Femtosecond laser glue-film-free welding production line for photovoltaic glass of double-glass photovoltaic module

Abstract

The invention relates to the technical field of photovoltaic module manufacturing, in particular to a femtosecond laser glue-film-free welding process for photovoltaic glass of a double-glass photovoltaic module and a matched automatic production line. By combining automatic feeding and pretreatment, a multi-station linkage welding workstation, real-time quality detection and intelligent production line control, the problems of aging, high cost, low efficiency and the like of traditional adhesive film welding are solved. Femtosecond laser welding avoids use of a glue film, the welding strength and the assembly reliability are improved, efficient cooperation is achieved through an automatic production line, the cost is reduced, the production efficiency and the product quality stability are improved, and intelligent upgrading of photovoltaic assembly manufacturing is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic module manufacturing, and in particular to a femtosecond laser adhesive-free film welding process for photovoltaic glass of a double-glass photovoltaic module and a matching automated production line. Background Art

[0002] In the production of double-glass photovoltaic modules, traditional photovoltaic glass welding methods mostly rely on adhesive films for connection, which has numerous drawbacks. Over long-term use, adhesive films are susceptible to aging and degradation due to environmental factors such as light, temperature fluctuations, and humidity. This can lead to debonding at the welds, seriously impacting the module's service life and power generation efficiency. Furthermore, traditional welding processes require complex steps such as adhesive coating and lamination, resulting in low production efficiency and high adhesive film material costs, which increase production costs.

[0003] On the production line, existing double-glass photovoltaic module production equipment has a low level of automation, with loose integration between processes and frequent manual intervention. This results in difficulty improving production efficiency and unstable product quality. For example, the glass loading, welding, and testing processes lack effective automated coordination, which can easily lead to problems such as material transfer jams and untimely welding parameter adjustments, making it impossible to meet the demands of large-scale, high-quality production. Furthermore, existing production lines have limited means of monitoring the quality of the welding process, making it difficult to detect and correct welding defects in real time, increasing the defect rate. Summary of the Invention

[0004] The purpose of this invention is to provide a femtosecond laser adhesive-free film welding and automated production line for double-glass photovoltaic modules, aiming to achieve efficient and high-quality double-glass photovoltaic module production through innovative technology. It includes:

[0005] Femtosecond laser welding module, using multi-parameter coordinated control model E d =k·f α ·v -β ·τ γ Achieve precise control of energy density;

[0006] Automated loading and pre-processing system, integrated with visual positioning and correction model;

[0007] Multi-station linkage welding workstation;

[0008] Real-time quality detection and feedback system;

[0009] Intelligent production line control system.

[0010] Furthermore, in the multi-parameter coordinated regulation model:

[0011] E d is the energy density (J / cm 2), f is the pulse frequency (kHz), v is the scanning speed (mm / s), and τ is the pulse width (fs);

[0012] Material correction coefficient k value: borosilicate glass 1.2-1.8, soda-lime glass 1.5-2.0;

[0013] Empirical indexes α = 0.2-0.3, β = 0.1-0.2, γ = -0.2-0.1.

[0014] Furthermore, the automated loading and pretreatment system includes:

[0015] Robotic arm loading device, positioning accuracy ±0.1mm;

[0016] Visual positioning system, using deviation calculation model:

[0017] Where (x w ,y w ) is the ideal coordinate, (x c ,y c ) is the detection coordinate, M 2D is a 2D transformation matrix.

[0018] Furthermore, the correction control equation of the visual positioning system is:

[0019] Where K is the proportional control matrix, v x / v y , is the linear velocity correction, ω is the angular velocity correction.

[0020] Furthermore, the process parameters of the femtosecond laser welding module are:

[0021] Wavelength 1030nm, average power 100-150W, pulse width 500-1000fs;

[0022] When welding 3mm borosilicate glass, the pulse frequency is 10-15kHz and the energy density is 12-18J / cm 2 , scanning speed 60-100mm / s.

