Novel efficient light conversion photovoltaic glass and preparation method thereof

Abstract

The invention belongs to the technical field of solar photovoltaics, and provides novel efficient light conversion photovoltaic glass and a preparation method thereof.The preparation method comprises the steps that glass colored glaze ink corresponding to the novel efficient light conversion photovoltaic glass is prepared, and glass powder, a light conversion material, pearl powder, light diffusion particles and an organic carrier are mixed; the content of the light conversion material is more than 5% and less than 30%; the particle size of the pearl powder is larger than 5 microns and smaller than 20 microns, and the content is larger than 1% and smaller than 10%; the content of the light diffusion particles is greater than 0.1% and less than 5%; printing the prepared glass colored glaze ink on the surface of the front plate glass layer; carrying out toughening treatment on the front plate glass layer printed with the glass colored glaze printing ink, and sintering the glass colored glaze printing ink to form a glass colored glaze layer; the battery piece is bonded to the front plate glass layer with the glass colored glaze layer through the upper adhesive film layer, the back plate glass layer is bonded to the battery piece through the lower polymer adhesive film layer, and preparation of the novel efficient light conversion photovoltaic glass is completed.

Description

Technical Field

[0001] This invention belongs to the field of solar photovoltaic technology, specifically relating to a novel high-efficiency light conversion photovoltaic glass and its preparation method. Background Technology

[0002] As a key component of photovoltaic modules, solar photovoltaic glass directly affects module performance through its absorption and conversion efficiency of sunlight. While traditional photovoltaic glass uses light conversion materials with high conversion efficiency and uniform color across all viewing angles without color distortion, the color richness of these materials is insufficient, limiting the selection. Furthermore, current technologies lack precise control over the refractive index and particle size of light-diffusing particles, resulting in poor light scattering uniformity and uneven light intensity distribution on the cell surface, further impacting conversion efficiency. Existing fabrication methods suffer from the following drawbacks: 1. Insufficient color richness of light conversion materials: Currently, high-efficiency light conversion materials are mainly green, yellow, orange and red, lacking high-efficiency blue light conversion materials, as well as some special colors, such as copper, gold, gray, etc.

[0003] 2. Although pearlescent powder has vibrant and rich colors, its particles are flaky and lack fluidity, making it poorly compatible with screen printing processes.

[0004] 3. Large color difference: The color of pearlescent powder has a strong angular selectivity, which can easily cause color difference with different angles, resulting in a large color difference in the glass enamel produced.

[0005] 4. Inefficient process parameters: The proportioning of glass enamel inks, printing path planning and tempering treatment rely on empirical parameters, and the material composition and process conditions are not optimized in a coordinated manner, making it difficult to guarantee the bonding strength and stability between the light conversion material and the glass substrate.

[0006] While existing technologies have attempted to improve photovoltaic performance by adding light-conversion materials, no technical solution has yet achieved a combination of light-conversion materials and pearlescent powder in a specific ratio, matched with light-diffusing particles of specific refractive index and particle size, while simultaneously employing refined processes to realize efficient light conversion, uniform light distribution, and low color difference. Therefore, there is an urgent need for a novel photovoltaic glass fabrication method that can comprehensively cover the solar spectrum, improve light scattering efficiency, and optimize the fabrication process. Summary of the Invention

[0007] This application provides a novel high-efficiency light conversion photovoltaic glass and its preparation method, aiming to solve the problem that although there are attempts in the prior art to improve photovoltaic performance by adding light conversion materials, there is no technical solution that combines light conversion materials with pearl powder in a specific ratio and matches light diffusion particles with specific refractive indices and particle sizes, while achieving high-efficiency light conversion, uniform light distribution and low color difference through refined processes.

[0008] In a first aspect, this application provides a method for preparing a novel high-efficiency light-conversion photovoltaic glass, the method comprising: A novel high-efficiency light-conversion photovoltaic glass-based colored enamel ink is prepared by mixing glass powder, light-conversion materials, pearlescent powder, light-diffusing particles, and an organic carrier. The light-conversion materials include green, yellow, orange, and red light-conversion materials, with a content greater than 5% and less than 30%. The pearlescent powder has a particle size greater than 5 μm and less than 20 μm, with a content greater than 1% and less than 10%. The light-diffusing particles have a particle size greater than 0.1 μm and less than 0.5 μm and a refractive index greater than 1.7 and less than 2.7, and include one or more of titanium dioxide, zinc oxide, aluminum oxide, and yttrium oxide, with a content greater than 0.1% and less than 5%. The organic carrier, composed of organic solvent, thickener, and dispersant, accounts for greater than 20% and less than 40% of the total ink mass to impart printability to the ink. The prepared glass enamel ink is printed onto the surface of the front glass layer; the front glass layer printed with glass enamel ink is tempered to sinter the glass enamel ink to form a glass enamel layer. The solar cells are bonded to the front glass layer with a glass glaze layer by an upper adhesive film layer, and the back glass layer is bonded to the solar cells by a lower polymer adhesive film layer, thus completing the preparation of a new type of high-efficiency light conversion photovoltaic glass.

[0009] In some embodiments, the green light conversion material includes Lu3(Al,Ga)5O 12 :Ce 3+ (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ β-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+ The yellow light-converting material comprises (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ Orange light conversion materials include α-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+ Red light conversion materials include (Sr,Ca)AlSiN3:Eu 2+ and (Sr,Ca)2Si5N8:Eu 2+ .

[0010] In some embodiments, the preparation of the glass enamel ink corresponding to the novel high-efficiency light-conversion photovoltaic glass involves mixing glass powder, light-conversion material, pearlescent powder, light-diffusing particles, and an organic carrier. This includes: determining the mixing ratio of glass powder, light-conversion material, pearlescent powder, light-diffusing particles, and an organic carrier using a preset machine learning model, wherein the machine learning model is trained based on historical mixing data and corresponding glass enamel ink performance parameters; and, according to the determined mixing ratio, first adding the glass powder, light-conversion material, pearlescent powder, and light-diffusing particles to a mixing container, stirring at a first preset speed for a first preset time, then adding the organic carrier, and continuing to stir at a second preset speed for a second preset time to form a uniform glass enamel ink.

[0011] In some embodiments, printing the prepared glass enamel ink onto the surface of the front glass layer includes: acquiring surface contour data of the front glass layer through an image recognition system; generating a printing path based on the surface contour data using a path planning algorithm; printing according to the generated printing path using a screen printing device; monitoring the printing pressure in real time through a pressure sensor during the printing process; and automatically adjusting the pressure of the printing squeegee according to a preset pressure threshold range through a control system.

[0012] In some embodiments, the tempering process of the front glass layer printed with glass enamel ink and the sintering of the glass enamel ink to form a glass enamel layer includes: placing the printed front glass layer into a tempering furnace, collecting furnace temperature data in real time through a temperature sensor, calculating the heating rate and holding time at the current temperature using a preset neural network model, wherein the input parameters of the neural network model include the component ratio of the glass enamel ink, the thickness and size of the front glass layer; heating the tempering furnace to a preset temperature range according to the calculated heating rate, holding it at the temperature for a third preset time, and then cooling it to room temperature at a preset cooling rate, so that the glass enamel ink is sintered and forms a bonding layer with the front glass layer.

[0013] In some embodiments, the process of bonding the battery cell to the front glass layer with the glass enamel layer via the adhesive film layer includes: obtaining the position coordinates of the front glass layer with the glass enamel layer using a machine vision system, determining the bonding position of the battery cell using a positioning algorithm; cutting the adhesive film layer to a size that matches the battery cell, laying the adhesive film layer on the surface of the glass enamel layer of the front glass layer using a robotic arm, placing the battery cell in a preset position on the adhesive film layer, fixing the position of the battery cell using a vacuum adsorption device, so that the battery cell and the adhesive film layer are completely bonded, and monitoring the bonding pressure through a pressure sensor during the bonding process to ensure uniform pressure distribution.

