision inspection mechanism 2, a label supply and peeling mechanism 3, a workpiece conveying and positioning mechanism 4, and a control system 5.

[0033] The visual inspection mechanism 2 is positioned above the workpiece conveying and positioning mechanism 4, and is used to collect three-dimensional point cloud data and texture information of the workpiece surface before labeling, and to collect position images of the workpiece and label in real time during the labeling process. The label supply and peeling mechanism 3 is positioned on one side of the multi-axis robotic arm 1, and is used to supply roll labels and peel the labels off the backing paper. The workpiece conveying and positioning mechanism 4 is positioned in front of the multi-axis robotic arm 1, and is used to transport the workpiece to be labeled to the labeling station and position the workpiece.

[0034] The multi-axis robotic arm 1 is equipped with an adaptive flexible labeling head 6 at its end effector, used for labeling curved workpieces. Figure 2-4 As shown, the adaptive flexible labeling head 6 includes a flexible bonding layer 61, a negative pressure adsorption cavity 62, and a deformation adjustment mechanism 63, stacked sequentially from the outside to the inside, as well as a pressure sensing and feedback mechanism 64 disposed between the flexible bonding layer 61 and the negative pressure adsorption cavity 62. The outer surface of the flexible bonding layer 61 is the bonding working surface, and the negative pressure adsorption cavity 62 is bonded to the inner side of the flexible bonding layer 61 and provides negative pressure adsorption force to the flexible bonding layer 61. The deformation adjustment mechanism 63 is disposed on the inner side of the negative pressure adsorption cavity 62 and is connected to the inner wall of the flexible bonding layer 61 through multiple push rods 65 passing through the negative pressure adsorption cavity 62, for adjusting the initial curvature shape of the flexible bonding layer 61.

[0035] The multi-axis robotic arm 1, the adaptive flexible labeling head 6, the vision inspection mechanism 2, the label supply and peeling mechanism 3, and the workpiece conveying and positioning mechanism 4 are all communicatively connected to the control system 5, which is used to receive data and control the coordinated operation of each mechanism.

[0036] In this embodiment, the flexible bonding layer 61 is a deformable sheet structure. In its natural state, the flexible bonding layer 61 has an outwardly convex, bowl-shaped arc and can conform to the surface irregularities of the curved workpiece when subjected to external pressure. Pressure sensors 01 are embedded in a grid or concentric ring pattern on the inner wall of the flexible bonding layer 61 to sense the contact pressure in different areas of the bonding surface. As an example, the pressure sensor can be an array-type flexible pressure sensor, such as the MF3216 array-type pressure sensor from Hanwei Technology Group. This sensor uses flexible pressure sensing technology to print nanomaterials on a flexible, thin material, achieving high-sensitivity detection of pressure distribution. Alternatively, a 4×8 or 6×6 flexible array tactile pressure sensor can be used, integrating multiple independent detection units within the sensing area to capture the pressure distribution and subtle changes at the contact point in real time. All of the above pressure sensors are mature flexible sensing elements in the field.

[0037] This invention embeds pressure sensors in a grid or concentric ring pattern on the inner wall of a flexible bonding layer. The sensing surface of each sensor is located at the center of the flexible bonding layer's thickness, enabling the sensing of contact pressure distribution across different areas of the bonding surface. Combined with a flexible circuit board, signal processing chipset, and metal shielding for pressure sensing and feedback, the resistance signals collected by each sensor are converted into voltage signals and transmitted to the control system after analog-to-digital conversion. This achieves high-precision, high-reliability real-time monitoring of contact pressure during the bonding process, providing a data foundation for precise control of labeling pressure and effectively avoiding labeling defects caused by uneven pressure.

