High-precision diaphragm automatic assembly equipment and process

Abstract

This invention discloses a high-precision automated assembly equipment and process for apertures, belonging to the field of automated assembly technology for optical components. It includes a magazine loading / unloading module, a dispensing module, an aperture tray feeding module, an aperture secondary positioning module, an assembly and bonding module, a segmented UV curing module, a large-area curing module, and a post-curing inspection and unloading module, arranged sequentially along the material conveying direction. It also includes an electrical control system, a vision system, and a MES data interaction unit, each electrically connected to each module. This invention improves the assembly precision of apertures from the traditional ±200μm to ±30μm, solving the problem of low assembly precision caused by the lack of secondary precision positioning and single-vision alignment in existing technologies. Simultaneously, through the addition of a process parameter self-optimization unit, using Bayesian optimization or neural network algorithms to dynamically iteratively optimize based on historical batch quality data, the bonding precision can be further converged to within ±20μm, achieving a further breakthrough in the upper limit of precision and effectively suppressing precision drift between batches.

Description

Technical Field

[0001] This invention relates to the field of automated assembly technology for optical components, and particularly to high-precision automated assembly equipment and processes for apertures. Background Technology

[0002] In the manufacturing of optical modules (such as TRX boards and lens barrel assemblies), the assembly, dispensing, UV curing, and testing of apertures are the core processes that determine the optical performance of the product. With the increasing demands for component precision in optical communication and precision imaging, the assembly accuracy, consistency, and traceability of apertures have become key factors restricting product quality. Currently, existing assembly methods mainly include manual assembly, semi-automatic equipment assembly, and a small number of fully automated equipment assembly methods.

[0003] Manual assembly relies on operators visually aligning, manually dispensing glue, and bonding the components. Its positioning error is typically greater than ±200μm, and glue dispensing uniformity is poor, easily leading to problems such as glue leakage, glue breaks, and inconsistent glue amounts. Furthermore, manual assembly is extremely inefficient, with each product taking over 60 seconds to assemble, and the lack of standardized testing results in a persistently high defect rate, making it difficult to meet the dual requirements of efficiency and yield for large-scale production.

[0004] Semi-automated equipment only automates the dispensing or curing processes; the loading, alignment, and bonding of the aperture still require manual assistance. This type of equipment typically uses simple fixtures for positioning, lacking a secondary precision positioning structure for the aperture, and relies solely on a single vision system for alignment. This results in generally low assembly accuracy, typically between ±100μm and ±200μm, failing to meet the ±50μm accuracy requirements of high-end optical modules. In the curing stage, existing equipment often uses single-source fixed-point irradiation, leading to uneven curing and incomplete curing of deep adhesives. Furthermore, the lack of real-time monitoring of UV lamp power attenuation, lamp temperature, and lifespan makes it difficult to guarantee adhesive curing quality, resulting in decreased product reliability after long-term operation. In addition, existing semi-automated production processes generally lack material barcode binding and process traceability mechanisms, making quality control extremely difficult. Once batch defects occur, it is impossible to effectively pinpoint the root cause of the problem.

[0005] In recent years, although a small number of fully automated equipment have achieved automated connection between various processes, they are essentially rigid automation with the following significant drawbacks: First, process parameters (such as dispensing pressure, bonding speed, and curing power) are fixed once set and cannot be adaptively adjusted according to batch differences in incoming materials, changes in environmental temperature and humidity, or equipment status drift, resulting in large fluctuations in product quality between batches. Second, the equipment lacks the ability to perceive and predict its own operating status in real time, and cannot provide early warnings before failures occur, leading to frequent sudden shutdowns and seriously affecting the overall efficiency of the equipment. Third, the final inspection stage relies solely on traditional image processing algorithms for defect judgment, which is insufficient in identifying subtle defects such as microcracks, microbubbles, and edge micro-defects, posing a certain risk of missed or false detections. Fourth, when switching product models, manual adjustment of the fixtures, trajectories, and parameters of each module is required, resulting in long changeover times and complex operations, making it difficult to adapt to the flexible production needs of multiple varieties and small batches.

[0006] In summary, existing technologies generally suffer from systemic defects such as low assembly precision, lack of adaptive control of process parameters, lack of predictive maintenance of equipment status, limited accuracy of defect detection, low changeover efficiency, and lack of process traceability. There is an urgent need to propose an automated assembly scheme that integrates high-precision mechanical structure and intelligent soft control system to achieve precision, intelligence, and flexibility in the aperture assembly process. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to overcome the problems of low assembly accuracy and poor yield of the aperture due to the lack of secondary fine positioning of the aperture and dual vision closed-loop alignment structure in the prior art.