[0023] Furthermore, the multi-station linkage welding workstation:

[0024] It consists of ≥3 welding units, and the laser focus spot diameter of each unit is ≤50μm;

[0025] Adopting parallel welding strategy, the transmission speed between adjacent workstations is 0.5-2m / s, and the efficiency is increased by 3-5 times.

[0026] Furthermore, the real-time quality detection and feedback system:

[0027] Integrated optical detection frame rate ≥1000fps, ultrasonic detection frequency 5-10MHz, thermal imaging detection temperature measurement 0-1000℃;

[0028] The PID algorithm is used to dynamically adjust the laser parameters with an adjustment range of 10-30%.

[0029] Furthermore, the intelligent production line control system:

[0030] Adopt OPC UA protocol to realize data interaction, with delay ≤50ms;

[0031] Use LSTM neural network to predict equipment failure with an accuracy rate of ≥90%;

[0032] Optimize production rhythm based on big data and increase OEE to over 90%.

[0033] Furthermore, the femtosecond laser welding module, automated loading system, and detection system all communicate with an intelligent control system to achieve:

[0034] The welding parameters are automatically retrieved from the database according to the glass material;

[0035] Abnormal operating conditions trigger multi-level warnings and activate emergency plans.

[0036] Furthermore, the multi-parameter collaborative control model is linked with the visual positioning and correction model to ensure that the strength of the 3mm glass welding joint is ≥20MPa and the positioning accuracy is ≤±0.1mm.

[0037] The present invention has the following beneficial effects:

[0038] Improve welding quality and reliability: Femtosecond laser film-free welding technology avoids problems such as film aging. The optimized welding process ensures the high quality of the welded joints and improves the overall reliability and service life of double-glass photovoltaic modules.

[0039] Improve production efficiency: The coordinated work of the automated loading and pretreatment system, the multi-station femtosecond laser welding workstation, and the intelligent production line control system has achieved efficient automation of the production process, significantly shortened the production cycle, and improved production efficiency. Compared with traditional production lines, production efficiency can be increased by more than 50%.

[0040] Reduce production costs: reduce the use of film and manual intervention, optimize the production process, improve equipment utilization, reduce raw material costs and labor costs, and improve the economic benefits of the enterprise.

[0041] Ensure product quality stability: The real-time online quality detection and feedback system and the intelligent production line control system can promptly detect and resolve quality problems in the production process, ensure the consistency and stability of product quality, and improve product market competitiveness.

[0042] Promote the intelligent upgrading of the industry: The automated production line of this invention adopts advanced industrial Internet and artificial intelligence technologies, providing a demonstration for the intelligent development of the photovoltaic module manufacturing industry, and helping to promote technological progress and industrial upgrading of the entire industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 Femtosecond laser film-free welding overall process flow chart;

[0044] Figure 2 Flowchart of automated pre-processing system;

[0045] Figure 3 Femtosecond laser welding parameter control flow chart;

[0046] Figure 4 Intelligent production line control flow chart. DETAILED DESCRIPTION

[0047] The following embodiments of the present invention are described in further detail with reference to the accompanying drawings and examples. The following examples are used to illustrate the present invention but are not intended to limit the scope of the present invention.

[0048] Example 1

[0049] Industrial Application of Femtosecond Laser Adhesive-Free Film Welding Technology in a 1.2GW Double-Glass Photovoltaic Module Production Line I. Project Background and Technical Transformation Goals

[0050] A photovoltaic company's original 1.2GW double-glass module production line used traditional EVA film welding technology, which had the following pain points:

[0051] The cost of adhesive film accounts for 9.2% of the component material cost, and the annual consumables cost reaches 20 million yuan;

[0052] Film aging causes the light transmittance of the modules to decrease by an average of 2.5% per year, significantly reducing power generation efficiency.

[0053] The welding yield rate is only 88.3%, and the annual loss of defective products is about 3.5 million yuan;

[0054] The production line cycle is 180 seconds per piece, which cannot meet the demand for production expansion.