[0014] In some embodiments, the process of bonding the backsheet glass layer to the battery cell via the lower polymer adhesive film layer includes: laying the lower polymer adhesive film layer on the surface of the battery cell facing away from the upper adhesive film layer; measuring the distance between the backsheet glass layer and the battery cell using a laser rangefinder; adjusting the descent speed and angle of the backsheet glass layer through a closed-loop control system to align the backsheet glass layer with the lower polymer adhesive film layer; and using a hot-pressing process to press the laid backsheet glass layer, lower polymer adhesive film layer, battery cell, and front sheet glass layer together. During the hot-pressing process, temperature and pressure data are collected in real time using temperature and pressure sensors. When the temperature and pressure reach preset conditions, these conditions are maintained for a fourth preset time to complete the bonding.

[0015] In some embodiments, the method further includes: before preparing the glass enamel ink, acquiring spectral distribution data of the target light source using a spectral analysis device, inputting the spectral distribution data into a pre-trained light conversion material ratio optimization algorithm model, wherein the light conversion material ratio optimization algorithm model is constructed based on the response spectrum of the photovoltaic cell and the excitation-emission spectrum data of the light conversion material, and outputting the optimal ratio combination of green light conversion material, yellow light conversion material, orange light conversion material and red light conversion material; and adjusting the content of the light conversion material according to the optimal ratio combination.

[0016] In some embodiments, the method further includes: after completing the preparation of the novel high-efficiency light conversion photovoltaic glass, obtaining the performance parameters of the photovoltaic glass such as short-circuit current, open-circuit voltage, and conversion efficiency through an electrical performance testing device, inputting the performance parameters into a preset process parameter feedback algorithm model, wherein the process parameter feedback algorithm model analyzes the correlation between the performance parameters and the process parameters such as the composition of the glass enamel ink, printing thickness, tempering temperature, and hot pressing time during the preparation process, outputting adjustment suggestions for the process parameters of the current batch, and storing the adjustment suggestions in the process database to guide the optimization of the preparation process for subsequent batches.

[0017] This invention utilizes four light conversion materials—green, yellow, orange, and red—to effectively convert short-wavelength ultraviolet and blue light from the solar spectrum into long-wavelength light within the sensitive wavelength range of photovoltaic cells. These materials (blue light excitation material for green excitation), green light excitation material for yellow excitation), orange light excitation material for orange excitation), and red light excitation material for red excitation. This results in a 10%-15% increase in short-circuit current and a 3%-5% improvement in conversion efficiency. Simultaneously, the high fluidity of the spherical light conversion material particles enhances the printability of the flake-shaped pearlescent powder. Furthermore, light-diffusing particles precisely control the light scattering of the glass enamel, reducing the color variations and angle-dependent color differences of the pearlescent powder. Tempering treatment chemically bonds the enamel layer to the glass substrate, and combined with the adhesive film bonding process, this enhances the mechanical strength (15% increase in impact resistance) and weather resistance of the photovoltaic glass (e.g., extending service life by 5-8 years).

[0018] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic flowchart illustrating the steps of a method for preparing a novel high-efficiency light-conversion photovoltaic glass according to an embodiment of this application; Figure 2 This is a schematic diagram of a novel high-efficiency light conversion photovoltaic glass provided in an embodiment of this application; Figure 3 This is a schematic block diagram of the structure of an electrical control box provided in one embodiment of this application.

[0021] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Detailed Implementation

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0024] It should be understood that, in order to clearly describe the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and the terms "first" and "second" are not necessarily different.

[0025] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0026] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0027] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0028] As a key component of photovoltaic modules, solar photovoltaic glass directly affects module performance through its absorption and conversion efficiency of sunlight. While traditional photovoltaic glass uses light conversion materials with high conversion efficiency and uniform color across all viewing angles without color distortion, the color richness of these materials is insufficient, limiting the selection. Furthermore, current technologies lack precise control over the refractive index and particle size of light-diffusing particles, resulting in poor light scattering uniformity and uneven light intensity distribution on the cell surface, further impacting conversion efficiency. Existing fabrication methods suffer from the following drawbacks: 1. Insufficient color richness of light conversion materials: Currently, high-efficiency light conversion materials are mainly green, yellow, orange and red, lacking high-efficiency blue light conversion materials, as well as some special colors, such as copper, gold, gray, etc.

[0029] 2. Although pearlescent powder has vibrant and rich colors, its particles are flaky and lack fluidity, making it poorly compatible with screen printing processes.

[0030] 3. Large color difference: The color of pearlescent powder has a strong angular selectivity, which can easily cause color difference with different angles, resulting in a large color difference in the glass enamel produced.

[0031] 4. Inefficient process parameters: The proportioning of glass enamel inks, printing path planning and tempering treatment rely on empirical parameters, and the material composition and process conditions are not optimized in a coordinated manner, making it difficult to guarantee the bonding strength and stability between the light conversion material and the glass substrate.

[0032] While existing technologies have attempted to improve photovoltaic performance by adding light-conversion materials, no technical solution has yet achieved a combination of light-conversion materials and pearlescent powder in a specific ratio, matched with light-diffusing particles of specific refractive index and particle size, while simultaneously employing refined processes to realize efficient light conversion, uniform light distribution, and low color difference. Therefore, there is an urgent need for a novel photovoltaic glass fabrication method that can comprehensively cover the solar spectrum, improve light scattering efficiency, and optimize the fabrication process.

[0033] To resolve the above issues, please refer to [link / reference]. Figure 1 , Figure 1 This is a schematic flowchart illustrating a method for preparing a novel high-efficiency light-conversion photovoltaic glass according to an embodiment of this application. This method for preparing such a high-efficiency light-conversion photovoltaic glass is used to prepare... Figure 2 The novel high-efficiency light conversion photovoltaic glass shown is shown.

[0034] like Figure 1 As shown, the provided method includes steps S101 to S103.

[0035] Step S101. Prepare a novel high-efficiency light-conversion photovoltaic glass-based colored enamel ink by mixing glass powder, light-conversion materials, pearlescent powder, light-diffusing particles, and an organic carrier. The light-conversion materials include green, yellow, orange, and red light-conversion materials, with a content greater than 5% and less than 30%. The pearlescent powder has a particle size greater than 5 μm and less than 20 μm, with a content greater than 1% and less than 10%. The light-diffusing particles have a particle size greater than 0.1 μm and less than 0.5 μm, a refractive index greater than 1.7 and less than 2.7, and include one or more of titanium dioxide, zinc oxide, aluminum oxide, and yttrium oxide, with a content greater than 0.1% and less than 5%. The organic carrier consists of an organic solvent, a thickener, and a dispersant, accounting for greater than 20% and less than 40% of the total ink mass, to impart printability to the ink.

[0036] Specifically, glass powder serves as the base binder for colored enamel inks. Lead-free, low-melting-point glass powder (such as the SiO2-B2O3-ZnO system) with a softening point of 500-600℃ is selected, with a particle size controlled at 1-5μm and a proportion of 50%-70% (by mass). Its function is to melt during the tempering and sintering process, firmly bonding the light conversion material and light diffusing particles to the surface of the front glass panel, forming a chemical bonding layer.

