Femtosecond adhesive-film-free edge adaptive welding system and welding method

Abstract

The invention discloses a femtosecond glue-film-free edge adaptive welding system and method, and relates to the technical field of laser welding. According to the system, through real-time correction of edge curl, enhancement of energy absorption of the light-transmitting film, rapid adaptation of small-batch multi-specification parameters and intelligent detection of welding edge quality, the problems of welding dislocation caused by edge curl in existing femtosecond glue-free film welding, high energy reflectivity of the light-transmitting film and low efficiency of small-batch multi-specification adaptation are solved. Through the combination of edge sensing, energy regulation and control and a parameter adaptation algorithm, accurate correction of curled edges, efficient welding of the light-transmitting film and small-batch rapid adaptation are achieved, and the edge welding precision, the welding strength of the light-transmitting film and the small-batch production efficiency are remarkably improved.

Description

TECHNICAL FIELD

[0001] The application discloses a femtosecond glue-free film edge adaptive welding system and a welding method, and relates to the technical field of laser welding. BACKGROUND

[0002] The existing femtosecond glue-free film welding technology has the following specific problems:

[0003] Edge curling leads to welding misalignment: after the film material is cut, the edge is prone to curling (the curling height is greater than or equal to 50 microns, and the curling angle is greater than or equal to 3 degrees), such as the edge of the aluminum film after cutting the lithium battery tab and the edge of the PET film of a flexible screen. The existing equipment cannot identify such curling, and the actual joint deviates from the visual positioning joint by more than ±30 microns during welding, leading to the deviation of the welding area from the effective edge and the occurrence of edge cracking (the cracking rate is more than 20%).

[0004] High energy reflectivity of light-transmitting film: for film materials with a light transmittance of greater than or equal to 70% (such as PC film and PMMA film), the laser energy reflectivity is more than 60%. The existing equipment uses a single energy output, and more than 30% of the energy needs to be increased to achieve the welding strength, resulting in ablation marks on the surface of the film material (the ablation area accounts for more than 15%), or the welding strength is less than 8 N / cm due to insufficient energy.

[0005] Low efficiency of small-batch multi-specification adaptation: in the production scene of small-batch multi-specification (such as custom film welding of medical consumables), when the film material specifications (thickness difference greater than or equal to 10 microns, material replacement) are changed, manual parameter adjustment is required, and the time consumed for single adjustment is more than 15 minutes. The first-time adjustment qualification rate is less than 60%, resulting in low production efficiency. SUMMARY

[0006] The application aims to solve the problems of edge curling leading to welding misalignment, high energy reflectivity of light-transmitting film and low efficiency of small-batch multi-specification adaptation in the existing femtosecond glue-free film welding technology.

[0007] A femtosecond glue-free film edge adaptive welding method, comprising:

[0008] (1) Collecting the height, angle and curvature data of the edge curling of the film material by a microscopic vision sensor, generating correction parameters and performing correction;

[0009] (2) Obtaining the light transmittance, thickness and refractive index of the light-transmitting film, and reducing the energy reflectivity by an energy enhancement algorithm;

[0010] (3) Inputting the film material specification parameters, calling an associated database, and calculating the initial welding parameters by the formula

[0011] P=P0×(1+α×Δh+β×M+γ×C)

[0012] (4) Perform welding, detect edge welding quality by laser profiler and ultrasonic detector;

[0013] (5) According to the quality detection result, optimize the welding parameter and correction parameter.

[0014] Further, the energy enhancement algorithm in step (2) includes: calculating the energy enhancement coefficient by the formula The reflectivity is controlled below 20% by combining the surface micro-texture pretreatment.

[0015] Further, the correction parameter in step (1) includes the correction force (0.1-1N) and the action time, and the edge curling height after correction is ≤10μm and the angle is ≤0.5°.

[0016] Further, the quality detection in step (4) includes the edge flatness (deviation ≤5μm) and the internal bonding strength (≥12N / cm) detection.

[0017] Further, the film material specification parameter includes the thickness (10-100μm), the material type (metal film / transparent film / complex film) and the edge curling coefficient (0-1).

[0018] Further, a femtosecond glue-free film edge adaptive welding system is provided, which includes:

[0019] (1) Edge curling correction module: integrated with microscopic vision sensor and micro-force actuator, used for collecting curling data and performing correction;

[0020] (2) Transparent energy regulation module: obtain the transparent film parameter, calculate the energy enhancement coefficient by the formula To reduce the reflectivity;

[0021] (3) Parameter rapid adaptation module: call the film material specification-parameter database, generate the initial welding parameter by the formula P=P0×(1+α×Δh+β×M+γ×C);

[0022] (4) Edge quality detection module: detect the welding edge quality by laser profiler and ultrasonic detector;

[0023] (5) Parameter optimization module: according to the quality detection result, optimize the welding and correction parameters.

[0024] Further, the resolution of the microscopic vision sensor of the edge curling correction module is ≥1μm, and the micro-force actuator force control precision is ±0.05N.

[0025] Further, the transparent energy regulation module includes a laser micro-etching unit, and the etching precision is ±2μm, which can generate micron-level texture (depth 1-5μm).