[0008] To achieve the above-mentioned objectives, the first aspect of the present invention proposes a high-precision automated assembly equipment for apertures, comprising a magazine loading and unloading module, a dispensing module, an aperture tray feeding module, an aperture secondary positioning module, an assembly and bonding module, a segmented UV curing module, a large-area curing module, and a post-curing inspection and unloading module arranged sequentially along the material conveying direction, and including an electrical control system, a vision system, and an MES data interaction unit that are electrically connected to each module respectively. The magazine loading and unloading module is used to complete the automatic loading, barcode scanning and binding, and conveying of the substrate carrier; The dispensing module is used to identify the dispensing area on the substrate, measure the substrate height, and complete automatic dispensing. The aperture tray feeding module is used to complete the automatic tray separation, aperture identification and handling of the aperture tray; The secondary positioning module of the aperture is used to clean, position and fix the aperture after it has been transported; The assembly and bonding module is used to grasp the positioned aperture and complete the precise bonding between the aperture and the substrate based on the visual positioning result. The segmented UV curing module is used for pre-curing and final curing of the bonded products. The large-area curing module is used to complete the overall curing of the product; The post-curing inspection and unloading module is used to inspect the product assembly quality and separate qualified and unqualified products for unloading. The MES data interaction unit is used to collect material information, process parameters, equipment status and test results, and to establish product full-process traceability information; Also includes: The intelligent process closed-loop control system includes a process parameter self-optimization unit, a digital twin mapping unit, an AI vision re-judgment unit, and an adaptive type change unit. The process parameter self-optimization unit dynamically adjusts dispensing pressure, bonding speed and UV curing power based on historical assembly data and real-time sensor signals, using fuzzy PID or neural network algorithms. The digital twin mapping unit constructs a virtual model of the device, synchronizes the motion status and sensor data of each module in real time, and performs fault prediction and health management. The AI ​​vision re-judgment unit performs deep learning re-inference on the judgment results of the detection and feeding module to identify micro-cracks, small displacements or adhesive layer bubble defects. The adaptive model change unit identifies the product model by scanning the code and automatically calls up the corresponding process formula and motion trajectory to achieve manual model change.

[0009] Optionally, the process parameter self-optimization unit includes a data acquisition subunit, a database subunit, an optimization algorithm subunit, and an execution subunit; The data acquisition subunit collects dispensing pressure, bonding pressure, UV power, and temperature signals in real time. The database sub-unit stores historical batch data and corresponding quality results; The optimization algorithm subunit is based on a genetic algorithm or a Bayesian optimization algorithm and outputs the optimal combination of process parameters. The execution subunit sends the optimized parameters to the corresponding execution mechanism through the electrical control system.

[0010] Optionally, the digital twin mapping unit includes a device model subunit, a data synchronization subunit, a simulation and deduction subunit, and an early warning subunit; The device model sub-units are constructed based on 3D modeling software to build the geometric and kinematic models of each module; The data synchronization subunit acquires device operation data in real time via OPC UA or Modbus protocol and maps it to the virtual model; The simulation and deduction subunit simulates the equipment operating status under different working conditions and predicts potential faults; The early warning subunit issues a maintenance prompt when the simulation result exceeds the threshold.

[0011] Optionally, the AI ​​visual re-judgment unit includes an image preprocessing subunit, a feature extraction subunit, a classification reasoning subunit, and a result fusion subunit; The image preprocessing subunit performs noise reduction and enhancement on the detected image; The feature extraction subunit uses a convolutional neural network to extract defect features; The classification reasoning subunit outputs the defect type and confidence level based on the Softmax classifier; The result fusion subunit performs a logical AND operation between the AI ​​inference result and the original detection result, and outputs the final judgment.

[0012] Optionally, the adaptive model switching unit includes a model identification subunit, a recipe library subunit, a trajectory generation subunit, and a switching execution subunit; The model identification subunit reads the product model by scanning a code or using RFID. The formula library subunit stores the dispensing trajectory, bonding pressure, curing time and testing standards corresponding to different models; The trajectory generation subunit automatically generates motion control commands for each module based on the recipe data; The switching execution subunit automatically completes recipe loading and parameter switching when the device is idle.

[0013] A second aspect of this invention provides a high-precision automated aperture assembly process, employing the aforementioned high-precision automated aperture assembly equipment, comprising the following steps: S1. The substrate loading and magazine unloading module completes the automatic transport of the carrier and scans the magazine code and carrier code. S2. Dispensing: The dispensing module identifies the dispensing area on the substrate, measures the substrate height, compensates for flatness errors, and then completes automatic dispensing. S3, Aperture feeding: The aperture tray feeding module automatically divides the tray and identifies the aperture position, and moves the aperture to the aperture secondary positioning module; S4. Aperture positioning: The secondary aperture positioning module completes the cleaning, positioning, and fixing of the aperture. S5. Assembly and bonding: The assembly and bonding module picks up the positioned aperture and completes the precise bonding of the aperture to the substrate according to the upper and lower visual positioning results, and performs pre-curing. S6. Curing: The product enters the segmented UV curing module and the large-area curing module in sequence to complete the curing process, while monitoring the UV lamp power, temperature and working life. S7. Inspection and unloading: After curing, the qualified and unqualified products are separated and unloaded after visual inspection and height inspection. S8. Data Upload: The MES data interaction unit binds and uploads material information, process parameters, equipment operating status, and test results to achieve full traceability of the product production process. Its characteristic is that it further includes: S9. Process self-optimization step: The process parameter self-optimization unit dynamically adjusts the dispensing pressure and curing power of subsequent batches based on the test results of the current batch. S10, Digital Twin Monitoring Steps: The digital twin mapping unit simulates the equipment's operating status in real time, predicts and warns of potential anomalies; S11, AI re-inspection step: The AI ​​vision re-judgment unit performs random re-inspection on products that have been determined to be qualified by the inspection and unloading module to confirm that there are no minor defects. S12, Adaptive Changeover Step: When the barcode scan identifies a new product model, the adaptive changeover unit automatically loads the corresponding formula and switches the motion trajectory.