[0037] Light conversion materials include: green light conversion materials (excitation wavelength 450-500nm, emission wavelength 500-550nm): preferably Lu3(Al,Ga)5O 12 :Ce 3+ Or β-sialon:Eu 2+ This material is used to convert blue light into green light, matching the green light response peak of crystalline silicon solar cells. The yellow light conversion material (excitation wavelength 500-550nm, emission wavelength 550-600nm) is made from (Y,Ga)3(Al,Ga)5O. 12 :Ce 3+This converts green light into yellow light, enhancing the battery's absorption of the mid-band. Orange light conversion material (excitation wavelength 550-600nm, emission wavelength 600-650nm): α-sialon:Eu 2+ This converts orange light into near-infrared light, broadening the spectral range. The red light conversion material (excitation wavelength 600-650nm, emission wavelength 650-700nm) uses (Sr,Ca)AlSiN3:Eu 2+ It converts red light into long-wavelength light that is sensitive to the battery. The compounding rules are: the total content of the four materials is 5%-30% (mass percentage), and the content of any single component is not less than 1%. By adjusting the ratio, the spectral response can be optimized for different lighting conditions (such as insufficient red light in high-latitude regions and excessive blue light in low-latitude regions).

[0038] The light-diffusing particles include: materials and parameters: titanium dioxide (refractive index 2.5), zinc oxide (refractive index 2.0), or aluminum oxide (refractive index 1.76) with a particle size of 0.1-0.5μm (preferably 0.2-0.3μm) and a refractive index greater than 1.7 and less than 2.7 (forming a difference with the glass substrate's refractive index of 1.5-1.6), used alone or in combination, with a content of 0.1%-5% (mass percentage). Light scattering is induced by the refractive index difference, causing incident light to form diffuse reflection within the colored glaze layer, avoiding direct light spots and improving the uniformity of light intensity on the solar cell surface; particle size control prevents agglomeration or sedimentation, ensuring dispersion.

[0039] The organic carrier consists of organic solvents (terpineol, butyl carbitol, accounting for 60%-80%), thickeners (ethyl cellulose, accounting for 10%-20%), and dispersants (lecithin, accounting for 5%-15%), accounting for more than 20% and less than 40% of the total mass of the ink, in order to impart printability to the ink and ensure uniform dispersion of inorganic particles.

[0040] Step S102. Print the prepared glass enamel ink onto the surface of the front glass layer; temper the front glass layer with the printed glass enamel ink, and sinter the glass enamel ink to form a glass enamel layer.

[0041] Specifically, the screen printing process uses 200-300 mesh stainless steel wire mesh with a screen thickness of 50-100μm. The graphic design features honeycomb or matrix dots with a light transmittance of 30%-70% to avoid excessive light absorption caused by full-area coverage. A fully automatic screen printing machine is used, with the squeegee pressure set to 10-20N / cm and the printing speed at 50-100mm / s to ensure uniform ink deposition (wet film thickness 80-120μm). After printing, the screen is pre-dried at 80-100℃ for 10-15 minutes to remove solvents from the organic carrier and form a semi-cured film layer, preventing ink flow during the tempering process.

[0042] The tempering furnace parameters corresponding to the tempering sintering process include: Heating stage: Heating to 600-650℃ (glass powder softening point) at a rate of 5-10℃ / min for 30-40 minutes, melting the glass powder and encapsulating the light conversion material and diffusion particles. Holding stage: Holding at 600-650℃ for 15-20 minutes to ensure ion exchange occurs between the colored enamel layer and the surface of the front glass (3-5mm thick, made of ultra-clear patterned glass or float glass), forming a chemical bonding layer (bonding strength ≥5N / mm). Cooling stage: Cooling to room temperature at a rate of 15-20℃ / min, using a rapid cooling process to improve the mechanical strength of the glass (surface hardness ≥6H) while simultaneously fixing the structure of the colored enamel layer. During tempering, the melted glass powder forms covalent bonds with the front glass, the light conversion material is uniformly embedded in the glass matrix to prevent detachment, and the light diffusion particles are uniformly dispersed, forming stable scattering centers.

[0043] Step S103. The solar cell is bonded to the front glass layer with the glass glaze layer through the upper adhesive film layer, and the back glass layer is bonded to the solar cell through the lower polymer adhesive film layer, thus completing the preparation of the new high-efficiency light conversion photovoltaic glass.

[0044] Specifically, the adhesive film layer uses EVA (ethylene-vinyl acetate copolymer) or POE (polyolefin elastomer) film with a thickness of 0.3-0.5mm, a light transmittance of ≥90%, and a melting point of 80-100℃. It is used to bond the front glass enamel layer to the battery cells (polycrystalline silicon / monocrystalline silicon, size 156mm×156mm or 210mm×210mm).

[0045] The machine vision system (accuracy ±0.1mm) acquires the positioning marks (such as edge fiducial points) of the front glass enamel layer, calculates the cell bonding position, and ensures that the main grid lines of the cell are aligned with the light-transmitting area of ​​the enamel layer. After the robotic arm lays the adhesive film, a vacuum adsorption device places the cell in the center of the adhesive film and applies 5-10kPa pressure to pre-fix it and prevent displacement.

[0046] For the backsheet glass bonding (lower polymer adhesive film layer), the backsheet material is selected as 3-4mm thick low-iron tempered glass or polycarbonate (PC) backsheet with a weather-resistant coating. After the lower adhesive film layer (same type EVA / POE, 0.3-0.5mm thick) is laid, the lowering height of the backsheet glass is adjusted using a laser rangefinder (accuracy ±0.05mm) to ensure that the alignment error with the edge of the solar cell is ≤0.5mm.

[0047] The material is then placed in a laminator for hot pressing. The parameters are set as follows: temperature 130-150℃ (film melting temperature), pressure 80-100kPa, and time 10-15 minutes. During hot pressing, the film melts and fills the interlayer gaps. After cooling, it forms an integrated structure with a peel strength ≥30N / cm. The final structure consists of five layers: front glass (including the colored enamel layer), upper film layer, solar cell, lower film layer, and back glass. The colored enamel layer, located inside the front glass, is 50-80μm thick. Light conversion materials and diffuser particles are uniformly distributed, achieving a highly efficient energy transfer path of "light incident - spectral conversion - scattering homogenization - cell absorption".

[0048] In some embodiments, the green light conversion material includes Lu3(Al,Ga)5O 12 :Ce 3+ (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ β-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+ The yellow light-converting material comprises (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ Orange light conversion materials include α-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+ Red light conversion materials include (Sr,Ca)AlSiN3:Eu 2+ and (Sr,Ca)2Si5N8:Eu 2+ .

[0049] In some embodiments, the preparation of the glass enamel ink corresponding to the novel high-efficiency light-conversion photovoltaic glass involves mixing glass powder, light-conversion material, pearlescent powder, light-diffusing particles, and an organic carrier. This includes: determining the mixing ratio of glass powder, light-conversion material, pearlescent powder, light-diffusing particles, and organic carrier using a preset machine learning model, wherein the machine learning model is trained based on historical mixing data and corresponding glass enamel ink performance parameters; and, according to the determined mixing ratio, first adding glass powder, light-conversion material, and light-diffusing particles to a mixing container, stirring at a first preset speed for a first preset time, then adding the organic carrier, and continuing stirring at a second preset speed for a second preset time to form a uniform glass enamel ink.

[0050] By introducing machine learning models to dynamically optimize the component ratio of glass enamel inks, replacing traditional empirical parameters, a precise match between material composition and ink performance can be achieved.

[0051] The machine learning model construction includes: Training data: collecting over 2000 sets of historical mixed data (glass powder content 50%-70%, total light conversion material content 5%-30%, light diffusing particles 0.1%-5%, pearlescent powder 1-10%, organic carrier greater than 20% and less than 40%), corresponding to ink performance parameters (viscosity, solid content, dispersibility index). Model architecture: using a BP neural network (3 hidden layers, 64-32-16 nodes per layer), with the input layer representing the proportion of each component, the output layer representing the performance prediction value, and the loss function being the mean squared error, trained until the convergence error is ≤0.05.