[0026] Furthermore, the membrane material specification-parameter database stores the reference parameters (energy density, welding speed) and correction coefficients for at least 50 types of membrane materials.

[0027] Furthermore, it also includes an execution module: receiving welding parameters and correction parameters, controlling laser output (energy accuracy ±2%) and the action of the micro-force actuator.

[0028] Beneficial effects:

[0029] The edge curling correction accuracy is improved, the welding misalignment deviation is controlled within ±10μm, the edge cracking rate is reduced, and the edge welding reliability of film materials such as lithium battery tabs is significantly improved.

[0030] The energy reflectivity of the light-transmitting film is reduced, so the welding strength can be achieved without increasing the energy. The peel strength of light-transmitting films such as PC film is improved after welding, and the surface ablation area ratio is controlled within 2%.

[0031] The adaptation time for small-batch, multi-specification parameters has been shortened to less than 1 minute, the first-time debugging pass rate has been increased to over 90%, and the production efficiency of scenarios such as customized membranes for medical consumables has been improved.

[0032] Through closed-loop detection optimization, the system's adaptability to changes in membrane edge condition, light transmission characteristics, and specifications is enhanced, making it suitable for welding requirements of more than 70 types of membrane materials. Attached image description:

[0033] Appendix Figure 1 Method principle flowchart. Detailed implementation method:

[0034] Example 1:

[0035] This invention provides a femtosecond adhesive-free edge-adaptive welding system and method, achieving technological breakthroughs through the following four innovative aspects:

[0036] Real-time edge curling correction module: It integrates a microscopic vision sensor (resolution ≥1μm) and a micro-force actuator to collect data on the height, angle and curvature of edge curling; it generates correction force parameters (force value 0.1-1N) and application time through a curling correction algorithm, drives the micro-force actuator to press the curled edge, and the curling height after correction is ≤10μm, providing a smooth joint for welding.

[0037] Algorithm for enhancing energy absorption of transparent film: Based on the transmittance (T), thickness (h), and refractive index (n) of the transparent film, the energy enhancement coefficient is calculated using formula (1). Combined with surface micro-texture pretreatment (laser micro-etching), the reflectivity is reduced to below 20%.

[0038]

[0039] (where, K is the energy enhancement coefficient, T is the light transmittance (0-1), h is the film thickness (pm), n is the refractive index, k0 is the basic correction coefficient)

[0040] Small batch multi-specification parameter rapid adaptation model: build a film specification-parameter correlation database, input the thickness, material, and edge state parameters of the new specification film, calculate the initial parameters through formula (2), and generate the optimal parameters within 10 seconds combined with the first welding quality feedback:

[0041] P = P0 x (1 + a x Ah + b x M + g x C)

[0042] (where, P is the new specification parameter, P0 is the baseline parameter, a is the thickness correction coefficient, Ah is the thickness deviation, b is the material correction coefficient, M is the material type coefficient, g is the edge correction coefficient, and C is the edge curling coefficient)

[0043] Welding edge quality intelligent detection system: after welding, the flatness of the edge welding (precision ±1 pm) is detected by a laser profiler, and the internal bonding strength is detected by an ultrasonic detector. Compared with the preset standard, the unqualified products are automatically marked and the parameter optimization suggestions are fed back, forming a "curling correction-energy regulation-parameter adaptation-quality detection" closed loop.

[0044] Lithium battery tab aluminum film-composite film edge welding scene (edge curling correction)

[0045] Step 1: Place the aluminum film (thickness 30 pm) with edge curling (height 80 pm, angle 5°) and composite film on the welding platform, and the edge curling real-time correction module collects the curling data;

[0046] Step 2: The correction algorithm generates parameters (correction force 0.5 N, action time 0.5 s), and the micro-force actuator presses the edge. After correction, the curling height is 9 pm;

[0047] Step 3: Call the parameter adaptation model to generate the initial energy parameter 1.6 J / cm 2 based on the film specification (aluminum film thickness 30 pm).

[0048] Step 4: Start welding, and the welding edge quality intelligent detection system monitors in real time.

[0049] Step 5: After welding, the detection shows that the welding misalignment deviation is ±8 pm, the edge cracking rate is 2%, and the peel strength is 21 N / cm.

[0050] Comparison with prior art: the existing equipment has no curling correction, the misalignment deviation is ±35 pm, the cracking rate is 25%, and the strength is only 9 N / cm; this embodiment reduces the cracking rate by 92% and increases the strength by 133% through curling correction.

[0051] Example 2: Transparent Display Screen PC Film-ITO Film Welding Scenarios (Light Transmitting Film Energy Regulation)

[0052] Step 1: Place a PC film with 80% transmittance (thickness 25μm, refractive index 1.58) and an ITO film on the platform. The energy absorption enhancement algorithm of the transmittance film calculates the energy enhancement coefficient K = 1.4 (basic correction coefficient k0 = 0.8) using formula (1).

[0053] Step 2: Trigger surface microtexturing pretreatment (laser micro-etching energy 0.2 J / cm) 2 After treatment, the reflectivity decreased to 18%.