[0038] In this embodiment, the negative pressure adsorption cavity 62 is a hollow shell concentrically arranged with the flexible adhesive layer 61. The side facing the flexible adhesive layer 61 is open, and the outline and size of this opening match the outline of the inner wall of the flexible adhesive layer 61. The edge of the inner wall of the flexible adhesive layer 61 is fixed to the edge of the opening, making the flexible adhesive layer 61 the bottom closed wall of the negative pressure adsorption cavity 62. A vacuum pipe interface 66 is provided on the side wall of the negative pressure adsorption cavity 62, which is connected to an external vacuum generator via an air pipe. The adhesive surface of the flexible adhesive layer 61 is densely covered with adsorption holes 67 penetrating the wall thickness of the flexible adhesive layer 61, used to adsorb and fix the label onto the adhesive surface through negative pressure after the vacuum generator is activated. As an example, the vacuum generator can be an Airtac XKCV series vacuum generator, which utilizes compressed air through a special nozzle to generate a Venturi effect, creating a local negative pressure within the cavity. It features a compact structure, fast response, and is oil-free and pollution-free, and is widely used in label adsorption in automated labeling equipment. The aforementioned vacuum generator is a very mature pneumatic component in this field.

[0039] The above design enables the vacuum generator to produce a uniform negative pressure adsorption force at the adsorption hole after it is started, which can firmly adsorb the label onto the bonding surface of the flexible bonding layer, ensuring that the label does not shift or fall off during label picking and movement; the negative pressure is gradually released during the labeling process, which can also make the label smoothly transferred to the curved surface of the workpiece, thus realizing a reliable connection between label picking and labeling.

[0040] In this embodiment, the diameter of the adsorption holes 67 is 0.3mm to 0.8mm, the total area of ​​the adsorption holes 67 on the bonding surface accounts for 6% to 16% of the total area of ​​the flexible bonding layer 61, and all the adsorption holes 67 are directly connected to the internal space of the negative pressure adsorption cavity 62.

[0041] In this embodiment, the deformation adjustment mechanism 63 includes a support plate 631 and a plurality of electric push rods 65 arranged in an array. The support plate 631 is arranged parallel above the negative pressure adsorption cavity 62, and each electric push rod 65 is fixedly installed on the side of the support plate 631 facing the negative pressure adsorption cavity 62. A through hole 632 is opened at the top of the negative pressure adsorption cavity 62 at a position corresponding to the output end of each electric push rod 65, and a guide sleeve 633 with a sealing ring is installed in the through hole 632. The front end of each push rod 65 passes through the corresponding guide sleeve 633 and extends into the interior of the negative pressure adsorption cavity 62. The front end of each electric push rod 65 is connected to a corresponding position on the inner wall surface of the flexible bonding layer 61 through a freely rotatable connector 634. Each of the electric actuators 65 is independently and controllably extended or retracted to push different areas of the flexible adhesive layer 61 along the normal direction of its arc surface before labeling, thereby adjusting the initial arc shape of the flexible adhesive layer 61 to a shape close to the curvature of the surface to be labeled. As an example, the electric actuator 65 can be an RMMPLA miniature platform actuator, which adopts a platform structure integrating a servo motor and a high-precision ball bearing guide; or it can be an RMNPLAHF precision force-controlled miniature electric cylinder, possessing high force control accuracy of ±0.01N and micron-level high repeatability. Both of these electric actuators are mature miniature linear actuators in the field.

[0042] In this embodiment, as Figure 4 As shown, the pressure sensing and feedback mechanism 64 includes a flexible circuit board 641, a signal processing chip set 642, and a metal shield 643. The flexible circuit board 641 is attached and fixed to the inner wall of the flexible bonding layer 61, and the pins of each pressure sensor are connected to corresponding solder points on the flexible circuit board 641 via conductive lines. The signal processing chip set 642 is soldered to the edge of the flexible circuit board 641, and each signal processing channel is connected to the corresponding sensor to convert the resistance signals collected by each sensor into voltage signals. The metal shield 643 covers the outside of the signal processing chip set 642 and is electrically connected to the ground plane of the flexible circuit board 641. The output terminal of the flexible circuit board 641 is connected to an analog-to-digital converter (ADC) via a signal line. The ADC converts the voltage signals of each channel into digital pressure values ​​and transmits them to the control system 5. As an example, the ADC can be an AD7606 series 8-channel synchronous sampling ADC from Analog Devices (ADI) or an ADS8588H series 16-bit synchronous sampling ADC from Texas Instruments (TI). Both of these chips are mature data acquisition devices in the field.