[0052] The mixing process control includes: First stage mixing: Glass powder, light conversion material, pearlescent powder, and light diffusing particles are added to a dual planetary mixer and stirred at a preset speed of 300-500 rpm for a preset time of 15-20 minutes to initially disperse the inorganic particles (dispersion ≥95%). Second stage mixing: After adding the organic carrier, the mixing speed is switched to a second preset speed of 1000-1500 rpm and stirred for a second preset time of 40-60 minutes, ultimately controlling the ink viscosity to 10-100 Pa·s (adjusted in real-time based on model feedback).

[0053] In some embodiments, printing the prepared glass enamel ink onto the surface of the front glass layer includes: acquiring surface contour data of the front glass layer through an image recognition system; generating a printing path based on the surface contour data using a path planning algorithm; printing according to the generated printing path using a screen printing device; monitoring the printing pressure in real time through a pressure sensor during the printing process; and automatically adjusting the pressure of the printing squeegee according to a preset pressure threshold range through a control system.

[0054] By generating personalized printing paths using surface contour data and combining them with real-time pressure monitoring, the problem of inconsistent ink thickness caused by fixed printing paths and uneven pressure in traditional printing methods can be solved.

[0055] Surface contour data acquisition uses a 3D line laser scanner (accuracy ±5μm) to scan the surface of the front glass layer, obtain flatness data (warpage ≤0.3mm / m), generate point cloud data and import it into the path planning algorithm (an optimized version of the Dijkstra algorithm).

[0056] The printing path generation automatically adjusts the printing trajectory based on the curvature of the glass surface, increasing the number of printing passes (1-2 times) in recessed areas and reducing the squeegee pressure (gradient 0.5N / cm) in raised areas, ensuring that the uniformity error of the wet film thickness is ≤±5μm.

[0057] Real-time pressure control is achieved by installing a piezoelectric ceramic pressure sensor (accuracy ±0.1N / cm) on the scraper. When the detected pressure exceeds the preset range (15±2N / cm), the servo motor automatically adjusts the scraper angle (±2°). The pressure is closed-loop controlled by a PID controller, with a response time ≤0.2 seconds.

[0058] In some embodiments, the tempering process of the front glass layer printed with glass enamel ink and the sintering of the glass enamel ink to form a glass enamel layer includes: placing the printed front glass layer into a tempering furnace, collecting furnace temperature data in real time through a temperature sensor, calculating the heating rate and holding time at the current temperature using a preset neural network model, wherein the input parameters of the neural network model include the component ratio of the glass enamel ink, the thickness and size of the front glass layer; heating the tempering furnace to a preset temperature range according to the calculated heating rate, holding it at the temperature for a third preset time, and then cooling it to room temperature at a preset cooling rate, so that the glass enamel ink is sintered and forms a bonding layer with the front glass layer.

[0059] By using a neural network model to dynamically calculate the tempering process parameters based on ink composition and glass parameters, the problem of insufficient bonding strength caused by traditional empirical parameters is solved.

[0060] The neural network model's input and output include: Input parameters: glass enamel ink composition (glass powder softening point, proportion of light conversion material, type of light diffusing particles), front glass layer thickness (3-5mm), dimensions (1200mm×1800mm, etc.). Output parameters: heating rate (dynamically adjusted from 5-10℃ / min), holding time (adaptively calculated from 15-20 minutes), preset temperature range (600-650℃, ±10℃ based on glass powder softening point).

[0061] The tempering furnace control process includes: a temperature sensor (K-type thermocouple, accuracy ±1℃) collects the furnace temperature in real time and uploads the data to the model every 10 seconds. The model calculates the optimal heating rate based on the current composition and glass size (e.g., 6℃ / min for 5mm thick glass and 8℃ / min for 3mm thick glass).

[0062] After the heat preservation stage, the glass is cooled at a preset cooling rate (15-20℃ / min). The temperature uniformity of the glass surface is monitored by an infrared thermometer (temperature difference ≤5℃) to ensure that the colored glaze layer and the glass substrate form a chemical bonding layer (bonding strength is ≥5N / mm through pull-out test).

[0063] In some embodiments, the process of bonding the battery cell to the front glass layer with the glass enamel layer via the adhesive film layer includes: obtaining the position coordinates of the front glass layer with the glass enamel layer using a machine vision system, determining the bonding position of the battery cell using a positioning algorithm; cutting the adhesive film layer to a size that matches the battery cell, laying the adhesive film layer on the surface of the glass enamel layer of the front glass layer using a robotic arm, placing the battery cell in a preset position on the adhesive film layer, fixing the position of the battery cell using a vacuum adsorption device, so that the battery cell and the adhesive film layer are completely bonded, and monitoring the bonding pressure through a pressure sensor during the bonding process to ensure uniform pressure distribution.

[0064] High-precision bonding of battery cells is achieved through machine vision positioning and vacuum adsorption technology, solving the offset problem caused by traditional manual positioning.

[0065] Position coordinate acquisition includes: using a line scan camera (12μm / pixel resolution) to capture fiducial marks on the edge of the front glass enamel layer, and calculating the coordinate system deviation (x / y axis error ≤ ±0.2mm, rotation angle ≤ ±0.1°) using the Halcon vision algorithm.

[0066] The adhesive film laying and cell fixing process includes: cutting the upper adhesive film layer (EVA / POE) using a laser cutter (dimensional accuracy ±0.3mm); laying it with a robotic arm (repeat positioning accuracy ±0.1mm) pre-reserving the position of the main busbar of the cell (deviation ≤ ±0.5mm); and using a vacuum adsorption device (suction force 5-10kPa) to pick up the cell and place it in the preset position of the adhesive film. During the bonding process, pressure sensors (distributed on the adsorption platform) monitor the pressure uniformity (single-point pressure difference ≤ ±1kPa) to ensure complete bonding between the cell and the adhesive film (bubble rate ≤ 0.1%).

[0067] In some embodiments, the process of bonding the backsheet glass layer to the battery cell via the lower polymer adhesive film layer includes: laying the lower polymer adhesive film layer on the surface of the battery cell facing away from the upper adhesive film layer; measuring the distance between the backsheet glass layer and the battery cell using a laser rangefinder; adjusting the descent speed and angle of the backsheet glass layer through a closed-loop control system to align the backsheet glass layer with the lower polymer adhesive film layer; and using a hot-pressing process to press the laid backsheet glass layer, lower polymer adhesive film layer, battery cell, and front sheet glass layer together. During the hot-pressing process, temperature and pressure data are collected in real time using temperature and pressure sensors. When the temperature and pressure reach preset conditions, these conditions are maintained for a fourth preset time to complete the bonding.

[0068] High-precision alignment of the backplate glass is achieved through laser ranging and closed-loop control, and the interlayer bonding strength is improved by real-time monitoring of hot pressing process parameters.

[0069] Spacing measurement and alignment include: a laser rangefinder (accuracy ±0.05mm) is installed on the back glass robotic arm to measure the distance between the back panel and the edge of the battery cell; a closed-loop control system (PID algorithm) adjusts the descent speed (0.5-1mm / s) and angle (±0.5°) to ensure that the edge alignment error is ≤±0.5mm.

[0070] The hot-pressing process control includes: the laminator has built-in temperature sensors (accuracy ±1℃) and pressure sensors (accuracy ±1kPa) to collect data in real time. When the temperature reaches the preset value (140℃ for EVA film, 135℃ for POE film) and the pressure reaches 80±5kPa, the heat preservation and pressure holding program is triggered, lasting for a fourth preset duration of 12-15 minutes (dynamically adjusted according to the film thickness). After cooling, the bonding effect is verified by peel strength testing (ASTM D3330, ≥30N / cm), and the data is synchronized to the MES system for process traceability.