[0055] Based on the technology of this invention, the company renovated the No. 3 production line with the following goals:

[0056] Realize the full process application of film-free welding process, and the welding joint strength is ≥20MPa;

[0057] The production line cycle was shortened to 90 seconds per piece, and the production capacity was increased by 100%;

[0058] The annual cost of film consumables is reduced to 0, and the overall production cost is reduced by more than 10%;

[0059] The welding yield rate has been increased to over 98%, and the product life has been extended to 30 years.

[0060] 2. Implementation steps of femtosecond laser film-free welding process

[0061] (1) Photovoltaic glass loading and visual positioning

[0062] 1.Automated feeding system configuration

[0063] Robotic arm: ABB IRB 6700 model, load 100kg, positioning accuracy ±0.05mm, equipped with vacuum suction cup (suction force ≥20kPa), adsorption contact area ≥200cm when grasping 3mm borosilicate glass 2 , ensuring no slipping during grasping;

[0064] Conveyor belt: driven by a servo motor, with a speed of 0.8m / min. The surface is covered with a silicone anti-slip layer (hardness Shore A 60) to prevent glass scratches.

[0065] Loading efficiency: The overall loading time from the glass storage rack to the pre-treatment station is controlled within 8 seconds, meeting the 90-second / piece cycle requirement.

[0066] 2. Implementation of the visual positioning system

[0067] Hardware configuration:

[0068] Industrial camera: Basler ace 640-120gm, resolution 1280 × 960, frame rate 120 fps;

[0069] Lens: Computar M0814-MP2, focal length 8mm, depth of field 5-100mm;

[0070] Light source: Ring white LED light source, color temperature 5500K, light intensity ≥1000lux.

[0071] Positioning algorithm application:

[0072] 1. Image acquisition: The camera collects three sets of images of the glass on the conveyor belt and calculates the average position deviation;

[0073] 2. Deviation calculation: Using the model in claim 3: Δx = x w ·x c ·M 2D , Δy=y w -y c ·M 2D

[0074] The transformation matrix M between the world coordinate system and the image coordinate system is 2D It is obtained by Zhang’s calibration method, with a calibration error of ≤0.1 pixel;

[0075] 3. Correction control: According to the correction equation of claim 4:

[0076] When the glass offset Δx = 0.5 mm, Δy = 0.3 mm, and the angle deviation θ = 0.8° is detected, the conveyor belt speed correction v is calculated. x =0.3mm / s, v y =0.18mm / s, rotation angular velocity ω=0.12° / s, ensuring positioning accuracy ≤±0.1mm.

[0077] (2) Welding interface pretreatment process

[0078] 1. Plasma physical cleaning

[0079] Equipment parameters:

[0080] Plasma generator: Pfeiffer Vacuum Plasmaline 300, power 300 W, frequency 40 kHz;

[0081] Gas parameters: Argon (99.99% purity) mixed with oxygen in a volume ratio of 3:1, with a total flow rate of 20 sccm;

[0082] Cleaning time: 4 minutes, glass transmission speed 0.5m / min;

[0083] Cleaning effect: The glass surface roughness Ra is reduced from 1.0μm to 0.3μm, the contact angle is reduced from 68° to 28°, and the surface C element content detected by XPS is reduced from 2.1% to below 0.4%.

[0084] 2. Chemical activation treatment

[0085] Activating solution formula:

[0086] Hydrofluoric acid (HF) concentration 5%, nitric acid (HNO3) concentration 8%, prepared in deionized water, temperature 40℃;

[0087] Soaking time: 3 minutes, followed by rinsing with deionized water (conductivity <1 μS / cm) for 30 seconds, and drying with hot air (temperature 80°C, wind speed 10 m / s);

[0088] Surface modification effect: The density of Si-OH active groups on the glass surface increases from 2.3×10 14 / cm 2 Increased to 6.2×10 14 / cm 2 , providing a good chemical bonding basis for femtosecond laser welding.