[0043] In this embodiment, as Figure 5As shown, the visual inspection mechanism 2 includes a gantry 21, a laser contour scanner 22, an industrial camera 23, and a grating projector 24. The laser contour scanner 22, the industrial camera 23, and the grating projector 24 are arranged sequentially along the length of the crossbeam of the gantry 21, with all three lenses facing downwards. The laser contour scanner 22 has a built-in laser emitter and photosensitive receiver, used to emit a linear laser line onto the surface of the workpiece on the workpiece conveying and positioning mechanism 4 and receive the reflected light to obtain contour point cloud data of the workpiece surface. The grating projector 24 is used to project an coded grating stripe pattern onto the workpiece surface to assist the laser contour scanner 22 in obtaining depth information of the workpiece surface. A filter is installed at the front end of the lens of the industrial camera 23 to filter out ambient light interference during the labeling process and to acquire images of the workpiece surface in real time. As an example, the laser profile scanner 22 can employ the Gocator 2500 series high-speed 3D line laser profile sensor from LMI, which excels in speed and compactness; alternatively, it can use the Gocator 2629 4K+ resolution line laser profile sensor, which offers high scanning speed and high-quality data output capabilities. The industrial camera 23 can be a CCD or CMOS industrial camera from brands such as Hikvision Robotics or Daheng Imaging, which converts light signals into electrical signals via a photosensitive element, and then generates digital images through analog-to-digital conversion. Both the laser profile scanner and the industrial camera are mature visual inspection components in this field. The laser profile scanner 22, industrial camera 23, and grating projector 24 work together to stably acquire high-precision 3D point cloud data and image data even in complex lighting environments, providing a reliable sensing foundation for 3D surface reconstruction and online pose correction.

[0044] In this embodiment, the label feeding and peeling mechanism 3 is a mature existing technology in the field, specifically including a frame, a feeding mechanism, a traction mechanism, a peeling mechanism, and a take-up mechanism. The feeding mechanism includes a feeding shaft rotatably mounted on the frame and a baffle plate fitted on the feeding shaft. The roll of labels is mounted on the feeding shaft and limited by the baffle plates on both sides. The traction mechanism includes an active traction roller and a driven traction roller driven by a stepper motor or a servo motor, which are positioned opposite each other to clamp and pull the label tape forward for step-by-step conveying. The peeling mechanism includes a label-splitting knife fixedly mounted on the frame. The front end of the label-splitting knife has a sharp-angled cutting edge. When the label tape passes around the sharp-angled cutting edge of the label-splitting knife, it bends at an acute angle, separating the label from the backing paper. The take-up mechanism includes a take-up shaft rotatably mounted on the frame for winding up the separated backing paper. To ensure that the label tape remains flat during conveying, a guiding mechanism composed of multiple guide rollers and a flattening mechanism composed of upper and lower pressure rollers are also provided between the feeding mechanism and the traction mechanism. Furthermore, the label-receiving platform is located at the outlet of the peeling mechanism. The upper surface of this platform has vacuum suction holes for receiving the peeled labels and using negative pressure to adhere them to the platform, keeping them flat for retrieval. The components and working principles of the aforementioned label supply and peeling mechanism are common knowledge in the field, and their specific principles will not be elaborated further.