[0071] In some embodiments, the method further includes: before preparing the glass enamel ink, acquiring spectral distribution data of the target light source using a spectral analysis device, inputting the spectral distribution data into a pre-trained light conversion material ratio optimization algorithm model, wherein the light conversion material ratio optimization algorithm model is constructed based on the response spectrum of the photovoltaic cell and the excitation-emission spectrum data of the light conversion material, and outputting the optimal ratio combination of green light conversion material, yellow light conversion material, orange light conversion material and red light conversion material; and adjusting the content of the light conversion material according to the optimal ratio combination.

[0072] Based on the target light source spectrum and the battery response spectrum, the ratio of light conversion materials is dynamically optimized through an algorithm model to solve the spectral matching problem under different lighting conditions.

[0073] Spectral data acquisition includes: acquiring the light source spectrum of the target area using a spectral analysis device (xenon lamp simulator + spectrometer, accuracy ±2%) (e.g., red light enhancement in high-latitude regions and blue light enhancement in low-latitude regions), with a resolution of 1nm and covering the 300-1100nm wavelength band.

[0074] The proportioning optimization algorithm model is based on the photovoltaic cell response spectrum (peak value of crystalline silicon cells 400-1100nm) and the excitation-emission spectrum database of light conversion materials (containing 100+ material parameters). The Particle Swarm Optimization (PSO) algorithm is used to solve for the optimal proportions: Objective function: Maximize conversion efficiency = Σ(emission spectrum of light conversion material × cell responsivity) - scattering loss. Constraints: Total content 5%-30%, single component ≥1%, calculation time ≤5 minutes / time. The output results are directly imported into the batching system, automatically adjusting the weighing ratio of each material (accuracy ±0.1g).

[0075] In some embodiments, the method further includes: after completing the preparation of the novel high-efficiency light conversion photovoltaic glass, obtaining the performance parameters of the photovoltaic glass such as short-circuit current, open-circuit voltage, and conversion efficiency through an electrical performance testing device, inputting the performance parameters into a preset process parameter feedback algorithm model, wherein the process parameter feedback algorithm model analyzes the correlation between the performance parameters and the process parameters such as the composition of the glass enamel ink, printing thickness, tempering temperature, and hot pressing time during the preparation process, outputting adjustment suggestions for the process parameters of the current batch, and storing the adjustment suggestions in the process database to guide the optimization of the preparation process for subsequent batches.

[0076] By using electrical performance test data to optimize manufacturing process parameters, a closed loop of "test-analysis-feedback" is formed, continuously improving product yield.

[0077] Performance parameter acquisition includes: acquiring short-circuit current (Isc), open-circuit voltage (Voc), and conversion efficiency (Eff) using electrical performance testing equipment (IV tester, accuracy ±0.5%). Each sample is tested 5 times and the average value is taken. The data is then uploaded to the PLC system.

[0078] The process parameter feedback model uses the Apriori algorithm to analyze the correlation between performance parameters and process parameters. For example, when Isc decreases by more than 5%, the correlation analysis points to "low light conversion material content" or "insufficient tempering temperature." When Eff fluctuates by more than 3%, it identifies "uneven printing thickness" or "hot pressing time deviation." Adjustment suggestions are output (such as increasing green material by 2% or increasing tempering temperature by 10°C) and stored in the process database (SQL Server). The optimized parameters are automatically loaded during the next production batch, achieving iterative optimization of process parameters.

[0079] In some embodiments, by constructing a real-time monitoring network for the status of tempering furnace equipment, and using a long short-term memory network (LSTM) to predict the failure probability of key components (heating wire, drive roller), the problem of over-maintenance or sudden downtime caused by traditional periodic maintenance is solved, and predictive maintenance is achieved.

[0080] The sensor network deployment involves installing thermocouples (temperature monitoring, accuracy ±1℃), vibration sensors (drive roller vibration frequency, accuracy ±0.1Hz), and current sensors (heating wire current, accuracy ±0.5A) in the heating zone of the tempering furnace. Data is collected every 2 seconds and uploaded to the cloud via an edge computing gateway (MQTT protocol).

[0081] The LSTM fault prediction model includes: Training data: Collecting 3 years of historical equipment fault data (including 5 types of faults such as heating wire aging and transmission roller bearing wear), extracting features such as temperature fluctuation curves, abnormal current pulses, and sudden changes in vibration amplitude, and constructing a time series dataset (sliding window size 500). Model architecture: 2-layer LSTM (128 units per layer) + fully connected layer, outputting fault probability (0-1), with a threshold of 0.8 triggering an early warning, and a prediction accuracy ≥92%. When the model predicts a heating wire fault probability >80%, the system automatically schedules a backup heating module to connect, and simultaneously generates a work order (including replacement location and spare part model) and pushes it to the maintenance terminal, reducing downtime by 70%; when the transmission roller is abnormal, the running speed of the tempering furnace is dynamically adjusted (±5%) to avoid uneven sintering of the colored glaze layer caused by vibration.

[0082] In some embodiments, by combining real-time meteorological data with real-time power generation data from photovoltaic power plants, the ratio of light conversion materials is dynamically optimized through deep reinforcement learning (DRL) to achieve real-time closed-loop optimization of "light conditions-material ratio-power generation efficiency," thus solving the problem that traditional fixed ratios cannot adapt to seasonal / regional light changes.

[0083] Input data includes: real-time spectral data of the target area (collected by a spectrometer deployed in the power plant, every 10 minutes), historical irradiance curves for 1 year, real-time IV characteristics of crystalline silicon cells (conversion efficiency, short-circuit current), and excitation-emission spectral parameters of various light conversion materials in the material library.

[0084] The DRL model construction includes: State space: current spectral energy distribution (300-1100nm, 10nm resolution), cell temperature (°C), and material inventory. Action space: adjustment step size for the ratio of four light conversion materials (±0.5%, total content constrained to 5%-30%). Reward function: Reward value = predicted power generation efficiency improvement rate × 0.6 + material cost saving rate × 0.4 - ratio adjustment range × 0.1. The model is trained using the PPO algorithm, with the strategy updated every 2 hours.

[0085] The system generates an initial ratio at 0:00 every day based on the 7-day weather forecast (light intensity and spectral distribution prediction). During the day, it makes fine adjustments every hour based on real-time spectral data (adjustment range ≤2%). The system automatically replenishes or reduces the amount of each material fed in through an industrial robot (accuracy ±0.05g). Actual measurements show that the conversion efficiency is improved in different seasons.

[0086] In some embodiments, by establishing a digital twin model of the front glass printing process, the printing status of the physical entity is mapped in real time, and the batch defect problem caused by the lag in traditional manual quality inspection is solved through defect identification and self-correction of process parameters.

[0087] The digital twin model was constructed using ABAQUS to establish a mechanical model of the glass substrate (considering a thickness of 3-5 mm and a warpage of ≤0.5 mm / m), combined with a printing press dynamics model (squeegee pressure-ink flow coupling equation), and a 1:1 virtual printing production line was built in Unity with a synchronization rate of ≥98%.

[0088] Real-time defect diagnosis includes: acquiring wet film images after printing using a line scan camera (12K resolution), identifying defect types (white exposure, ink accumulation, edge jaggedness, accuracy ≥95%) using the YOLOv8 model, mapping defect coordinates to a digital twin in real time, and tracing the causes through finite element analysis (such as ink accumulation at the edge caused by a +0.3° deviation in the doctor blade angle).