[0054] Step 3: Parameter adaptation model generates welding energy of 1.2 J / cm². 2 (No additional energy required);

[0055] Step 4: Start welding, and the quality inspection system monitors the surface condition;

[0056] Step 5: Welding completed, the ablation area of ​​the PC film surface is 1.5%, and the peel strength is 19 N / cm.

[0057] Compared to existing technologies: Existing equipment does not handle reflection and requires 1.8 J / cm² of energy. 2 The ablation area was 20% and the strength was only 7 N / cm; in this embodiment, through energy regulation, the ablation area was reduced by 92.5% and the strength was increased by 171%.

[0058] Example 3: Customized PE film-silicone film welding scenario for medical consumables (small batch adaptation)

[0059] Step 1: Replace the PE film and silicone film with a thickness of 20μm, and quickly adapt the model to the database for small batches with multiple specifications and parameters;

[0060] Step 2: Input the new specification parameters (thickness 20μm, material PE-silicone), and calculate the initial parameters (reference parameter 1.0J / cm) using formula (2). 2 Thickness deviation -5μm, generating P = 0.9J / cm 2 );

[0061] Step 3: After the first weld, the strength is tested at 15 N / cm (standard ≥ 12 N / cm), and the quality system feedback indicates that no adjustment is needed;

[0062] Step 4: Batch welding of 50 pieces, total adaptation time of 50 seconds (including the first welding);

[0063] Step 5: Batch testing pass rate 98%, no secondary debugging required.

[0064] Example 2

[0065] (I) Algorithm Principle

[0066] In femtosecond adhesive-free edge-adaptive welding systems, real-time edge curling correction algorithms play a crucial role. They effectively address the issue of material edge curling during welding, ensuring high-quality welds. This algorithm primarily relies on the collaborative operation of an integrated microscopic vision sensor and a micro-force actuator.

[0067] The microscopic vision sensor, acting as the "eye" of the entire correction system, possesses extremely high resolution, enabling it to accurately acquire key data on the edge curling of the welding material, including height, angle, and curvature. Based on high-precision optical imaging technology, it converts the acquired edge images into digital signals and, through complex image processing algorithms, precisely identifies subtle features of edge curling, thereby obtaining accurate curling data. This data forms the basis for subsequent analysis and processing by the correction algorithm.

[0068] Micro-force actuators are the direct executors for edge curling correction. Based on instructions generated by the correction algorithm, they apply precise forces to the curled edges. Micro-force actuators are typically manufactured using advanced microelectromechanical systems (MEMS) technology, possessing high-precision force control capabilities. They can achieve precise adjustment within an extremely small force range, generally controllable between 0.1 and 1 N, to meet the correction needs of different materials and degrees of curling.

[0069] The correction algorithm is the core of the entire real-time edge curling correction module. It generates corresponding correction force parameters and application times by deeply analyzing and calculating the curling data collected by microscopic vision sensors. The algorithm's design is based on a thorough understanding of material mechanical properties and edge curling behavior, employing a series of complex mathematical models and intelligent algorithms. For example, finite element analysis is used to simulate the deformation of the material under different correction forces. By establishing a mechanical model of the material, the relationship between parameters such as the elastic modulus and yield strength and the correction force is analyzed to determine the optimal correction force parameters. Simultaneously, machine learning algorithms are combined with extensive welding experimental data for learning and training. This allows the algorithm to automatically identify different curling patterns and predict the optimal correction parameters based on historical and real-time data, achieving intelligent and adaptive correction processes.

[0070] In actual operation, after the microscopic vision sensor acquires edge curling data, it transmits this data to the correction algorithm module in real time. The correction algorithm module quickly processes the data, calculating appropriate correction force parameters (force value 0.1-1N) and application time based on a preset mathematical model and machine learning algorithm. These parameters are then sent to the micro-force actuator, which, according to the received instructions, precisely applies a force of appropriate magnitude and duration to the curled edge, effectively correcting it. After correction, the edge curling height can be controlled within a very small range, generally reduced to ≤10μm, providing a smooth joint for subsequent welding processes and greatly improving the quality and stability of the weld.

[0071] (II) Efficiency Enhancement Analysis

[0072] Compared to traditional welding processes, real-time edge curling correction algorithms demonstrate significant advantages in improving welding quality, particularly in reducing welding misalignment and cracking rates, and enhancing welding strength. Taking the edge welding of the aluminum film and composite film on lithium-ion battery tabs as an example, traditional welding equipment, lacking an effective curling correction mechanism, often encounters severe welding problems when welding aluminum films and composite films with curled edges. Welding misalignment deviations are large, typically reaching ±35μm, leading to inaccurate connection positions between the tab and the composite film, affecting the battery's electrical performance. Simultaneously, the cracking rate is also high, reaching 25%, posing significant safety hazards during battery use and making it prone to open circuits and other malfunctions. Furthermore, the welding strength is only 9N / cm, which cannot meet the requirements of lithium-ion batteries under complex operating conditions.