[0045] In this embodiment, the workpiece conveying and positioning mechanism 4 is also a mature existing technology in the field, specifically including a frame, a double-row parallel transmission belt module, a width adjustment mechanism, and a positioning and locking module. The double-row parallel transmission belt module includes two parallel annular synchronous belts, which are tensioned and driven by a driving pulley and a driven pulley, respectively. A conveying plane for carrying the workpiece is formed between the two synchronous belts, used to convey the workpiece to be labeled to the labeling station in a straight line. The width adjustment mechanism includes a bidirectional lead screw rotatably mounted on the frame and sliding seats respectively disposed on the outer sides of the two synchronous belts. The two ends of the bidirectional lead screw have positive and negative threads, respectively. The two sliding seats are threadedly engaged with the two ends of the bidirectional lead screw. By rotating the bidirectional lead screw, the distance between the two synchronous belts can be adjusted synchronously to accommodate workpieces of different widths. The positioning and locking module is located between two synchronous belts and directly below the labeling station. It includes a vertically mounted lifting cylinder and a support plate fixed to the top of the piston rod of the lifting cylinder. The upper surface of the support plate has a positioning groove or positioning pin that matches the bottom contour of the workpiece. When the photoelectric sensor detects that the workpiece has been conveyed to the labeling station, the lifting cylinder pushes the support plate upward, lifting the workpiece off the synchronous belt and locking it in a predetermined position. After the labeling operation is completed, the lifting cylinder drives the support plate downward, placing the workpiece back onto the synchronous belt and continuing to convey it forward. In addition, the drive pulley of the double-row parallel transmission belt module is driven by a servo motor, which is equipped with an encoder for real-time feedback of conveying speed and position. The components and working principles of the above-mentioned workpiece conveying and positioning mechanism 4 are common knowledge in the field, and the specific principles will not be described in detail.

[0046] In this embodiment, as Figure 6 As shown, the control system 5 includes a main control module 51, a vision data processing module 52, and a bus communication module 53.

[0047] The main control module 51 includes an industrial control motherboard, a central processing unit (CPU), and memory. The main control module 51 is electrically connected to the visual data processing module 52 and the bus communication module 53, respectively, and is used to perform computational tasks such as 3D point cloud reconstruction, surface model generation, and labeling path planning, and to issue control commands to each module. As an example, the industrial control motherboard can be a Huabei Industrial Control EMB3948 4-inch embedded industrial motherboard based on an Intel Celeron J1900 processor; alternatively, it can be a Korad EPC302A wide-temperature industrial computer equipped with a 12th generation Intel Core low-power embedded CPU. Both of these industrial control motherboards are mature industrial computing platforms in this field.

[0048] The visual data processing module 52 includes an image acquisition card and a graphics processor. The input terminal of the image acquisition card is electrically connected to the image output terminal of the visual inspection mechanism 2, and the graphics processor is electrically connected to the output terminal of the image acquisition card. It is used to perform parallel accelerated processing on the acquired 3D point cloud data and image data, and transmit the processed data to the main control module 51. As an example, the graphics processor can be an NVIDIA Jetson AGXOrin embedded AI computing module or an Intel Arc series discrete graphics card; both of these graphics processors are mature parallel computing devices in the field.

[0049] The bus communication module 53 includes a bus controller and a bus transceiver. The bus controller is electrically connected to the main control module 51, and the bus transceiver is electrically connected to the bus controller and connected to a fieldbus network. This allows for data interaction and clock synchronization between the main control module 51 and external devices. As an example, the fieldbus network can be an EtherCAT industrial Ethernet bus, and the bus controller can be a Beckhoff ET1200 or ET1810 EtherCAT slave controller chip. These bus solutions represent mature high-speed real-time communication technologies in the field.

[0050] In this embodiment, the control system 5 further includes a motion control module 54 and a logic control module 55.

[0051] The motion control module 54 includes a servo control card and a pulse output interface. The servo control card is electrically connected to the servo drivers of each joint of the multi-axis robotic arm 1 through the pulse output interface. It receives path planning data from the main control module 51 via the bus communication module 53 and generates servo drive pulse signals for each joint to drive the robot to perform labeling motion. As an example, the servo driver can be the CoolDrive RC6 series integrated network servo driver from Qingneng Dechuang, a product tailored for industrial robots, featuring a multi-axis integrated design and built-in vibration suppression algorithms and feedforward functions; alternatively, it can be the JD series AC servo driver from Boke or a low-voltage servo driver from Mingzhi Electric. All of the above servo drivers are mature robot drive components in this field.