[0089] The self-correction mechanism includes: when three consecutive glass sheets exhibit the same defect, the system automatically triggers parameter corrections: For white showing defects: increase the organic carrier content by 1% (to improve ink flowability), and simultaneously increase the doctor blade pressure by 1 N / cm. For ink accumulation defects: reduce the printing speed by 5 mm / s, and switch the screen mesh from 250 mesh to 300 mesh. After correction, the digital twin pre-verifies the correction effect (time ≤ 2 minutes), and upon confirmation, automatically sends the results to the printing press PLC, resulting in a decrease in the defect rate.

[0090] In some embodiments, by using transfer learning technology to rapidly transfer the process parameter experience of mature production lines to new production lines for multiple factories and multiple models of photovoltaic glass production lines, the problem of high parameter trial and error costs and long debugging cycles during the introduction of new products can be solved.

[0091] Source and target domain data processing includes: Source domain: collecting historical data (200,000+ batches, including composition, printing, tempering parameters, and performance indicators) for 1800×1200mm glass from Factory A. Target domain: data from newly produced 2100×1500mm glass from Factory B, with initial data limited to only 500 batches. A Domain-Adversarial Neural Network (DANN) is used to eliminate distribution shifts caused by size and thickness differences, while preserving core process logic (such as the relationship between glass powder softening point and tempering temperature).

[0092] The transfer learning execution process includes: freezing the first three convolutional layers of the source domain model (such as the tempering temperature prediction model in Example 4), fine-tuning only the last two fully connected layers, and using a small amount of data (500 batches) from the target domain for fine-tuning. The number of training iterations is reduced from the traditional 2000 to 300, and the convergence speed is improved by 80%. During the new production line debugging phase, the system automatically compares the deviation between the predicted parameters and the actual performance (e.g., triggering local retraining when the conversion efficiency prediction error is >2%), shortening the debugging cycle from the traditional 2 weeks to 3 days, and achieving a yield improvement of over 95% in the initial stage.

[0093] In some embodiments, by constructing a knowledge graph covering material formulation, production process, equipment status, and quality inspection data, rapid root cause analysis and process traceability of quality problems can be achieved, solving the problem of low traceability efficiency caused by traditional data silos.

[0094] The knowledge graph construction includes: Entity types: materials (glass powder, phosphor), equipment (printing press, tempering furnace), process parameters (mixing speed, hot pressing time), and quality indicators (bonding strength, conversion efficiency), totaling 50+ entity categories; and Relationship types (such as "influence," "association," "cause"): 30+ types. Data access: Data is extracted from MES, LIMS, and SCADA systems using ETL tools and stored in the Neo4j graph database, with over 100,000 nodes and over 500,000 relationships.

[0095] Root cause analysis applications include: When the bonding strength of a batch of products fails to meet the standard (<5N / mm), the user inputs the problem node, and the system automatically traverses the graph: Path 1: Aging of the tempering furnace heating wire (equipment entity) → Fluctuation in heating rate (process entity) → Insufficient melting of glass powder (material entity) → Decreased bonding strength (quality entity). Path 2: Agglomeration of light-diffusing particles (material entity) → Uneven thickness of printed wet film (process entity) → Stress concentration during sintering (physical phenomenon) → Defects in the bonding layer. The system outputs the top 3 root causes and rectification suggestions (such as replacing the heating wire, adjusting the dispersant content to 0.5%), reducing the average source tracing time from 4 hours manually to 8 minutes.

[0096] In some embodiments, by utilizing generative adversarial networks to autonomously design novel light conversion materials, the research and development bottlenecks of traditional trial-and-error methods are overcome, and reverse optimization of material composition and spectral conversion efficiency is achieved.

[0097] The GAN model architecture includes: A generator: inputs a random vector (100 dimensions) and outputs the chemical formula of the material (e.g., adjusting the cation ratio parameters x and y), using a Transformer architecture to process the chemical formula sequence. A discriminator: inputs real material spectral data (from the JCPDS database, containing various phosphor parameters) and generated material spectral predictions (emission spectra simulated using density functional theory (DFT)), with the training objective of distinguishing between real and generated data.

[0098] The materials development process includes: Pre-training stage: A GAN is trained using data from over 100,000 known materials. The generator outputs feasible chemical formulas, and their emission peak wavelength and full width at half maximum (FWHM) are calculated using DFT. Candidate materials meeting the requirement of "emission peak covering the battery response blind zone (e.g., 700-750nm)" are screened. The top 5 generated materials are prepared into inks, and their actual emission spectra are tested using a spectrometer, verifying an accuracy of ≥85%. One novel red material further improves conversion efficiency by 1.2%, shortening the development cycle from the traditional 18 months to 3 months.

[0099] In some embodiments, reinforcement learning is used to optimize the layout algorithm for the front glass cutting process, maximizing the utilization rate of glass raw materials while meeting the cutting accuracy requirements, thus solving the problem of insufficient adaptation of traditional heuristic algorithms to complex sizes.

[0100] The state and action definitions include: State space: current glass sheet size (2400×3600mm), remaining cutable area outline, and set of order sizes to be cut (including various specifications such as 156mm×156mm and 210mm×210mm). Action space: select the next order size to cut, determine the cutting position (accuracy ±0.1mm), and use a quadtree data structure to represent the cutting area division. Reward function: Reward value = raw material utilization rate × 0.7 + cutting path length saving rate × 0.3, penalizing invalid cutting actions (such as when the remaining area cannot accommodate the smallest order size).

[0101] The model is trained and deployed using the Proximal Policy Optimization (PPO) algorithm, which simulates multiple cutting processes in a virtual environment to learn the optimal layout strategy. In actual production, the system dynamically adjusts the layout according to order priority, increasing the raw material utilization rate from 82% to 88% compared to the traditional algorithm. It also supports real-time insertion of urgent orders (response time ≤ 30 seconds), and the optimized cutting path reduces the cutting time of each piece of glass by 15%.

[0102] In some embodiments, by deploying edge computing nodes in key processes such as printing and tempering, and carrying lightweight deep learning models to detect glass surface defects in real time, the problems of high latency and strong network dependence in cloud processing are solved.

[0103] The edge computing nodes use NVIDIA Jetson AGX Orin (200 TOPS computing power) and integrate a line scan camera (12K resolution, 100fps acquisition speed). Each node independently processes the inspection tasks of one production line and transmits the inspection results back to the server via 5G.

[0104] The lightweight model design is based on the MobileNetV3+FPN architecture to design the detection model. The number of parameters is compressed to 15MB, the detection speed is ≥150fps (meeting the real-time detection requirements of glass transmission speed of 1.5m / s), and it supports the detection of 6 types of defects such as exposed white glaze, bubbles, and cracks, with mAP@0.5≥90%.

[0105] The detection and feedback closed loop includes: when three consecutive pieces of glass show the same defect, the edge node automatically triggers production line linkage; for defects in the printing process: the doctor blade pressure is adjusted in real time (±0.5N / cm) and the ink stirring system is notified to check the viscosity (target value ±5Pa*s); for cracks in the tempering process: a signal is sent to the tempering furnace temperature control system to force the heating rate fluctuation ≤±1℃ / min, while reducing the cooling stage air pressure by 10%.

[0106] This invention utilizes the synergistic effect of four light conversion materials—green, yellow, orange, and red—to effectively convert blue light (excited by green light materials), green light (excited by yellow light materials), orange light (excited by orange light materials), and red light (excited by red light materials) from the solar spectrum into the sensitive wavelengths of photovoltaic cells. This results in a 10%-15% increase in short-circuit current and a 3%-5% improvement in conversion efficiency. Precise control of the particle size and refractive index of the light-diffusing particles (0.1-0.5 μm, greater than 1.7 and less than 2.7) forms a highly efficient scattering structure within the glass enamel layer, reducing direct light loss and improving the uniformity of light intensity on the cell surface by 20%-30%, thus mitigating the risk of hot spot effects. Tempering treatment chemically bonds the enamel layer to the glass substrate, and combined with the adhesive film bonding process, enhances the mechanical strength (15% increase in impact resistance) and weather resistance of the photovoltaic glass (extending its service life by 5-8 years).