[0073] The femtosecond adhesive-free edge-adaptive welding system, employing a real-time edge curling correction algorithm, demonstrated superior performance when handling the same welding tasks. By acquiring edge curling data in real time and performing precise correction, the welding misalignment deviation was significantly reduced to ±8μm, ensuring the connection position accuracy between the tab and the composite film and effectively improving the electrical performance stability of the battery. The cracking rate was significantly reduced to only 2%, greatly improving the safety and reliability of the battery. The welding strength was further increased to 21N / cm, a 133% improvement compared to traditional welding processes, enabling the lithium battery to withstand greater current and mechanical stress, thus extending the battery's lifespan.

[0074] In other similar welding scenarios, such as the welding of electronic components and precision mechanical parts, the real-time edge curling correction algorithm can also play an important role in effectively solving welding quality problems caused by edge curling, improving product qualification rate and performance, and providing strong technical support for the development of related industries.

[0075] Algorithm for Enhancing Energy Absorption of Translucent Film

[0076] (I) Algorithm Principle

[0077] The energy absorption enhancement algorithm for transparent films is another key innovative algorithm in femtosecond adhesive-free edge-adaptive welding systems. It mainly focuses on solving the problems of energy absorption and reflection of transparent films during the welding process, in order to achieve more efficient and higher-quality welding. The principle of this algorithm is based on in-depth research on the optical properties of transparent films and precise understanding of the energy transfer mechanism during the welding process.

[0078] The transmittance (T), thickness (h), and refractive index (n) of a transparent film are crucial parameters affecting its energy absorption and reflection. The algorithm calculates the energy enhancement coefficient through a comprehensive analysis of these parameters. Its core principle lies in utilizing optical interference and thin-film optics theory. By precisely controlling the film's thickness and refractive index, the optical path difference between the reflected light from the upper and lower surfaces of the film satisfies the condition for destructive interference under illumination of a specific wavelength. When light enters the transparent film from air, reflection occurs on the upper surface due to the difference in refractive index between the film and air. After passing through the film, reflection occurs again at the interface between the film and the welded material. By adjusting the film thickness, the optical path difference between these two reflected beams is precisely half a wavelength. According to the interference principle, these two reflected beams cancel each other out, significantly reducing the energy of the reflected light, increasing the energy of the transmitted light, and improving the energy absorption efficiency of the transparent film. For example, in common optical components, such as the anti-reflective coating on the surface of camera lenses, this principle is utilized. By precisely controlling the thickness of the coating layer to one-quarter of the target wavelength, the reflectivity of light of a specific wavelength on the coating layer surface is greatly reduced, thereby improving the light transmission performance of the lens.

[0079] Surface microtexturing preprocessing is another important component of the energy absorption enhancement algorithm for transparent films. It primarily involves creating tiny textured structures on the surface of the transparent film using techniques such as laser micro-etching. The mechanisms of these microtextures are mainly reflected in the following aspects: First, they increase light scattering. When light shines on the microtextured film surface, the light is scattered by the irregular surface of the microtexture, lengthening the propagation path within the film and thus increasing the interaction time between the light and the film material, thereby improving light absorption efficiency. Second, they change the direction of light reflection. The microtexture structure makes the reflected light more dispersed, reducing the intensity of reflected light in a specific direction and further reducing reflectivity. Third, they enhance the adhesion between the film and the welded materials. The microtexture structure increases the surface roughness of the film, increasing the contact area between the film and the welded materials, thereby improving the adhesion between them and facilitating the welding process. For example, in the manufacturing of some solar panels, microtexturing the surface of the transparent film can effectively improve the absorption efficiency of sunlight by the solar panel, thereby improving the conversion efficiency of the solar cell.

[0080] (II) Efficiency Enhancement Analysis

[0081] In practical applications, the energy absorption enhancement algorithm for transparent films has demonstrated significant synergistic effects, particularly in reducing reflectivity, decreasing ablation area, and improving peel strength, providing strong support for improving welding quality. Taking the welding of PC film-ITO film for transparent displays as an example, traditional welding equipment, due to its ineffective handling of the reflection problem of the transparent film, requires high energy input to achieve good welding results. Typically, traditional equipment requires 1.8 J / cm². 2 The energy input is high, but excessive energy input will bring a series of problems. The most prominent one is that the ablation area on the PC film surface is large, accounting for up to 20%. This not only affects the appearance quality of the display, but may also have a negative impact on the performance of the display. At the same time, the peel strength after welding is only 7N / cm, which cannot meet the reliability requirements of the display during long-term use.

[0082] The femtosecond adhesive-free edge-adaptive welding system employing a light-transmitting film energy absorption enhancement algorithm demonstrates significant advantages in the same welding tasks. By accurately calculating the energy enhancement coefficient through the algorithm and combining it with surface micro-texturing preprocessing, the system can effectively reduce reflectivity. After processing, the reflectivity can be reduced to below 18%, meaning that more energy can be absorbed by the light-transmitting film and used in the welding process. In this case, the system requires only 1.2 J / cm². 2 The high energy consumption achieves excellent welding results, reducing energy requirements compared to traditional equipment. Simultaneously, due to efficient energy utilization and reduced reflectivity, the ablation area on the PC film surface is significantly reduced to only 1.5%, greatly improving the display's appearance quality and performance stability. The peel strength after welding is also significantly increased to 19 N / cm, a 171% improvement compared to traditional welding processes, ensuring the connection strength between the PC film and the ITO film and meeting the reliability requirements of the display under various complex operating environments.