[0052] The logic control module 55 includes a programmable logic device (PLD) and an input / output expansion interface. The PLD is electrically connected via the PLD to the drive motors of the workpiece conveying and positioning mechanism 4, the vacuum pneumatic solenoid valve of the adaptive flexible labeling head 6, and each drive motor of the label supply and peeling mechanism 3, respectively, to control the timing logic actions of workpiece conveying, workpiece positioning, label supply, and label peeling. As an example, the PLD can be a Ruineng XM series PLC, which has high-speed I / O input capability to ensure rapid response of the photoelectric sensor; alternatively, it can be an Yike LogicX0 series bus-based small programmable controller or a Mitsubishi Electric MELSEC series PLC. All of the above-mentioned programmable logic controllers are mature industrial control devices in this field.

[0053] The modules described above have clearly defined roles and work together efficiently. The visual data processing module uses a graphics processor to perform parallel acceleration processing on the acquired 3D point cloud data and image data, significantly improving data processing speed; the main control module performs calculations for 3D point cloud reconstruction, surface model generation, and labeling path planning; the motion control module and logic control module are responsible for the robotic arm's motion control and timing logic control, respectively. This modular architecture results in a fast system response, high reliability, and facilitates future functional expansion, upgrades, and maintenance.

[0054] Example 2: This invention also provides a robot flexible online labeling path planning method for curved workpieces, applied to the robot flexible online labeling device for curved workpieces as described in Embodiment 1 above, comprising the following steps: Step 1: The visual inspection mechanism performs a full-range scan of the workpiece to be labeled, acquiring 3D point cloud data and texture information of the workpiece surface. The laser contour scanner emits a linear laser line onto the workpiece surface and receives the reflected light to acquire contour point cloud data of the workpiece surface; the grating projector projects an coded grating stripe pattern onto the workpiece surface to assist in acquiring depth information of the workpiece surface. The visual data processing module performs parallel accelerated processing on the acquired point cloud data and image data and transmits it to the main control module, which reconstructs the 3D digital surface model of the workpiece and extracts the boundary features and curvature distribution information of the area to be labeled. Simultaneously, the in-situ pose information of the workpiece is generated based on the scan data and fed back to the workpiece conveying and positioning mechanism, which is then driven by the logic control module to precisely position the workpiece.

[0055] Step 2: The main control module sends commands to the motion control module via the bus communication module. The motion control module drives the multi-axis robotic arm 1 to move to the label supply and peeling mechanism. The logic control module controls the label supply and peeling mechanism to complete the label peeling operation. The peeled label is then received on the adsorption platform. The adaptive flexible labeling head uses the negative pressure generated by the negative pressure adsorption chamber to create an adsorption force at the adsorption holes, picking up the peeled label from the adsorption platform and fixing it to the outer surface of the flexible adhesive layer.

[0056] Step 3: The main control module calculates the target curvature compensation amount required for each region of the flexible bonding layer based on the curvature distribution information of the area to be labeled extracted in Step 1. Before the adaptive flexible labeling head contacts the workpiece, the main control module controls the independent movement of each electric push rod of the deformation adjustment mechanism through the logic control module. Each electric push rod extends or retracts by different lengths according to the calculated target curvature compensation amount, pushing different regions of the flexible bonding layer to move along the normal direction of its arc surface, so that the initial arc surface shape of the flexible bonding layer is deformed until its initial curvature shape matches the macroscopic curvature of the surface to be labeled.

[0057] Step 4: The main control module uses the reference point of the picked label (such as the geometric center of the label) as a reference to plan the label-applying motion path of the multi-axis robotic arm on the three-dimensional digital surface model. The path planning includes: Within the labeling area of ​​the 3D digital surface model, a series of path points are generated based on the surface curvature distribution. Each path point contains the position coordinates and attitude angles (roll, pitch, yaw) of the multi-axis robotic arm 1 in Cartesian space. The attitude angles are determined based on the normal vector direction of the surface at that point to ensure that the angle between the bonding surface of the adaptive flexible labeling head and the surface meets the labeling requirements. The main control module uses B-spline curves or NURBS curves to interpolate and fit the discrete path points, forming a smooth and continuous robot motion trajectory. Simultaneously, collision detection and obstacle avoidance planning are performed in conjunction with the spatial position of the workpiece conveying and positioning mechanism. The final generated path point sequence is sent to the motion control module via the bus communication module.