[0089] (3) Femtosecond laser multi-parameter collaborative welding

[0090] 1. Laser parameter database call and optimization

[0091] For 3mm borosilicate glass (transmittance 91.5%):

[0092] οRetrieve basic parameters from the database:

[0093] Pulse frequency f = 12kHz, energy density E d =15J / cm 2 , scanning speed v = 80 mm / s, pulse width τ = 800 fs;

[0094] οComputational verification of multi-parameter coordinated regulation model:

[0095] E d =k·f α ·v β ·τ γ =1.6×12 0.25 ×80 0.15 ×800 0.1 ≈15.2J / cm 2

[0096] The material correction coefficient k = 1.6, the empirical exponents α = 0.25, β = 0.15, and γ = -0.1, which are consistent with the database parameters.

[0097] For 2.8mm soda-lime glass (transmittance 89.2%)

[0098] οAdjustment parameters: f=10kHz, E d =18J / cm 2 , v=60mm / s, τ=1000fs, k=1.8;

[0099] Calculated Ed = 18.1 J / cm 2 , meeting the higher melting energy requirements of soda-lime glass.

[0100] 2. Implementation of welding process

[0101] Device configuration:

[0102] Femtosecond laser: TRUMPF TruMicro 7050, wavelength 1030 nm, average power 150 W, repetition rate 1-100 kHz;

[0103] Galvanometer scanning system: SCANLAB HurryScan 14, positioning accuracy ±5μm, maximum scanning speed 7000mm / s;

[0104] Focusing system: focal length 160mm, focused spot diameter 30μm, energy uniformity ≥90%;

[0105] Coaxial blowing: nitrogen (purity 99.99%), pressure 0.2MPa, to prevent oxidation of the welding area.

[0106] Welding process implementation:

[0107] Scanning path: Zigzag scanning, line spacing 50μm, overlap rate 70%, welding speed 80mm / s (borosilicate glass);

[0108] Line energy control: borosilicate glass 0.6J / mm, soda-lime glass 0.8J / mm, real-time adjustment through laser energy feedback system (accuracy ±2%);

[0109] Welding time: The welding time of a single piece of 1.6m×1.2m glass is controlled within 45 seconds, meeting the production line's beat requirements.

[0110] (4) Deployment of multi-station linkage welding workstations

[0111] 1. Workstation layout and equipment configuration

[0112] Workstation design:

[0113] Number of welding units: 5, U-shaped layout, suitable for glass size of 1.6m×1.2m;

[0114] Unit spacing: 2.5m, ensuring no interference in glass transmission;

[0115] Transmission device: high-precision linear motor conveyor, positioning accuracy ±0.1mm, transmission speed 1m / s.

[0116] Single welding unit configuration:

[0117] Femtosecond laser head: same as above, focusing spot 30μm;

[0118] Vision alignment system: KEYENCE VR-3000, resolution 5μm, alignment time ≤ 2 seconds;

[0119] Pneumatic clamping device: clamping force 5-10N, ensuring no displacement during glass welding.

[0120] 2. Collaborative working mechanism

[0121] Task allocation algorithm: The intelligent control system automatically allocates welding units according to the size of the glass. For example, a 1.6m×1.2m glass is welded simultaneously by three units, with the edge areas being handled by the units on both sides and the center area being covered by the middle unit.

[0122] Parameter synchronization: Each unit synchronizes parameters in real time through the OPC UA protocol to ensure that the difference in welding energy density is ≤3%;

[0123] Production efficiency: Multi-station linkage increases welding efficiency by 3.8 times, shortening the welding time of a single piece of glass from 170 seconds (traditional single station) to 45 seconds, meeting the 90-second / piece cycle requirement.

[0124] (5) Real-time online quality detection and feedback

[0125] 1. Detection system configuration

[0126] Optical inspection:

[0127] Camera: Phantom Miro M110, frame rate 1000fps, resolution 1280×800;

[0128] Image algorithm: YOLOv5 model based on deep learning, detecting weld width (accuracy ±0.05mm) and flatness (deviation ±0.1mm);

[0129] Detection speed: 5 seconds per piece of glass, defect recognition rate ≥99%.