[0083] In other scenarios involving the welding of transparent films, such as the welding of lenses for optical instruments and the packaging of electronic displays, the algorithm for enhancing the energy absorption of transparent films can also play an important role. It can effectively solve the problems of energy absorption and reflection of transparent films during the welding process, improve welding quality and production efficiency, and provide key technical support for the development of related industries.

[0084] Algorithm for Fast Adaptation of Small Batch Multi-Specification Parameters

[0085] (I) Algorithm Principle

[0086] The rapid adaptation algorithm for small-batch, multi-specification parameters is a key algorithm designed for welding different specifications of film materials in femtosecond adhesive-free edge-adaptive welding systems, aiming to achieve rapid and accurate adaptation of welding parameters, thereby improving production efficiency and product quality.

[0087] The core of this algorithm lies in constructing a comprehensive and efficient database linking membrane material specifications and parameters. During the database construction phase, key characteristic information of various membrane materials of different specifications is collected through extensive experimental and actual production data accumulation. This includes parameters such as membrane thickness, material composition, and edge condition, as well as the corresponding optimal welding parameters. These parameters cover multiple aspects, including welding energy, welding time, and welding speed, and they are interrelated, collectively affecting the quality and effect of the weld. For example, the required welding energy and time often differ for membrane materials of different thicknesses; thicker membrane materials require higher energy and longer welding times to ensure a strong weld connection. Furthermore, the differences in the physical and chemical properties of membrane materials, such as melting point and coefficient of thermal expansion, also affect the welding parameters, requiring targeted adjustments.

[0088] When faced with welding tasks using new membrane materials, operators simply input parameters such as the thickness, material, and edge condition of the new membrane material into the system. The algorithm quickly searches and matches data in the relevant database, using preset matching algorithms and models to find the most similar existing membrane material specifications. Then, based on this similarity data, specific calculation methods and rules are used to preliminarily calculate the initial welding parameters suitable for the new membrane material. This calculation is not a simple direct application, but rather fully considers the differences between the new and existing membrane materials in various parameters, using mathematical models and algorithms for reasonable adjustments and optimizations.

[0089] After the initial welding is completed, the intelligent weld edge quality inspection system performs a comprehensive and detailed inspection of the weld quality, acquiring various post-weld quality index data, such as weld smoothness, internal bonding strength, and the presence of defects. This quality feedback data is transmitted in real-time back to the small-batch, multi-specification parameter rapid adaptation algorithm module. Based on this feedback data, the algorithm uses intelligent optimization algorithms to further optimize and adjust the initial welding parameters. Intelligent optimization algorithms are typically based on advanced technologies such as machine learning and artificial intelligence. They can learn and analyze large amounts of welding quality data and parameter adjustment history, thereby automatically identifying the complex relationships and patterns between parameters and quality. For example, by training welding quality data under different parameter combinations using deep learning algorithms, a nonlinear mapping model between parameters and quality is established. Based on model predictions and actual quality feedback, the parameters are automatically adjusted to continuously bring the welding quality closer to the optimal state.

[0090] In subsequent batch welding processes, the algorithm will continuously and dynamically fine-tune the welding parameters based on the real-time detected welding quality data, ensuring that each welding is carried out under optimal parameter conditions, thereby guaranteeing the stability and consistency of welding quality.

[0091] (II) Efficiency Enhancement Analysis

[0092] In the scenario of welding customized PE film and silicone film for medical consumables, the algorithm for rapid adaptation of small batches and multiple specifications of parameters has shown significant efficiency improvement and has many advantages compared with traditional equipment.

[0093] Traditional equipment relies heavily on manual experience for parameter adjustments when handling small-batch, multi-specification welding tasks. This process is often time-consuming; for example, manual adjustments typically take 18 minutes for welding customized PE and silicone films for medical consumables. During this time, the production equipment is idle and unable to perform effective production operations, resulting in a significant waste of production time and resources and reduced production efficiency. Furthermore, the accuracy and consistency of manual adjustments are difficult to guarantee. Different operators may set different parameters due to differences in experience and judgment, leading to a low first-time weld pass rate, typically only 55%. A low pass rate not only means more raw material waste and increased production costs but can also affect the overall production schedule and product delivery time.

[0094] The femtosecond adhesive-free edge-fitting welding system, employing a rapid adaptation algorithm for small-batch, multi-specification parameters, demonstrated extremely high efficiency and stability under the same welding conditions. When switching to 20μm thick PE and silicone films, the system could quickly access the database and calculate initial parameters in a short time using the algorithm. After the first weld, based on feedback from the quality inspection system, the algorithm could quickly determine whether parameter adjustments were needed. If no adjustments were required, batch welding could proceed directly; if adjustments were necessary, parameter optimization could be completed in a very short time. In practical applications, the total adaptation time for batch welding 50 products was only 50 seconds (including the first weld), a 95% reduction compared to the 18 minutes of traditional equipment, significantly improving production efficiency and enabling the equipment to complete more production tasks in a shorter time.