[0058] Step 5: The motion control module drives the multi-axis robotic arm to perform the labeling motion based on the received path point sequence. During the labeling motion performed by the multi-axis robotic arm and the contact of the flexible bonding layer with the curved workpiece: On one hand, the industrial camera of the vision inspection mechanism acquires images of feature points or marker points on the surface of the workpiece in real time. The vision data processing module processes the acquired images and transmits them to the main control module. The main control module calculates the pose deviation of the end effector of the multi-axis robotic arm through a vision servo control algorithm and corrects the motion trajectory of the robotic arm in real time through the motion control module.

[0059] On the other hand, the pressure sensors on the flexible circuit board of the pressure sensing and feedback mechanism sense the contact pressure of each area of ​​the flexible bonding layer in real time. The signal processing chip converts the resistance signals collected by each sensor into voltage signals, which are then converted into digital pressure values ​​by an analog-to-digital converter and transmitted to the main control module. Based on the contact pressure distribution data, the main control module dynamically adjusts the feed speed and end effector posture of the multi-axis robotic arm, and can also fine-tune the output force of each electric push rod of the deformation adjustment mechanism in real time, so that the label is uniformly attached to the curved surface from the center to the edge, ensuring that the contact pressure between the label and each area of ​​the curved surface is consistent.

[0060] Step Six: After labeling is completed, the multi-axis robotic arm moves away, and the industrial camera of the vision inspection mechanism captures images of the labeling area. The main control module uses image processing algorithms to detect the labeling quality, including label position deviation, presence of air bubbles, and presence of wrinkles. If the detection result is qualified, the workpiece conveying and positioning mechanism conveys the workpiece to the next station; if the detection is unqualified, the main control module marks the current workpiece as a defective product and triggers an alarm or controls the rejection mechanism to divert defective products through the logic control module.

[0061] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A robotic flexible online labeling device for curved workpieces, characterized in that, The system includes a multi-axis robotic arm, a vision inspection mechanism, a label supply and peeling mechanism, a workpiece conveying and positioning mechanism, and a control system. The vision inspection mechanism is located above the workpiece conveying and positioning mechanism and is used to collect three-dimensional point cloud data and texture information of the workpiece surface before labeling, and to collect position images of the workpiece and label in real time during the labeling process. The label supply and peeling mechanism is located on one side of the multi-axis robotic arm and is used to supply roll labels and peel the labels off the backing paper. The workpiece conveying and positioning mechanism is located in front of the multi-axis robotic arm and is used to convey the workpiece to be labeled to the labeling station and position the workpiece. The multi-axis robotic arm is equipped with an adaptive flexible labeling head at its end effector for labeling curved workpieces. The adaptive flexible labeling head includes a flexible bonding layer, a negative pressure adsorption cavity, and a deformation adjustment mechanism stacked sequentially from the outside in, as well as a pressure sensing and feedback mechanism disposed between the flexible bonding layer and the negative pressure adsorption cavity. The outer surface of the flexible bonding layer serves as the bonding working surface. The negative pressure adsorption cavity is fitted to the inner side of the flexible bonding layer and provides negative pressure adsorption force. The deformation adjustment mechanism is disposed inside the negative pressure adsorption cavity and connected to the inner wall of the flexible bonding layer via multiple push rods passing through the negative pressure adsorption cavity, for adjusting the initial curvature of the flexible bonding layer. The multi-axis robotic arm, the adaptive flexible labeling head, the vision inspection mechanism, the label supply and peeling mechanism, and the workpiece conveying and positioning mechanism are all communicatively connected to the control system, which is used to receive data and control the coordinated operation of each mechanism.