[0014] Optionally, in step S9, the process parameter self-optimization unit adopts a Bayesian optimization algorithm, with the bonding accuracy and the hardness of the cured adhesive layer as the objective functions, and iteratively optimizes to bring the bonding accuracy to within ±20μm.

[0015] Optionally, in step S10, the digital twin mapping unit maps the motion curves and vibration spectrum of each axis of the device in real time, and triggers an early warning when the vibration amplitude exceeds 30% of the historical baseline.

[0016] Optionally, in step S11, the AI ​​visual re-judgment unit uses a ResNet-50 convolutional neural network to classify defects in the detected image, with a classification accuracy of not less than 99.5%.

[0017] Optionally, in step S12, the adaptive type-changing unit completes the switching of all process parameters and motion trajectories within 30 seconds, and the first piece after type-changing enters the automatic production state.

[0018] The beneficial effects of this invention are: 1. The high-precision automated aperture assembly equipment and process of this invention achieves coarse positioning through asymmetrical positioning pins in the aperture secondary positioning module, followed by visual fine correction through a visual correction unit. This, combined with the upper and lower visual units in the assembly and bonding module, forms a dual-vision closed-loop alignment system, improving the aperture assembly accuracy from the traditional ±200μm to ±30μm. This solves the problem of low assembly accuracy caused by the lack of secondary fine positioning and single-vision alignment in existing technologies. Furthermore, through the added process parameter self-optimization unit, which uses Bayesian optimization or neural network algorithms to dynamically iterate and optimize based on historical batch quality data, the bonding accuracy can be further converged to within ±20μm, achieving a further breakthrough in the upper limit of accuracy and effectively suppressing batch-to-batch accuracy drift.

[0019] 2. The high-precision automated aperture assembly equipment and process of the present invention, by introducing pressure sensors and displacement sensors into the assembly and bonding module, achieves dual closed-loop control of pressure and displacement during the bonding process, precisely controlling the bonding pressure between 5N and 20N. This avoids aperture deformation due to pressure overload or incomplete bonding due to insufficient pressure, reducing the product defect rate to below 0.5%. Simultaneously, through the self-optimization unit of process parameters, the adaptive adjustment of bonding speed and holding time further improves the consistency of adhesive layer thickness and bonding strength, resulting in significantly better performance than existing technologies in subsequent environmental reliability testing.

[0020] 3. The high-precision aperture automated assembly equipment and process of this invention, through the design of a segmented UV curing module composed of a pre-curing unit and a final curing unit, and equipped with a power detection sensor, a temperature sensor, and a lifespan timing unit, performs real-time monitoring and alarm for three parameters: curing energy, temperature, and equipment status. This ensures the uniformity and stability of the curing effect and avoids poor adhesive curing or peeling due to UV lamp attenuation or abnormal temperature. Furthermore, by using a digital twin mapping unit to simulate and predict the UV lamp power attenuation trend, maintenance prompts can be proactively triggered before the curing quality deteriorates, reducing unplanned equipment downtime by more than 60%, effectively ensuring the continuity of mass production and the long-term stability of curing quality.

[0021] 4. The high-precision aperture automated assembly equipment and process of this invention, by integrating a MES data interaction unit, binds and uploads material information, process parameters, equipment status, and detection data throughout the entire process, constructing a full-chain traceability system and realizing quality traceability and control for each product's production process. Furthermore, by introducing an AI visual re-judgment unit, and employing a convolutional neural network to perform deep learning re-inference on the judgment results of the detection and unloading module, it can accurately identify minute defects such as micro-cracks, minor offsets, or internal bubbles in the adhesive layer, increasing the defect detection rate to over 99.5%. This compensates for the shortcomings of traditional algorithm detection capabilities and provides dual assurance for product quality.

[0022] 5. The high-precision aperture automated assembly equipment and process of the present invention, by setting an adaptive type-changing unit, automatically calls the corresponding process formula and motion trajectory after scanning and identifying the product model. It can complete all parameter switching within 30 seconds, and the first piece after type-changing enters the qualified production state. It completely changes the cumbersome mode of traditional equipment that requires manual replacement of fixtures and manual adjustment of parameters. The type-changing efficiency is improved by more than 98%, enabling the equipment to flexibly adapt to the flexible production needs of multiple varieties and small batches, and significantly enhancing the adaptability of the production line and its market competitiveness. Attached Figure Description

[0023] Figure 1 This is a top view of the high-precision aperture automated assembly equipment provided in an embodiment of the present invention; Figure 2 for Figure 1 Schematic diagram of the loading and unloading module for the medium-lift clip; Figure 3 for Figure 1 A schematic diagram of the dispensing module; Figure 4 for Figure 1 Schematic diagram of the structure of the central aperture tray feeding module; Figure 5 for Figure 1 Schematic diagram of the secondary positioning module of the central aperture; Figure 6 for Figure 1 A schematic diagram of the structure of the mid-assembly bonding module; Figure 7 for Figure 1 Schematic diagram of the mid-segment UV curing module; Figure 8 for Figure 1 Schematic diagram of the structure of the material feeding and inspection module after intermediate curing; Figure 9 This is a schematic diagram of the external structure of a high-precision aperture automated assembly equipment provided in an embodiment of the present invention; Figure 10 The system structure block diagram of the high-precision aperture automated assembly equipment provided in the embodiments of the present invention is shown.