[0130] Ultrasonic testing:

[0131] Probe: Olympus V318-SU, frequency 5 MHz, resolution 0.1 mm;

[0132] Scanning mode: linear scanning, covering 100% of the weld area, detecting internal defects such as pores and cracks;

[0133] Detection sensitivity: Can identify pores with a diameter of ≥0.2mm.

[0134] Thermal imaging detection:

[0135] Thermal imager: FLIR A655sc, temperature measurement range -20°C-1500°C, accuracy ±2°C;

[0136] Detection content: Temperature distribution in welding area, ensuring melting uniformity, and immediate alarm when temperature is abnormal.

[0137] 2. Feedback Control Case

[0138] 1. The thermal imaging system sends a temperature anomaly signal to the PLC;

[0139] 2. The control system calculates the parameter adjustment according to the PID algorithm:

[0140] Where the deviation e = 30°C, the calculated laser energy attenuation is 15%, while the scanning speed is increased to 65 mm / s;

[0141] 3. After adjustment, the temperature dropped back to the 650°C threshold within 2 seconds, and the strength of the weld joint did not change significantly (maintained at 18.7 MPa).

[0142] 3. Implementation of intelligent production line control system

[0143] 1. System architecture construction

[0144] Hardware configuration:

[0145] Industrial computer: Advantech ARK-3500, CPU i7-11700, 16GB memory, 512GB SSD storage;

[0146] Data acquisition module: Advantech ADAM-4000 series, sampling frequency 100Hz, precision 16 bits;

[0147] Communication network: Gigabit Ethernet, switch Huawei S5720, data transmission delay ≤ 30ms.

[0148] Software system:

[0149] Operating system: Windows 10 IoT Enterprise;

[0150] Control software: developed based on Python, integrating OPC UA server, data processing algorithm, and device control module;

[0151] Database: InfluxDB, stores production data (storage period ≥ 2 years).

[0152] 2. Data-driven production optimization

[0153] Equipment Failure Prediction: Using an LSTM neural network model with input parameters including laser power fluctuations, motor current, and bearing temperature, the system can predict equipment failures 72 hours in advance with 92% accuracy. For example, if the system predicts abnormal galvanometer motor bearing temperature, it automatically schedules maintenance to avoid downtime losses.

[0154] Production cadence optimization: Based on historical production data (over 100,000 pieces), a reinforcement learning algorithm was used to optimize the tact time of each workstation, increasing the production line's overall equipment efficiency (OEE) from 67% before the transformation to 92%. Specifically, the following performance was achieved:

[0155] The efficiency of the loading station increased by 35%, the efficiency of the welding station increased by 42%, and the efficiency of the inspection station increased by 28%.

[0156] 4. Test Verification and Long-term Reliability Evaluation

[0157] 1. Welding joint performance test

[0158]

[0159]

[0160] 2. Accelerated aging test results

[0161] Damp heat test (85℃ / 85%RH, 1000h):

[0162] Light transmittance decreased by 0.7% (conventional process decreased by 5.3%);

[0163] The weld joint strength retention rate is 98.5% (82.1% for traditional process).

[0164] Thermal cycle test (-40℃~85℃, 1000 times):

[0165] Femtosecond laser welded components are crack-free and have a stress change of ≤4%;

[0166] The traditional film component had five cracks and a stress attenuation of 25%.

[0167] 3. Power generation comparison

[0168] Deployment of 2MW double-glass modules in a photovoltaic power station in Qinghai:

[0169] Femtosecond laser welding components: first-year power generation of 2.92 million kWh, with an average annual attenuation rate of 0.32%;

[0170] Traditional film modules: First-year power generation: 2.78 million kWh, with an average annual degradation rate of 2.2%;

[0171] Over a 25-year period, the power generation of femtosecond laser components increased by approximately 16.5%.