[0107] like Figure 2 As shown, the novel high-efficiency light conversion photovoltaic glass provided by the present invention includes a front glass layer 12 and a glass enamel layer 11 printed and sintered on the front glass layer 12. The glass enamel layer 11 is formed by mixing and sintering glass powder, pigment, light diffusing particles, pearl powder and organic carrier. The solar cell 14 is bonded to the front glass layer 12 and the back glass layer 16 through an upper adhesive film layer 13 and a lower polymer adhesive film layer 15, respectively.

[0108] A novel high-efficiency light-conversion photovoltaic glass-based colored enamel ink is prepared by mixing glass powder, light-conversion materials, pearlescent powder, light-diffusing particles, and an organic carrier. The light-conversion materials include green, yellow, orange, and red light-conversion materials, with a content greater than 5% and less than 30%. The light-diffusing particles have a particle size greater than 0.1 μm and less than 0.5 μm, a refractive index greater than 1.7 and less than 2.7, and include one or more of titanium dioxide, zinc oxide, aluminum oxide, and yttrium oxide, with a content greater than 0.1% and less than 5%. The organic carrier, composed of organic solvent, thickener, and dispersant, accounts for greater than 20% and less than 40% of the total ink mass to impart printability to the ink. The prepared glass enamel ink is printed onto the surface of the front glass layer; the front glass layer printed with glass enamel ink is tempered to sinter the glass enamel ink to form a glass enamel layer. The solar cells are bonded to the front glass layer with a glass glaze layer by an upper adhesive film layer, and the back glass layer is bonded to the solar cells by a lower polymer adhesive film layer, thus completing the preparation of a new type of high-efficiency light conversion photovoltaic glass.

[0109] In some embodiments, the green light conversion material includes Lu3(Al,Ga)5O 12 :Ce 3+ (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ β-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+ Mixed with green pearlescent pigment; the yellow light-converting material comprises (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ Mixed with yellow pearlescent pigments; orange light-converting materials include α-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+ Mixed with orange pearlescent pigments; red light-converting materials include (Sr,Ca)AlSiN3:Eu 2+ and (Sr,Ca)2Si5N8:Eu 2+ When mixed with red pearlescent pigments, metallic pearlescent powders such as gray, silver, gold, and copper can be added for special color requirements.

[0110] In some embodiments, the mixing ratio of glass powder, light conversion material, pearlescent powder, light diffusing particles, and organic carrier is determined by using a preset machine learning model, which is trained based on historical mixing data and corresponding glass enamel ink performance parameters. According to the determined mixing ratio, the glass powder, light conversion material, and light diffusing particles are first added to a mixing container and stirred at a first preset speed for a first preset time. Then, the organic carrier is added and stirred at a second preset speed for a second preset time to form a uniform glass enamel ink.

[0111] In some embodiments, surface contour data of the front glass layer is obtained through an image recognition system, and a printing path is generated based on the surface contour data using a path planning algorithm; a screen printing device is used to print according to the generated printing path, and the printing pressure is monitored in real time by a pressure sensor during the printing process, and the pressure of the printing squeegee is automatically adjusted by the control system according to a preset pressure threshold range.

[0112] In some embodiments, the printed front glass layer is placed in a tempering furnace, and the temperature data inside the furnace is collected in real time by a temperature sensor. A preset neural network model is used to calculate the heating rate and holding time at the current temperature. The input parameters of the neural network model include the composition ratio of the glass enamel ink, the thickness and size of the front glass layer. The tempering furnace heats the glass to a preset temperature range according to the calculated heating rate, holds it for a third preset time, and then cools it to room temperature at a preset cooling rate, so that the glass enamel ink is sintered and forms a bonding layer with the front glass layer.

[0113] In some embodiments, the position coordinates of the front glass layer with the glass enamel layer are obtained by using a machine vision system, and the bonding position of the battery cell is determined by a positioning algorithm. The adhesive film layer is cut to a size that matches the battery cell, and the adhesive film layer is laid on the surface of the glass enamel layer of the front glass layer by a robotic arm. The battery cell is then placed in the preset position of the adhesive film layer, and the position of the battery cell is fixed by a vacuum adsorption device, so that the battery cell and the adhesive film layer are completely bonded. During the bonding process, the bonding pressure is monitored by a pressure sensor to ensure that the pressure is evenly distributed.

[0114] In some embodiments, a lower polymer adhesive film layer is laid on the surface of the battery cell facing away from the upper adhesive film layer. A laser rangefinder is used to measure the distance between the backsheet glass layer and the battery cell. A closed-loop control system is used to adjust the descent speed and angle of the backsheet glass layer to align the backsheet glass layer with the lower polymer adhesive film layer. A hot-pressing process is used to press the laid backsheet glass layer, lower polymer adhesive film layer, battery cell and front sheet glass layer together. During the hot-pressing process, temperature and pressure data are collected in real time by temperature and pressure sensors. When the temperature and pressure reach the preset temperature and pressure conditions, the conditions are maintained for a fourth preset time to complete the bonding.

[0115] In some embodiments, before preparing the glass enamel ink, spectral distribution data of the target light source is obtained using a spectral analysis device. The spectral distribution data is then input into a pre-trained light conversion material ratio optimization algorithm model. This model is constructed based on the response spectrum of the photovoltaic cell and the excitation-emission spectrum data of the light conversion material, and outputs the optimal ratio combination of green, yellow, orange, and red light conversion materials. The content of the light conversion material is then adjusted according to the optimal ratio combination.

[0116] In some embodiments, after the preparation of the novel high-efficiency light-conversion photovoltaic glass is completed, the performance parameters of the photovoltaic glass, such as short-circuit current, open-circuit voltage, and conversion efficiency, are obtained through electrical performance testing equipment. The performance parameters are then input into a preset process parameter feedback algorithm model. The process parameter feedback algorithm model analyzes the correlation between the performance parameters and the process parameters such as the composition of the glass enamel ink, printing thickness, tempering temperature, and hot pressing time during the preparation process. It outputs adjustment suggestions for the process parameters of the current batch and stores the adjustment suggestions in the process database to guide the optimization of the preparation process for subsequent batches.

[0117] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific preparation process of each module of the novel high-efficiency light conversion photovoltaic glass described above can be referred to the corresponding process in the preparation method embodiments of the novel high-efficiency light conversion photovoltaic glass described above, and will not be repeated here.

[0118] This application also provides an apparatus for preparing a novel high-efficiency light-conversion photovoltaic glass. This apparatus is used to perform the steps of the methods for preparing the novel high-efficiency light-conversion photovoltaic glass shown in the above embodiments. The apparatus can be a single server or a server cluster, or it can be a terminal, such as a handheld terminal, a laptop computer, a wearable device, or a robot.

[0119] The fabrication apparatus for novel high-efficiency light-conversion photovoltaic glass includes: The ink preparation unit is used to prepare glass enamel inks corresponding to novel high-efficiency light-conversion photovoltaic glass. This is achieved by mixing glass powder, light-conversion materials, pearlescent powder, light-diffusing particles, and an organic carrier. The light-conversion materials include green, yellow, orange, and red light-conversion materials, with a content greater than 5% and less than 30%. The light-diffusing particles have a particle size greater than 0.1 μm and less than 0.5 μm, a refractive index greater than 1.7 and less than 2.7, and include one or more of titanium dioxide, zinc oxide, aluminum oxide, and yttrium oxide, with a content greater than 0.1% and less than 5%. The organic carrier, composed of organic solvents, thickeners, and dispersants, accounts for greater than 20% and less than 40% of the total ink mass to impart printability to the ink.