[0024] Explanation of reference numerals in the attached figures: 1. Magazine loading / unloading module; 11. Magazine loading / unloading bin; 12. Servo-driven pusher cylinder; 13. Carrier hook mechanism; 14. Barcode scanner; 15. Automatic cover plate picking and placing mechanism; 2. Dispensing module; 21. Three-axis motion platform; 22. Pressure dispensing valve; 23. Upper CCD vision unit; 24. Laser displacement sensor; 25. Brush cleaning mechanism; 26. Glue balance monitoring sensor; 3. Tray feeding module; 31. Drawer-type dual material bin; 32. Pulling mechanism; 33. Upper CCD guide unit; 34. Four-axis handling module; 4. Aperture secondary positioning module; 41. Blowing and suction cleaning module; 42. Transfer platform; 43. Moving cylinder; 5. Assembly and bonding module; 51. Three-axis transport module; 52. Vacuum suction nozzle; 53. Upper vision unit; 54. Lower CCD vision unit; 55. Pressure sensor; 6. Segmented UV curing module; 61. Lifting pre-curing mechanism; 62. Pre-curing UV lamp; 63. Surface light source UV lamp; 7. Large-area solidification module; 8. Curing inspection and unloading module; 81. High-definition CCD detection unit; 82. Second laser displacement sensor; 83. Unloading mechanism; 84. Defective product rejection mechanism.

[0025] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0026] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0027] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection, a direct connection, or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances. Example 1

[0028] This embodiment provides a high-precision automated aperture assembly device. Figure 1 and Figure 9 The overall structure of the equipment is shown. The equipment is located along the material conveying direction ( Figure 1 (In the direction of the middle arrow), the equipment is configured sequentially as follows: magazine loading / unloading module 1, dispensing module 2, aperture tray feeding module 3, aperture secondary positioning module 4, assembly and bonding module 5, segmented UV curing module 6, large-area curing module 7, and post-curing inspection and unloading module 8. The equipment is also equipped with an electrical control system, a vision system, and an MES data interaction unit (not shown in the figure) connected to the electrical control system, to complete the electrical control, visual positioning and inspection, and full-process data interaction of the entire assembly process.

[0029] The following is combined Figures 2 to 8 The structure and working principle of each functional module are explained in detail.

[0030] Figure 2This is a schematic diagram of the magazine loading / unloading module 1. The module mainly includes a magazine loading / unloading bin 11, a servo-driven pusher cylinder 12, a carrier hook mechanism 13, a barcode scanner 14, and an automatic cover plate handling mechanism 15. The magazine loading / unloading bin 11 is a drawer-type design, including three loading buffer positions and two unloading buffer positions for storing fully loaded carrier magazines and empty carrier magazines. During operation, the servo-driven pusher cylinder 12 precisely pushes the magazine to the designated position, and the carrier hook mechanism 13 pulls a single carrier out of the magazine to the loading conveyor track, or pushes a carrier returning from downstream into the unloading magazine. Both the carrier hook mechanism 13 and the carrier positioning components use a contour jig, which can be replaced according to different product models of carriers, enabling rapid model changeover. The barcode scanner 14 is located on one side of the hook mechanism, scanning the magazine code on the magazine and the carrier code on the carrier, and the electrical control system binds the two information to establish a material association. The automatic cover plate pick-and-place mechanism 15 is used to remove the protective cover plate before the carrier enters the dispensing module 2 and to reapply the cover plate upon its return. This module realizes full automation of substrate loading, initial binding of material information, and rapid changeover, solving the problems of low efficiency and lack of traceability in manual loading.

[0031] Figure 3 This is a schematic diagram of the dispensing module 2. The module mainly includes a three-axis motion platform 21, a pressure dispensing valve 22, an upper CCD vision unit 23, a laser displacement sensor 24, a brush cleaning mechanism 25, and a glue balance monitoring sensor 26. The three-axis motion platform 21 has a positioning accuracy of no more than 20μm and is used to move the pressure dispensing valve 22 within the XYZ space. Before dispensing, the upper CCD vision unit 23 captures a QR code on the substrate surface to identify the product and retrieve the corresponding dispensing program. It also identifies features such as gold lines on the substrate to accurately locate the area requiring dispensing. The laser displacement sensor 24 measures the height of three preset points on the substrate surface. The electrical control system calculates the flatness error of the substrate based on the measurement data and controls the Z-axis of the three-axis motion platform 21 for real-time height compensation during the dispensing process. Subsequently, the three-axis motion platform 21 drives the pressure dispensing valve 22 to complete the dispensing according to the set trajectory and height. The pressure dispensing valve 22 stabilizes the dispensing volume by precisely controlling air pressure and is compatible with 365nm UV adhesive. The brush cleaning mechanism 25 periodically cleans the dispensing valve needle and prevents clogging by periodically draining adhesive. The adhesive level monitoring sensor 26 issues an alarm when the adhesive level is insufficient. This module ensures high precision in dispensing position and volume control through visual positioning and automatic compensation for flatness errors, solving the problem of large errors in traditional manual dispensing.