2. The robotic flexible online labeling device according to claim 1, characterized in that, The flexible bonding layer is a deformable sheet structure; in its natural state, the flexible bonding layer is a convex bowl-shaped arc, and when subjected to external extrusion, it can conform to the surface undulations of the curved workpiece to produce an adhesive deformation; on the inner wall surface of the flexible bonding layer, pressure sensors are embedded in a grid or concentric ring pattern to sense the contact pressure of each area of ​​the bonding surface.

3. The robotic flexible online labeling device according to claim 1 or 2, characterized in that, The negative pressure adsorption chamber is a hollow shell arranged concentrically with the flexible adhesive layer. The side facing the flexible adhesive layer is open, and the outline and size of the open are consistent with the outline of the inner wall of the flexible adhesive layer. The edge of the inner wall of the flexible adhesive layer is fixed to the edge of the open, so that the flexible adhesive layer serves as the bottom closed wall of the negative pressure adsorption chamber. A vacuum pipe interface is provided on the side wall of the negative pressure adsorption chamber, which is connected to an external vacuum generator through an air pipe. The adhesive surface of the flexible adhesive layer is densely covered with adsorption holes that penetrate the wall thickness of the flexible adhesive layer, which are used to adsorb and fix the label to the adhesive surface by negative pressure after the vacuum generator is started.

4. The robotic flexible online labeling device according to claim 3, characterized in that, The diameter of the adsorption pores is 0.3mm to 0.8mm, and the total area of ​​the adsorption pores on the bonding surface accounts for 5% to 15% of the total area of ​​the flexible bonding layer. All the adsorption pores are directly connected to the internal space of the negative pressure adsorption cavity.

5. The robotic flexible online labeling device according to claim 3, characterized in that, The deformation adjustment mechanism includes a support plate and multiple electric push rods arranged in an array. The support plate is arranged parallel above the negative pressure adsorption cavity, and each electric push rod is fixedly installed on the side of the support plate facing the negative pressure adsorption cavity. A through hole is opened at the top of the negative pressure adsorption cavity corresponding to the output end of each electric push rod. A guide sleeve with a sealing ring is installed in the through hole. The front end of each electric push rod passes through the corresponding guide sleeve and extends into the interior of the negative pressure adsorption cavity. The front end of each electric push rod is connected to a corresponding position on the inner wall surface of the flexible adhesive layer through a freely rotatable connector. Each electric push rod extends or retracts independently and in a controlled manner to push different areas of the flexible adhesive layer along the normal direction of its arc surface before labeling, so as to adjust the initial arc shape of the flexible adhesive layer to a shape close to the curvature of the surface to be labeled.

6. The robotic flexible online labeling device according to claim 4 or 5, characterized in that, The pressure sensing and feedback mechanism includes a flexible circuit board, a signal processing chipset, and a metal shield. The flexible circuit board is bonded and fixed to the inner wall of the flexible bonding layer, and the pins of each pressure sensor are connected to corresponding solder points on the flexible circuit board via conductive lines. The signal processing chipset is soldered to the edge of the flexible circuit board, and each signal processing channel is connected to its corresponding sensor to convert the resistance signals collected by each sensor into voltage signals. The metal shield is placed over the signal processing chipset and electrically connected to the ground plane of the flexible circuit board. The output of the flexible circuit board is connected to an analog-to-digital converter via a signal line, and the analog-to-digital converter converts the voltage signals of each channel into digital pressure values ​​and transmits them to the control system.

7. The robotic flexible online labeling device according to claim 1, characterized in that, The visual inspection mechanism includes a gantry frame, a laser contour scanner, an industrial camera, and a grating projector. The laser contour scanner, the industrial camera, and the grating projector are arranged sequentially along the length of the crossbeam of the gantry frame, with their lenses all facing downwards. The laser contour scanner has a built-in laser emitter and a photosensitive receiver, used to emit a linear laser line onto the surface of the workpiece on the workpiece conveying and positioning mechanism and to receive the reflected light to obtain contour point cloud data of the workpiece surface. The grating projector is used to project an coded grating stripe pattern onto the workpiece surface to assist the laser contour scanner in obtaining depth information of the workpiece surface. The industrial camera has a filter installed at the front end of its lens to filter out ambient light interference during the labeling process and to acquire images of the workpiece surface in real time.