[0095] Meanwhile, because the algorithm can dynamically adjust parameters based on real-time quality feedback, ensuring that each weld is performed under optimal parameter conditions, the product pass rate is significantly improved. In this welding scenario, the batch inspection pass rate reached 98%, an increase of 78% compared to the 55% of traditional equipment. This high pass rate not only reduces raw material waste and increases production costs but also improves product quality and reliability, enhancing the company's competitiveness in the market.

[0096] In other small-batch, multi-specification welding scenarios, such as welding electronic device components and automotive parts, the rapid parameter adaptation algorithm for small-batch, multi-specification welding can also play an important role. It can effectively solve the problem of parameter adaptation in traditional equipment, improve production efficiency and product quality, and bring significant economic benefits and market competitive advantages to enterprises.

[0097] Intelligent detection algorithm for weld edge quality

[0098] (I) Algorithm Principle

[0099] The intelligent detection algorithm for weld edge quality is an important component of the femtosecond adhesive-free edge-adaptive welding system. It mainly uses two key devices, a laser profilometer and an ultrasonic detector, to comprehensively and accurately detect and evaluate the quality of the weld edge, thereby achieving effective monitoring of weld quality and parameter optimization.

[0100] Laser profilometers operate on the principle of laser triangulation. A laser beam emitted from a semiconductor laser is expanded by a lens system to form a laser line projected onto the surface of the welding edge. The laser light scattered back from the welding edge is collected by another lens and projected onto a two-dimensional CMOS image sensor. An FPGA and signal processor analyze the resulting object contour image, calculating the distance (Z-coordinate) from each point on the laser line to the object, thus obtaining the three-dimensional contour information of the welding edge. With its high-precision measurement capabilities, the laser profilometer can accurately detect the flatness of the welding edge, with an accuracy of ±1μm. By analyzing and processing the acquired contour data and comparing it with a preset flatness standard, it can be determined whether the flatness of the welding edge meets the requirements. For example, in the welding production of automotive parts, the laser profilometer can monitor the flatness of the welding edge in real time, ensuring the connection accuracy between parts and improving the overall performance and safety of the vehicle.

[0101] Ultrasonic testing instruments detect the internal bond strength of welds based on the characteristics of ultrasonic waves propagating through materials. The instrument emits high-frequency ultrasonic waves through a probe towards the weld area. When these waves encounter defects within the weld (such as cracks, porosity, or lack of fusion) or interfaces between different media, reflection, refraction, and scattering occur. The reflected ultrasonic waves are received by the probe, converted into electrical signals, and displayed on the instrument as waveforms or images. Inspectors or intelligent algorithms analyze the internal structure and defects of the weld based on information such as the time, amplitude, and waveform characteristics of the reflected waves, thereby determining the internal bond strength. For example, in the welding quality inspection of pressure vessels, ultrasonic testing instruments can effectively detect minute defects within the weld, assess the weld strength and sealing performance, and ensure the safe operation of the pressure vessel under high-pressure environments.

[0102] After the laser profilometer and ultrasonic testing instrument complete the inspection of the weld edge quality, the inspection data is transmitted to the intelligent inspection algorithm module. The algorithm module first preprocesses this data to remove noise and interference signals, improving the accuracy and reliability of the data. Then, it performs a detailed comparative analysis with preset standard data. The preset standard data is typically formulated based on extensive experimental data, industry standards, and actual production needs; it includes the acceptable range and ideal values ​​for various quality indicators such as weld edge flatness and internal bond strength. Through comparison, the algorithm can quickly and accurately determine whether the weld quality is acceptable. If the inspection data exceeds the preset standard range, the algorithm automatically marks the welded product as unacceptable and further analyzes the inspection data to identify the root cause of the weld quality problem, such as unreasonable welding parameter settings or unstable factors in the welding process. Based on these analysis results, the algorithm generates targeted parameter optimization suggestions and feeds them back to the welding system to adjust and optimize subsequent welding processes, ensuring the stability and reliability of weld quality.

[0103] (II) Efficiency Enhancement Analysis

[0104] The intelligent detection algorithm for weld edge quality has demonstrated significant efficiency gains in practical applications, making important contributions to improving overall welding quality and production efficiency.

[0105] In terms of improving welding quality, this algorithm can promptly and accurately detect various quality problems that occur during the welding process. Taking the welding of lithium battery tab aluminum film-composite film as an example, under traditional welding quality inspection methods, due to the limitations of inspection means, it is difficult to detect some minor welding defects in a timely manner, such as micro-cracks and loose internal bonding. These potential quality problems may lead to performance degradation, shortened lifespan, or even safety hazards in lithium batteries during use. However, by adopting the intelligent detection algorithm for welding edge quality, through the collaborative work of a laser profilometer and an ultrasonic detector, comprehensive and high-precision inspection of the welding edge can be performed, promptly detecting these minor quality problems. Once a defective product is detected, the algorithm immediately marks it and provides parameter optimization suggestions. The welding system can adjust the welding parameters in a timely manner based on these suggestions to avoid similar quality problems in subsequent welded products, thereby greatly improving the stability and reliability of welding quality.