[0032] Figure 4This is a structural diagram of the aperture tray feeding module 3. This module mainly includes a drawer-type dual-bin hopper 31, a pulling mechanism 32, an upper CCD guide unit 33, and a four-axis transport module 34. The drawer-type dual-bin hopper 31 is divided into a full hopper and an empty hopper. After the operator places a stack of trays containing apertures into the full hopper, the pulling mechanism 32 automatically separates the bottom tray and pulls it to the picking station. The upper CCD guide unit 33 is positioned above the picking station, scanning the QR code on the tray to bind material information and scanning the apertures within the tray to identify their position and orientation. Based on the visual positioning information provided by the upper CCD guide unit 33, the four-axis transport module 34 accurately grasps the apertures and transports them to the aperture secondary positioning module 4. This module realizes automatic aperture feeding and visually guided grasping, replacing manual placement and improving feeding accuracy and efficiency.

[0033] Figure 5 This is a schematic diagram of the secondary positioning module 4 for the aperture. This module includes a blow-suction cleaning module 41, a transfer platform 42, a moving cylinder 43, an asymmetric positioning pin, a vision correction unit, and a vacuum adsorption fixing structure. When the four-axis transport module 34 places the aperture onto the transfer platform 42, the blow-suction cleaning module 41 simultaneously blows and suctions its surface to remove foreign objects. Subsequently, the moving cylinder 43 pushes the transfer platform 42 into the coarse positioning station. In the coarse positioning station, the aperture is placed on a fixture with an asymmetric positioning pin. The asymmetric positioning pin engages with pre-set asymmetric holes or features on the aperture, achieving unique coarse positioning of the aperture's initial position and orientation through mechanical limiting. Afterward, the transfer platform 42 carries the aperture into the fine correction station, where the vacuum adsorption fixing structure adsorbs and fixes the aperture to prevent movement. The vision correction unit includes a high-resolution industrial camera and an image processing unit. It takes a picture of the fixed aperture from above, calculates the offset and rotation angle of the aperture relative to the standard position using image processing algorithms, and sends this data to the electrical control system. This module improves the positional accuracy of the aperture from an uncertain state during material feeding to no more than ±50μm through secondary positioning, laying a key foundation for subsequent high-precision bonding.

[0034] Figure 6This is a schematic diagram of the assembly and bonding module 5. This module mainly includes a three-axis transport module 51, a vacuum suction nozzle 52, an upper vision unit 53, a lower CCD vision unit 54, a pressure sensor 55, and a displacement sensor. The three-axis transport module 51 drives the vacuum suction nozzle 52 to accurately grasp the aperture on the secondary positioning module 4 based on the precise offset data provided by the vision correction unit. After the nozzle grasps the aperture, the upper vision unit 53 photographs the bottom surface of the aperture from below to obtain its precise spatial orientation. Simultaneously, the lower CCD vision unit 54 photographs the surface of the substrate that has flowed into the bonding station to obtain the precise coordinates of the dispensing area. The electrical control system comprehensively processes the visual data from the upper vision unit 53 and the lower CCD vision unit 54, forming a dual-vision closed-loop alignment system. It calculates the deviation and angular deviation between the aperture and the substrate dispensing area in the X and Y directions, and controls the three-axis transport module 51 to perform precise translation and rotation compensation, ensuring an alignment accuracy of no more than ±30μm.

[0035] After alignment, the Z-axis servo motor of the three-axis transport module 51 drives the vacuum suction nozzle 52 to descend, performing the bonding action. During bonding, the pressure sensor 55 monitors the contact pressure between the nozzle and the substrate in real time, and the displacement sensor monitors the Z-axis descent stroke in real time. Both signals are fed back to the electrical control system. The electrical control system adopts a dual closed-loop control strategy of pressure and displacement: when the pressure detected by the pressure sensor 55 reaches a preset value (e.g., 10N), or when the downward pressure depth detected by the displacement sensor reaches the safety limit set according to the substrate height, the system immediately stops the Z-axis feed to ensure that the aperture is neither damaged or deformed due to overpressure, nor does it experience incomplete bonding due to insufficient pressure. The bonding pressure is stably maintained within an adjustable range of 5N to 20N. After bonding, the product is immediately moved to the subsequent pre-curing unit, which is part of the segmented UV curing module 6 and located downstream of the assembly bonding station. The pre-curing unit performs UV pre-curing on the product for approximately 10 seconds to initially fix the relative position of the aperture and the substrate, preventing displacement during subsequent transfers.

[0036] Figure 7This is a schematic diagram of the segmented UV curing module 6. This module includes a lifting pre-curing mechanism 61, a pre-curing UV lamp 62, a surface UV lamp 63, a power detection sensor, a temperature sensor, and a lifespan timing unit. After initial pre-curing, the product flows to this module. The lifting pre-curing mechanism 61 descends, bringing the pre-curing UV lamp 62 closer to the product. Simultaneously, the surface UV lamp 63 provides comprehensive irradiation to the product, completing a final curing process of approximately 30 seconds. The surface UV lamp 63 has a light intensity of up to 3000 mw / cm², ensuring deep curing of the adhesive. Throughout the curing process, the power detection sensor monitors the output power of the UV lamp in real time, the temperature sensor monitors the lamp's operating temperature, and the lifespan timing unit accumulates the UV lamp's operating time. The system automatically issues an alarm when power decay, abnormal temperature, or the accumulated time exceeds the set lifespan of 20,000 hours. The large-area curing module 7 is located after the segmented UV curing module 6. It includes one or more high-power surface light sources to provide comprehensive supplementary irradiation to the product, ensuring complete curing of the adhesive even in potentially shaded areas, further improving curing reliability. This module ensures uniform curing depth and stable performance of the adhesive through segmented curing and real-time monitoring of multiple parameters, solving the problems of unstable curing and lack of monitoring in existing technologies.