[0172] V. Economic Benefit Analysis

[0173]

[0174] VI. Implementation and Verification of Technological Innovations

[0175] The multi-parameter collaborative control model was matched with the database through model calculation. The standard deviation of borosilicate glass welding strength was reduced from 1.9MPa to 0.4MPa, and the parameter debugging time was shortened from the traditional 2 weeks to 1 day, verifying the accuracy and efficiency of the model.

[0176] The positioning accuracy of the visual positioning and correction system has been improved from ±1mm in the traditional process to ±0.1mm, meeting the micron-level requirements of femtosecond laser welding. The offset of the welding joint is ≤0.3mm, which is a significant improvement compared with the traditional process (offset ≤1.5mm).

[0177] The coordinated operation of the five welding units in the multi-station linkage workstation increased production efficiency by 3.8 times, reaching a production capacity of 2.4GW / year, and increasing equipment utilization from 55% to 89%, verifying the industrial feasibility of the parallel welding strategy.

[0178] The LSTM fault prediction model of the intelligent control system successfully warned of 23 equipment anomalies, avoiding downtime losses of approximately 1.5 million yuan; production rhythm optimization stabilized the production line beat at 90 seconds per piece, with fluctuations ≤5 seconds, realizing adaptive production.

[0179] This invention adopts femtosecond laser adhesive-free film welding technology, which takes advantage of the ultra-short pulse and high peak power of femtosecond laser to locally heat and melt photovoltaic glass in a very short time, thus achieving direct welding between glasses. By precisely controlling the pulse frequency, energy density, scanning speed and other parameters of the femtosecond laser, personalized welding solutions can be developed for photovoltaic glass of different thicknesses and materials. For example, for ordinary photovoltaic glass with a thickness of 3mm, the pulse frequency is set to 10-15kHz and the energy density is controlled at 12-18J / cm 2 , the scanning speed is set to 60-100mm / s to ensure that the glass melts evenly, forming a firm and well-sealed welding joint, avoiding the connection failure caused by the film problem in traditional welding.

[0180] An automated loading and pretreatment system was designed to automatically load, position, and pre-treat photovoltaic glass. The system includes a robotic loading device, a visual positioning system, and plasma cleaning equipment. The robotic loading device automatically grabs the photovoltaic glass and places it in the designated position according to the production rhythm. The visual positioning system uses a high-definition camera to capture and analyze images of the glass, precisely adjusting its position to ensure welding accuracy. The plasma cleaning equipment cleans the glass welding surface, removing impurities such as oil, dust, and oxides, enhancing the glass's surface activity and providing a good foundation for subsequent welding. The entire process requires no human intervention, improving production efficiency and loading accuracy.

[0181] A multi-station linkage femtosecond laser welding workstation is constructed. This workstation consists of multiple femtosecond laser welding units, each of which can weld multiple pieces of photovoltaic glass simultaneously and automatically adjust welding parameters according to the glass size and welding requirements. Automatic transmission devices are installed between adjacent stations to achieve fast and accurate transmission of glass. By optimizing the welding path and timing, the various welding units can work together to greatly improve welding efficiency. For example, when welding large double-glass photovoltaic modules, multiple welding units can weld different areas of the glass at the same time, which can increase production efficiency by 3-5 times compared to a single welding unit, while ensuring the consistency of welding quality.

[0182] A real-time online quality inspection and feedback system has been established to monitor welding quality in real time during the welding process. This system integrates multiple inspection technologies, including optical, ultrasonic, and thermal imaging. Optical inspection uses a high-speed camera to capture images of the weld area and utilizes image recognition algorithms to assess the weld's appearance, such as weld width and flatness. Ultrasonic inspection detects internal defects such as pores and cracks in the weld joint. Thermal imaging monitors the temperature distribution during welding to determine weld uniformity. If quality issues are detected, the system immediately sends a signal back to the femtosecond laser welding equipment, automatically adjusting welding parameters and correcting defects in real time to ensure product quality.