[0106] This algorithm also plays a crucial role in improving production efficiency. In traditional welding production, inspecting weld quality often requires significant time and manpower, and the accuracy and timeliness of the results are difficult to guarantee. This can lead to a large number of defective products during production, requiring rework or scrapping, thus wasting considerable production time and resources and reducing efficiency. The intelligent weld edge quality detection algorithm, however, automates and intelligently detects weld quality, offering fast and high-precision inspection. After welding is completed, it quickly provides inspection results, automatically marks defective products, and provides parameter optimization suggestions. This allows production personnel to adjust the welding process promptly, reducing rework and scrap due to quality issues and improving production efficiency. For example, in electronic equipment manufacturing companies, after adopting this algorithm, the production efficiency of the welding production line increased by more than 30%, and the product qualification rate rose from 80% to over 95%, bringing significant economic benefits to the company.

[0107] Furthermore, intelligent detection algorithms for weld edge quality can provide enterprises with abundant quality data and analysis reports. By accumulating and analyzing large amounts of welding quality data, enterprises can gain a deeper understanding of the trends and patterns of quality changes during the welding process, providing strong data support for further optimizing welding processes and improving product design, thereby promoting the enterprise's sustainable development and technological innovation.

[0108] Innovative Algorithm Collaborative Operation Mechanism

[0109] (I) Closed-loop control process

[0110] The closed-loop control process of "curling correction - energy regulation - parameter adaptation - quality inspection" in the femtosecond adhesive-free edge-adaptive welding system is a highly intelligent and collaborative operating system. Each link is closely connected and interacts with each other to jointly ensure the efficiency, accuracy and stability of the welding process.

[0111] Before welding begins, the real-time edge curling correction module takes center stage. When the material to be welded is placed on the welding platform, the integrated microscopic vision sensor, with its extremely high resolution (≥1μm), rapidly scans the material edges, accurately acquiring data on the height, angle, and curvature of the edge curling. This data, like a detailed "problem list," is immediately transmitted to the curling correction algorithm module. Based on a pre-established complex mathematical model and intelligent algorithm, the correction algorithm module performs in-depth analysis and calculation on the acquired data, quickly generating targeted correction force parameters (force value 0.1-1N) and application time. Subsequently, the micro-force actuator applies precise force to the curled edges according to these parameters, correcting the curled edges to an ideal state and ensuring the flatness of the weld joint (curling height ≤10μm after correction). This process is like a skilled craftsman meticulously trimming the edges of the material, laying a solid foundation for subsequent welding work.

[0112] Next, the energy absorption enhancement module for the transparent film begins operation. For the transparent film material, the system uses an energy absorption enhancement algorithm based on optical interference and thin-film optics theory, based on its characteristic parameters such as transmittance (T), thickness (h), and refractive index (n), to accurately calculate the energy enhancement coefficient. Simultaneously, micro-textured pretreatment of the transparent film surface is performed using laser micro-etching and other techniques to increase light scattering and change the reflection direction, further reducing reflectivity to below 20% and improving the film's energy absorption efficiency. This operation is like putting an "energy absorption enhancement coat" on the transparent film, enabling it to better absorb the energy required for welding and providing sufficient energy for high-quality welding.

[0113] After completing the above preparations, the small-batch, multi-specification parameter rapid adaptation module quickly searches and matches parameters such as the thickness, material, and edge condition of the new membrane material in the constructed membrane material specification-parameter association database. Through preset matching algorithms and models, combined with mathematical calculations and intelligent optimization, it rapidly generates initial welding parameters suitable for the membrane material. After the first weld is completed, the intelligent welding edge quality detection system performs a comprehensive inspection of the weld quality, acquiring key quality indicator data such as weld flatness and internal bonding strength. This quality feedback data is promptly transmitted back to the small-batch, multi-specification parameter rapid adaptation module. The algorithm optimizes and adjusts the initial welding parameters based on the feedback data, ensuring that subsequent welding processes are performed under optimal parameter conditions. This process acts like an intelligent "parameter adjustment master," continuously optimizing parameters based on the actual welding situation to maintain optimal welding quality.

[0114] The intelligent welding edge quality inspection system plays a crucial supervisory and feedback role throughout the welding process. During welding, it utilizes a laser profilometer and ultrasonic testing instrument to perform real-time, precise inspection of the weld edge flatness (accuracy ±1μm) and internal bond strength. Once an abnormality in welding quality is detected, such as flatness exceeding the allowable range or insufficient internal bond strength, the system immediately and automatically marks the welded product as defective and deeply analyzes the inspection data to identify the root cause of the quality problem. Based on these analysis results, the system generates detailed parameter optimization suggestions, which are fed back to other modules of the welding system, prompting corresponding adjustments and optimizations throughout the welding process, forming a complete closed-loop control loop. This process acts like a strict "quality gatekeeper," constantly monitoring welding quality to ensure that every welded product meets high-quality standards.

[0115] (II) Overall Efficiency Improvement

[0116] The collaborative work of these innovative algorithms has comprehensively improved welding quality and production efficiency from multiple dimensions, bringing significant upgrades and changes to industrial manufacturing.