[0037] Figure 8 This is a schematic diagram of the post-curing inspection and unloading module 8. This module mainly includes a high-definition CCD inspection unit 81, a second laser displacement sensor 82, an unloading mechanism 83, and a defective product rejection mechanism 84. Cured products are transported to this module. The high-definition CCD inspection unit 81 photographs the product and automatically detects appearance defects such as mounting misalignment, excess adhesive, and insufficient adhesive through image processing algorithms. The second laser displacement sensor 82 measures the height of six preset points on the product surface to calculate the height tolerance and flatness between the aperture and the substrate. The electrical control system comprehensively judges the inspection results: if the product's height tolerance is within ±0.1mm, flatness is within ±0.15mm, and there are no appearance defects, it is judged as a qualified product and is transported to the qualified product unloading magazine station by the unloading mechanism 83; if any indicator fails, the defective product rejection mechanism 84 diverts it to the defective product recycling area.

[0038] The electrical control and data interaction process of the equipment operation is as follows: the electrical control system (such as using a Siemens 1515TPLC, in conjunction with Huichuan servo motors and ET200 distributed IO modules) controls the sequential actions of the above modules. The vision system includes multiple industrial cameras such as dispensing CCD, aperture picking CCD, assembly CCD, and inspection CCD, OPT lenses, and dedicated light sources. It uses a 13-point calibration method for high-precision calibration to achieve visual guidance and inspection functions. The MES data interaction unit, as the interface between the equipment and the factory's manufacturing execution system, binds material information such as magazine codes, carrier codes, tray codes, and adhesive batch codes obtained from barcode scanning, real-time process parameters such as dispensing pressure, UV lamp power, and curing temperature, as well as inspection data such as OK / NG results and height tolerance values ​​determined by the inspection module, to the MES system in real time. This establishes a complete production process file for each product, enabling full-process traceability. Example 2

[0039] This embodiment provides a high-precision automated assembly process for apertures. This process is based on the equipment described in the first embodiment, and the specific steps and preferred parameters are as follows: S1: Substrate loading step, the carrier hook mechanism 13 of the magazine loading / unloading module 1 pulls out the carrier, the barcode scanner 14 scans the magazine code and carrier code and binds them, and the carrier is transported to the dispensing station.

[0040] S2: Dispensing step, the upper CCD vision unit 23 of the dispensing module 2 positions the dispensing area of ​​the substrate, the laser displacement sensor 24 performs 3-point height measurement to compensate for flatness error, and then the three-axis motion platform 21 drives the pressure dispensing valve 22 to complete automatic dispensing.

[0041] S3: Aperture feeding step, the material pulling mechanism 32 of the aperture tray feeding module 3 automatically divides the tray, the upper CCD guiding unit 33 identifies the position of the aperture inside the tray, and the four-axis transport module 34 grabs the aperture and transfers it to the aperture secondary positioning module 4.

[0042] S4: Secondary positioning step of the aperture. The secondary positioning module 4 of the aperture performs coarse positioning of the aperture by using asymmetric positioning pins. The vacuum adsorption fixing structure fixes it. The visual correction unit takes pictures of the aperture and calculates the position offset and angle, and performs visual fine correction to improve the positioning accuracy to no more than ±50μm.

[0043] S5: Assembly and bonding step. The vacuum suction nozzle 52 of the assembly and bonding module 5 picks up the aperture. Through the dual vision closed-loop alignment system composed of the upper vision unit 53 and the lower CCD vision unit 54, the deviation between the aperture and the substrate dispensing area is calculated and compensated to achieve precise alignment. The bonding accuracy is no more than ±30μm. Subsequently, under the dual closed-loop control of the pressure sensor 55 and the displacement sensor, the precision bonding is performed with a preset pressure between 5N and 20N, and a pre-curing time of 10 seconds is performed.

[0044] S6: Segmented and Large-Area Curing Steps. After pre-curing, the product first enters the segmented UV curing module 6. The lifting pre-curing mechanism 61 descends, and the surface light source UV lamp 63 is turned on for 30 seconds of final curing. Then, the product enters the large-area curing module 7 for supplementary curing. Throughout the curing process, power detection sensors, temperature sensors, and lifespan timing units monitor the entire process.

[0045] S7: Inspection and unloading step. After curing, the product enters the post-curing inspection and unloading module 8. The high-definition CCD inspection unit 81 detects mounting misalignment and adhesive defects, and the second laser displacement sensor 82 performs 6-point height measurement to control the height tolerance within ±0.1mm and the flatness within ±0.15mm. Qualified and unqualified products are then separated and unloaded.

[0046] During the execution of each step from S1 to S7, the MES data interaction unit collects material binding information, process parameters, equipment status and test results data of each process in real time, and uploads them to the manufacturing execution system to establish a complete full-process traceability archive. Example 3

[0047] like Figure 10 As shown, this embodiment, based on Embodiments 1 and 2, adds an intelligent process closed-loop control system and a digital twin monitoring platform. This system is integrated into the equipment's central control unit in the form of a system block diagram, without relying on additional mechanical structures. The following provides a detailed description of each functional unit.