[0183] Develop an intelligent production line control system, leveraging the Industrial Internet and artificial intelligence technologies to provide unified management and control of the entire production line. This system monitors the operating status of each production line device in real time and collects production data, such as equipment operating parameters, product quality data, and production progress. Using big data analytics and machine learning algorithms, this system conducts in-depth mining and analysis of production data to predict equipment failures and optimize production parameters and processes. For example, based on historical production data and current production status, it automatically adjusts the production rhythm of each process, achieving adaptive production line operation, improving production efficiency and equipment utilization, while reducing labor and management costs.

[0184] The embodiments of the present invention are presented for purposes of illustration and description and are not intended to be exhaustive or to limit the invention to the disclosed forms. Many modifications and variations will be apparent to those skilled in the art. The embodiments are chosen and described in order to better illustrate the principles of the invention and its practical application and to enable those skilled in the art to understand the invention and design various embodiments with various modifications as suited for specific applications.

Claims

1. A femtosecond laser adhesive-free film welding and automated production line for double-glass photovoltaic modules, characterized in that: include: Femtosecond laser welding module, using multi-parameter coordinated control model E d =k·f a ·v -β ·τ γ Achieve precise control of energy density; Automated loading and pre-processing system, integrated with visual positioning and correction model; Multi-station linkage welding workstation; Real-time quality detection and feedback system; Intelligent production line control system.

2. The production line according to claim 1, characterized in that: In the multi-parameter coordinated regulation model: E d is the energy density (J / cm 2 ), f is the pulse frequency (kHz), v is the scanning speed (mm / s), τ is the pulse width (fs); the material correction coefficient k is 1.2-1.8 for borosilicate glass and 1.5-2.0 for soda-lime glass; Empirical indexes α = 0.2-0.3, β = 0.1-0.2, γ = -0.2-0.

1.

3. The production line according to claim 1, characterized in that The automatic loading and pretreatment system includes: Robotic arm loading device, positioning accuracy ±0.1mm; The visual positioning system uses a deviation calculation model: Where (x w ,y w ) is the ideal coordinate, (x c ,y c ) is the detection coordinate, M 2D is a 2D transformation matrix.

4. The production line according to claim 3, characterized in that: The correction control equation of the visual positioning system is: Where K is the proportional control matrix, v x / v y is the linear velocity correction, and ω is the angular velocity correction.

5. The production line according to claim 1, characterized in that: The process parameters of the femtosecond laser welding module are: Wavelength 1030nm, average power 100-150W, pulse width 500-1000fs; When welding 3mm borosilicate glass, the pulse frequency is 10-15kHz and the energy density is 12-18J / cm 2 , scanning speed 60-100mm / s.

6. The production line according to claim 1, characterized in that: The multi-station linkage welding workstation: It consists of ≥3 welding units, and the laser focus spot diameter of each unit is ≤50μm; Adopting parallel welding strategy, the transmission speed between adjacent workstations is 0.5-2m / s, and the efficiency is increased by 3-5 times.

7. The production line according to claim 1, characterized in that: The real-time quality detection and feedback system: Integrated optical detection frame rate ≥1000fps, ultrasonic detection frequency 5-10MHz, thermal imaging detection temperature measurement 0-1000℃; The PID algorithm is used to dynamically adjust the laser parameters with an adjustment range of 10-30%.

8. The production line according to claim 1, characterized in that: The intelligent production line control system: Adopt OPC UA protocol to realize data interaction, with delay ≤50ms; Use LSTM neural network to predict equipment failure with an accuracy rate of ≥90%; Optimize production rhythm based on big data and increase OEE to over 90%.

9. The production line according to claim 1, characterized in that: The femtosecond laser welding module, automated loading system, and detection system all communicate with the intelligent control system to achieve: The welding parameters are automatically retrieved from the database according to the glass material; Abnormal operating conditions trigger multi-level warnings and activate emergency plans.

10. The production line according to any one of claims 1 to 9, characterized in that: The multi-parameter collaborative control model is linked with the visual positioning and correction model to ensure that the strength of the 3mm glass welding joint is ≥20MPa and the positioning accuracy is ≤±0.1mm.

Images