[0117] Regarding welding quality, the various algorithms work together to effectively solve many problems existing in traditional welding processes. The real-time edge curling correction algorithm ensures the flatness of the weld joint, greatly reducing welding misalignment deviation and cracking rate. Taking the welding of lithium battery tab aluminum film-composite film as an example, the misalignment deviation of traditional welding process can reach ±35μm, and the cracking rate is as high as 25%. However, after adopting this algorithm, the misalignment deviation is reduced to ±8μm, and the cracking rate is only 2%, effectively improving the connection stability and reliability between the battery tab and the composite film. The light-transmitting film energy absorption enhancement algorithm reduces energy loss and ablation during the welding process by reducing reflectivity and improving energy absorption efficiency. In the welding scenario of PC film-ITO film for transparent displays, traditional processes require higher energy (1.8J / cm²). 2 Furthermore, with an ablation area accounting for 20%, the energy requirement is reduced to 1.2 J / cm² after adopting this algorithm. 2 With an ablation area of ​​only 1.5%, the appearance quality and performance stability of the display screen are significantly improved. The small-batch, multi-specification parameter rapid adaptation algorithm can quickly generate optimal welding parameters based on different film material specifications and adjust them promptly based on welding quality feedback, ensuring the consistency and stability of welding quality. In the scenario of welding customized PE film-silicone film for medical consumables, this algorithm achieves a batch inspection pass rate of 98%, a significant improvement compared to the 55% of traditional manual adjustments. The intelligent detection algorithm for welding edge quality promptly identifies and resolves welding quality problems through real-time, accurate quality detection and feedback, further ensuring the reliability of welding quality.

[0118] In terms of production efficiency, the collaborative work of innovative algorithms also performs exceptionally well. The rapid adaptation algorithm for small-batch, multi-specification parameters enables quick adaptation of welding parameters, significantly reducing pre-production debugging time. In customized welding scenarios for medical consumables, traditional manual debugging takes 18 minutes, while with this algorithm, the total adaptation time is only 50 seconds (including the initial weld), a 95% reduction. This allows production equipment to be put into production more quickly, improving equipment utilization and production efficiency. The automated and intelligent detection algorithm for weld edge quality reduces the time and workload of manual inspection, while timely parameter optimization suggestions prevent rework and scrap due to quality issues, further improving production efficiency. For example, in electronic equipment manufacturing companies, the production efficiency of welding production lines has increased by more than 30% after adopting this algorithm.

[0119] The above are merely preferred embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the invention by those skilled in the art. All modifications and variations within the spirit and principles of the present invention are permitted.

[0120] Any modifications, equivalent substitutions, improvements, etc., made should be included within the scope of protection of this invention.

Claims

1. A femtosecond adhesive-free edge-adaptive welding method, characterized in that, include: (1) Collect the height, angle and curvature data of the edge curling of the membrane material using a microscopic vision sensor, generate correction parameters and perform correction; (2) Obtain the transmittance, thickness and refractive index of the transparent film, and reduce the energy reflectivity through the energy enhancement algorithm; (3) Input the membrane material specifications and parameters, call the associated database, and calculate the initial welding parameters using the formula P=P0×(1+α×Δh+β×M+γ×C); (4) Perform welding and inspect the edge welding quality using a laser profilometer and an ultrasonic testing instrument; (5) Optimize welding parameters and correction parameters based on the quality inspection results.

2. The method according to claim 1, characterized in that, The energy enhancement algorithm described in step (2) includes: using the formula The energy enhancement coefficient was calculated, and combined with surface microtexturing preprocessing, the reflectivity was controlled to below 20%.

3. The method according to claim 1, characterized in that, The correction parameters mentioned in step (1) include the correction force and the application time. After correction, the edge curling height is ≤10μm and the angle is ≤0.5°.

4. The method according to claim 1, characterized in that, The quality inspection mentioned in step (4) includes edge flatness and internal bonding strength inspection.

5. The method according to claim 1, characterized in that, The membrane material specifications include thickness, material type, and edge curling coefficient.

6. A femtosecond adhesive-free edge-adaptive welding system, characterized in that, include: (1) Edge curling correction module: integrates microscopic vision sensor and micro-force actuator to collect curling data and perform correction; (2) Light Transmittance Energy Control Module: Obtains light transmittance film parameters and uses formulas... Calculate the energy enhancement coefficient to reduce reflectivity; (3) Parameter quick adaptation module: Call the membrane material specification-parameter database and generate initial welding parameters by formula P=P0×(1+α×Δh+β×M+γ×C); (4) Edge quality inspection module: The quality of the welded edges is inspected using a laser profilometer and an ultrasonic detector; (5) Parameter optimization module: Optimize welding and correction parameters based on quality inspection results.

7. The system according to claim 6, characterized in that, The edge curling correction module has a microscopic vision sensor resolution ≥1μm and a micro-force actuator force control accuracy of ±0.05N.

8. The system according to claim 6, characterized in that, The light transmission energy control module includes a laser micro-etching unit with an etching accuracy of ±2μm, which can generate micron-level textures.

9. The system according to claim 6, characterized in that, The membrane material specification-parameter database stores the baseline parameters and correction coefficients for at least 50 types of membrane materials.

10. The system according to claim 6, characterized in that, It also includes an execution module: receiving welding parameters and correction parameters, and controlling the laser output and the action of the micro-force actuator.

Images