[0048] The process parameter self-optimization unit includes a data acquisition subunit, a database subunit, an optimization algorithm subunit, and an execution subunit. The data acquisition subunit acquires dispensing pressure, bonding pressure, UV power, and temperature signals in real time through a sensor network. The database subunit stores historical batch data and corresponding quality inspection results. The optimization algorithm subunit uses a Bayesian optimization algorithm, with bonding accuracy and post-cured adhesive layer hardness as objective functions, to iteratively optimize the process. The execution subunit sends the optimized parameters to the dispensing valve 22, the three-axis transport module 51, and the segmented UV curing module 6 via the electrical control system.

[0049] The digital twin mapping unit includes an equipment model subunit, a data synchronization subunit, a simulation deduction subunit, and an early warning subunit. The equipment model subunit constructs geometric and kinematic models of each module based on SolidWorks or Unity 3D. The data synchronization subunit acquires data of PLC and servo drivers in real time through the OPC UA protocol and maps the data to the virtual model. The simulation deduction subunit simulates the equipment operation state under different working conditions, such as acceleration and deceleration impact, thermal expansion influence, etc. The early warning subunit triggers a maintenance prompt when the simulation result exceeds the historical baseline by 30%. In actual operation, this unit successfully predicted the power attenuation trend of the segmented UV curing module 6 for 3 times, avoiding batch curing defects.

[0050] The AI visual recheck unit includes an image preprocessing subunit, a feature extraction subunit, a classification inference subunit, and a result fusion subunit. The image preprocessing subunit performs Gaussian filtering and contrast enhancement on the detected image. The feature extraction subunit adopts the ResNet-50 convolutional neural network to extract defect features. The classification inference subunit outputs the defect type and confidence based on the Softmax classifier. The result fusion subunit performs a logical AND operation on the AI inference result and the original detection result, and outputs the final judgment. After testing, the recognition accuracy of this unit for microcracks and tiny bubbles is increased from 95% to 99.5%, and the false detection rate is reduced to below 0.1%.

[0051] The adaptive model change unit includes a model identification subunit, a formula library subunit, a trajectory generation subunit, and a switching execution subunit. The model identification subunit reads the product model through the code scanner 14. The formula library subunit stores the dispensing trajectory, bonding pressure, curing time and detection standard corresponding to different models. The trajectory generation subunit automatically generates motion control instructions for each module based on the formula data. The switching execution subunit automatically completes formula loading and parameter switching when the equipment is idle. In practical applications, the model change time is shortened from the original 30 minutes of manual operation to within 30 seconds, and the first piece is qualified.

[0052] In this embodiment, by introducing the intelligent process closed-loop regulation system and the digital twin monitoring platform, on the basis of the original high-precision mechanical structure, the present invention further realizes adaptive optimization of process parameters, real-time prediction of equipment status, intelligent recheck of defect detection and automation of model switching, significantly improves the intelligent level of the equipment and the stability of product quality, increases the product yield to over 99.9%, and improves the overall equipment efficiency by 25%. At the same time, the technical solution in the form of the system block diagram does not rely on adding new complex structures, is easy to transform and upgrade on existing equipment, and has good industrial promotion value.

[0053] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A high-precision automated assembly equipment for apertures, comprising a magazine loading and unloading module, a dispensing module, an aperture tray feeding module, an aperture secondary positioning module, an assembly and bonding module, a segmented UV curing module, a large-area curing module, and a post-curing inspection and unloading module arranged sequentially along the material conveying direction, and including an electrical control system, a vision system, and an MES data interaction unit that are electrically connected to each module respectively. Its features are, The magazine loading and unloading module is used to complete the automatic loading, barcode scanning and binding, and conveying of the substrate carrier; The dispensing module is used to identify the dispensing area on the substrate, measure the substrate height, and complete automatic dispensing. The aperture tray feeding module is used to complete the automatic tray separation, aperture identification and handling of the aperture tray; The secondary positioning module of the aperture is used to clean, position and fix the aperture after it has been transported; The assembly and bonding module is used to grasp the positioned aperture and complete the precise bonding between the aperture and the substrate based on the visual positioning result. The segmented UV curing module is used for pre-curing and final curing of the bonded products. The large-area curing module is used to complete the overall curing of the product; The post-curing inspection and unloading module is used to inspect the product assembly quality and separate qualified and unqualified products for unloading. The MES data interaction unit is used to collect material information, process parameters, equipment status and test results, and to establish product full-process traceability information; Also includes: The intelligent process closed-loop control system includes a process parameter self-optimization unit, a digital twin mapping unit, an AI vision re-judgment unit, and an adaptive type change unit. The process parameter self-optimization unit dynamically adjusts dispensing pressure, bonding speed and UV curing power based on historical assembly data and real-time sensor signals, using fuzzy PID or neural network algorithms. The digital twin mapping unit constructs a virtual model of the device, synchronizes the motion status and sensor data of each module in real time, and performs fault prediction and health management. The AI ​​vision re-judgment unit performs deep learning re-inference on the judgment results of the detection and feeding module to identify micro-cracks, small displacements or adhesive layer bubble defects. The adaptive model change unit identifies the product model by scanning the code and automatically calls up the corresponding process formula and motion trajectory to achieve manual model change.