8. The robotic flexible online labeling device according to claim 1, characterized in that, The control system includes a main control module, a vision data processing module, and a bus communication module; The main control module includes an industrial control motherboard, a central processing unit, and memory. The main control module is electrically connected to the visual data processing module and the bus communication module, respectively, and is used to perform calculation tasks such as 3D point cloud reconstruction, surface model generation, and labeling path planning, and to issue control commands to each module. The visual data processing module includes an image acquisition card and a graphics processor. The input end of the image acquisition card is electrically connected to the image output end of the visual inspection mechanism, and the graphics processor is electrically connected to the output end of the image acquisition card. It is used to perform parallel accelerated processing on the acquired 3D point cloud data and image data, and transmit the processed data to the main control module. The bus communication module includes a bus controller and a bus transceiver. The bus controller is electrically connected to the main control module, and the bus transceiver is electrically connected to the bus controller and connected to the fieldbus network to realize data interaction and clock synchronization between the main control module and external devices.

9. The robotic flexible online labeling device according to claim 8, characterized in that, The control system also includes a motion control module and a logic control module; The motion control module includes a servo control card and a pulse output interface. The servo control module is electrically connected to the servo drivers of each joint of the multi-axis robotic arm through the pulse output interface. It is used to receive path planning data sent by the main control module through the bus communication module and generate servo drive pulse signals for each joint to drive the robot to perform labeling motion. The logic control module includes a programmable logic device and an input / output expansion interface. The programmable logic device is electrically connected to the drive motor of the workpiece conveying and positioning mechanism, the vacuum pneumatic solenoid valve of the adaptive flexible labeling head, and each drive motor of the label supply and peeling mechanism through the input / output expansion interface, and is used to control the timing logic actions of workpiece conveying, workpiece positioning, label supply and label peeling.

10. A robot flexible online labeling path planning method for curved workpieces, applied to the robot flexible online labeling device for curved workpieces as described in any one of claims 19, characterized in that, Includes the following steps: Step 1: Scan the workpiece using the vision inspection mechanism to obtain the three-dimensional point cloud data and texture information of the workpiece surface, reconstruct the three-dimensional digital surface model of the workpiece, and extract the boundary features and curvature distribution information of the area to be labeled; generate workpiece pose information based on the scan data and feed it back to the workpiece conveying and positioning mechanism to position the workpiece. Step 2: Control the multi-axis robotic arm to move to the label supply and peeling mechanism, and pick up the peeled label through the adaptive flexible labeling head; Step 3: Based on the curvature distribution information of the area to be labeled extracted in Step 1, before the adaptive flexible labeling head contacts the workpiece, control the deformation adjustment mechanism to drive the flexible bonding layer to deform, so that the initial curvature shape of the flexible bonding layer matches the macroscopic curvature of the surface to be labeled. Step 4: Using the reference point of the picked-up label as a reference, plan the labeling motion path of the multi-axis robotic arm on the three-dimensional digital surface model, generate a path point sequence containing position coordinates and attitude angles, and perform interpolation fitting on the path point sequence to form a smooth and continuous robot motion trajectory. Step 5: During the labeling motion of the multi-axis robotic arm, the position and posture of the robotic arm end effector are corrected in real time by the vision detection mechanism; at the same time, the contact pressure data of each area of ​​the flexible bonding layer are acquired in real time by the pressure sensing and feedback mechanism, and the motion parameters of the multi-axis robotic arm are dynamically adjusted according to the contact pressure data. Step 6: After labeling is completed, the labeling quality is checked by the vision inspection mechanism. If the inspection fails, the current workpiece is marked as a defective product.

Images