2. The high-precision aperture automated assembly equipment according to claim 1, characterized in that, The process parameter self-optimization unit includes a data acquisition subunit, a database subunit, an optimization algorithm subunit, and an execution subunit. The data acquisition subunit collects dispensing pressure, bonding pressure, UV power, and temperature signals in real time. The database sub-unit stores historical batch data and corresponding quality results; The optimization algorithm subunit is based on a genetic algorithm or a Bayesian optimization algorithm and outputs the optimal combination of process parameters. The execution subunit sends the optimized parameters to the corresponding execution mechanism through the electrical control system.

3. The high-precision aperture automated assembly equipment according to claim 1, characterized in that, The digital twin mapping unit includes a device model subunit, a data synchronization subunit, a simulation and deduction subunit, and an early warning subunit; The device model sub-units are constructed based on 3D modeling software to build the geometric and kinematic models of each module; The data synchronization subunit acquires device operation data in real time via OPC UA or Modbus protocol and maps it to the virtual model; The simulation and deduction subunit simulates the equipment operating status under different working conditions and predicts potential faults; The early warning subunit issues a maintenance prompt when the simulation result exceeds the threshold.

4. The high-precision aperture automated assembly equipment according to claim 1, characterized in that, The AI ​​visual re-judgment unit includes an image preprocessing subunit, a feature extraction subunit, a classification reasoning subunit, and a result fusion subunit. The image preprocessing subunit performs noise reduction and enhancement on the detected image; The feature extraction subunit uses a convolutional neural network to extract defect features; The classification reasoning subunit outputs the defect type and confidence level based on the Softmax classifier; The result fusion subunit performs a logical AND operation between the AI ​​inference result and the original detection result, and outputs the final judgment.

5. The high-precision aperture automated assembly equipment according to claim 1, characterized in that, The adaptive model switching unit includes a model identification subunit, a recipe library subunit, a trajectory generation subunit, and a switching execution subunit; The model identification subunit reads the product model by scanning a code or using RFID. The formula library subunit stores the dispensing trajectory, bonding pressure, curing time and testing standards corresponding to different models; The trajectory generation subunit automatically generates motion control commands for each module based on the recipe data; The switching execution subunit automatically completes recipe loading and parameter switching when the device is idle.

6. A high-precision automated assembly process for apertures, characterized in that, The high-precision aperture automated assembly equipment according to any one of claims 1 to 5 includes the following steps: S1. The substrate loading and magazine unloading module completes the automatic transport of the carrier and scans the magazine code and carrier code. S2. Dispensing: The dispensing module identifies the dispensing area on the substrate, measures the substrate height, compensates for flatness errors, and then completes automatic dispensing. S3, Aperture feeding: The aperture tray feeding module automatically divides the tray and identifies the aperture position, and moves the aperture to the aperture secondary positioning module; S4. Aperture positioning: The secondary aperture positioning module completes the cleaning, positioning, and fixing of the aperture. S5. Assembly and bonding: The assembly and bonding module picks up the positioned aperture and completes the precise bonding of the aperture to the substrate according to the upper and lower visual positioning results, and performs pre-curing. S6. Curing: The product enters the segmented UV curing module and the large-area curing module in sequence to complete the curing process, while monitoring the UV lamp power, temperature and working life. S7. Inspection and unloading: After curing, the qualified and unqualified products are separated and unloaded after visual inspection and height inspection. S8. Data Upload: The MES data interaction unit binds and uploads material information, process parameters, equipment operating status, and test results to achieve full traceability of the product production process. Its characteristic is that it further includes: S9. Process self-optimization step: The process parameter self-optimization unit dynamically adjusts the dispensing pressure and curing power of subsequent batches based on the test results of the current batch. S10, Digital Twin Monitoring Steps: The digital twin mapping unit simulates the equipment's operating status in real time, predicts and warns of potential anomalies; S11, AI re-inspection step: The AI ​​vision re-judgment unit performs random re-inspection on products that have been determined to be qualified by the inspection and unloading module to confirm that there are no minor defects. S12, Adaptive Changeover Step: When the barcode scan identifies a new product model, the adaptive changeover unit automatically loads the corresponding formula and switches the motion trajectory.

7. The high-precision aperture automated assembly process according to claim 6, characterized in that, In step S9, the process parameter self-optimization unit adopts a Bayesian optimization algorithm, with the bonding accuracy and the hardness of the cured adhesive layer as the objective functions, and iteratively optimizes to bring the bonding accuracy to within ±20μm.

8. The high-precision aperture automated assembly process according to claim 6, characterized in that, In step S10, the digital twin mapping unit maps the motion curves and vibration spectrum of each axis of the device in real time, and triggers an early warning when the vibration amplitude exceeds 30% of the historical baseline.

9. The high-precision aperture automated assembly process according to claim 6, characterized in that, In step S11, the AI ​​visual re-judgment unit uses a ResNet-50 convolutional neural network to classify defects in the detected image, with a classification accuracy of no less than 99.5%.

10. The high-precision aperture automated assembly process according to claim 6, characterized in that, In step S12, the adaptive type-changing unit completes the switching of all process parameters and motion trajectories within 30 seconds, and the first piece after type-changing enters the automatic production state.

Images