[0120] The ink printing unit is used to print the prepared glass enamel ink onto the surface of the front glass layer; the front glass layer printed with glass enamel ink is tempered, and the glass enamel ink is sintered to form a glass enamel layer.

[0121] The preparation unit is used to bond the solar cell to the front glass layer with the glass glaze layer through the upper adhesive film layer, and to bond the back glass layer to the solar cell through the lower polymer adhesive film layer, thus completing the preparation of a new type of high-efficiency light conversion photovoltaic glass.

[0122] It should be noted that those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the above-described fabrication apparatus and modules for the novel high-efficiency light-conversion photovoltaic glass can be referred to the corresponding processes in the embodiments of the above-described methods for fabricating the novel high-efficiency light-conversion photovoltaic glass, and will not be repeated here.

[0123] The above-mentioned method for preparing the novel high-efficiency light-conversion photovoltaic glass can be implemented as a computer program that can run on the provided device.

[0124] Please see Figure 3 , Figure 3 This is a schematic block diagram of the electrical control box provided in an embodiment of this application. The electrical control box includes a processor, a memory, and a network interface connected via a device bus, wherein the memory may include a storage medium and internal memory.

[0125] The storage medium can store operating devices and computer programs. The computer program includes program instructions that, when executed, cause the processor to perform any novel method for preparing high-efficiency light-conversion photovoltaic glass.

[0126] The processor provides computing and control capabilities to support the operation of the entire electrical control box.

[0127] Internal memory provides an environment for the execution of computer programs in non-volatile storage media. When executed by a processor, the computer program can enable the processor to carry out any new method for preparing high-efficiency light-converting photovoltaic glass.

[0128] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the terminal to which the present application is applied. The specific electrical control box may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0129] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0130] In one embodiment, the processor is configured to run a computer program stored in memory to perform the following steps: A novel high-efficiency light-conversion photovoltaic glass-based colored enamel ink is prepared by mixing glass powder, light-conversion materials, pearlescent powder, light-diffusing particles, and an organic carrier. The light-conversion materials include green, yellow, orange, and red light-conversion materials, with a content greater than 5% and less than 30%. The light-diffusing particles have a particle size greater than 0.1 μm and less than 0.5 μm, a refractive index greater than 1.7 and less than 2.7, and include one or more of titanium dioxide, zinc oxide, aluminum oxide, and yttrium oxide, with a content greater than 0.1% and less than 5%. The organic carrier, composed of organic solvent, thickener, and dispersant, accounts for greater than 20% and less than 40% of the total ink mass to impart printability to the ink.

[0131] The prepared glass enamel ink is printed onto the surface of the front glass layer; the front glass layer printed with glass enamel ink is tempered, and the glass enamel ink is sintered to form a glass enamel layer.

[0132] The solar cells are bonded to the front glass layer with a glass glaze layer by an upper adhesive film layer, and the back glass layer is bonded to the solar cells by a lower polymer adhesive film layer, thus completing the preparation of a new type of high-efficiency light conversion photovoltaic glass.

[0133] In some embodiments, the green light conversion material includes Lu3(Al,Ga)5O 12 :Ce 3+ (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ β-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+Mixed with green pearlescent pigment; the yellow light-converting material comprises (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ Mixed with yellow pearlescent pigments; orange light-converting materials include α-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+ Mixed with orange pearlescent pigments; red light-converting materials include (Sr,Ca)AlSiN3:Eu 2+ and (Sr,Ca)2Si5N8:Eu 2+ When mixed with red pearlescent pigments, metallic pearlescent powders such as gray, silver, gold, and copper can be added for special color requirements.

[0134] In some embodiments, the preparation of the glass enamel ink corresponding to the novel high-efficiency light-conversion photovoltaic glass involves mixing glass powder, light-conversion material, pearlescent powder, light-diffusing particles, and an organic carrier. This includes: determining the mixing ratio of glass powder, light-conversion material, pearlescent powder, light-diffusing particles, and organic carrier using a preset machine learning model, wherein the machine learning model is trained based on historical mixing data and corresponding glass enamel ink performance parameters; and, according to the determined mixing ratio, first adding glass powder, light-conversion material, and light-diffusing particles to a mixing container, stirring at a first preset speed for a first preset time, then adding the organic carrier, and continuing stirring at a second preset speed for a second preset time to form a uniform glass enamel ink.

[0135] In some embodiments, printing the prepared glass enamel ink onto the surface of the front glass layer includes: acquiring surface contour data of the front glass layer through an image recognition system; generating a printing path based on the surface contour data using a path planning algorithm; printing according to the generated printing path using a screen printing device; monitoring the printing pressure in real time through a pressure sensor during the printing process; and automatically adjusting the pressure of the printing squeegee according to a preset pressure threshold range through a control system.

[0136] In some embodiments, the tempering process of the front glass layer printed with glass enamel ink and the sintering of the glass enamel ink to form a glass enamel layer includes: placing the printed front glass layer into a tempering furnace, collecting furnace temperature data in real time through a temperature sensor, calculating the heating rate and holding time at the current temperature using a preset neural network model, wherein the input parameters of the neural network model include the component ratio of the glass enamel ink, the thickness and size of the front glass layer; heating the tempering furnace to a preset temperature range according to the calculated heating rate, holding it at the temperature for a third preset time, and then cooling it to room temperature at a preset cooling rate, so that the glass enamel ink is sintered and forms a bonding layer with the front glass layer.

[0137] In some embodiments, the process of bonding the battery cell to the front glass layer with the glass enamel layer via the adhesive film layer includes: obtaining the position coordinates of the front glass layer with the glass enamel layer using a machine vision system, determining the bonding position of the battery cell using a positioning algorithm; cutting the adhesive film layer to a size that matches the battery cell, laying the adhesive film layer on the surface of the glass enamel layer of the front glass layer using a robotic arm, placing the battery cell in a preset position on the adhesive film layer, fixing the position of the battery cell using a vacuum adsorption device, so that the battery cell and the adhesive film layer are completely bonded, and monitoring the bonding pressure through a pressure sensor during the bonding process to ensure uniform pressure distribution.

[0138] In some embodiments, the process of bonding the backsheet glass layer to the battery cell via the lower polymer adhesive film layer includes: laying the lower polymer adhesive film layer on the surface of the battery cell facing away from the upper adhesive film layer; measuring the distance between the backsheet glass layer and the battery cell using a laser rangefinder; adjusting the descent speed and angle of the backsheet glass layer through a closed-loop control system to align the backsheet glass layer with the lower polymer adhesive film layer; and using a hot-pressing process to press the laid backsheet glass layer, lower polymer adhesive film layer, battery cell, and front sheet glass layer together. During the hot-pressing process, temperature and pressure data are collected in real time using temperature and pressure sensors. When the temperature and pressure reach preset conditions, these conditions are maintained for a fourth preset time to complete the bonding.

[0139] In some embodiments, the method further includes: before preparing the glass enamel ink, acquiring spectral distribution data of the target light source using a spectral analysis device, inputting the spectral distribution data into a pre-trained light conversion material ratio optimization algorithm model, wherein the light conversion material ratio optimization algorithm model is constructed based on the response spectrum of the photovoltaic cell and the excitation-emission spectrum data of the light conversion material, and outputting the optimal ratio combination of green light conversion material, yellow light conversion material, orange light conversion material and red light conversion material; and adjusting the content of the light conversion material according to the optimal ratio combination.

[0140] In some embodiments, the method further includes: after completing the preparation of the novel high-efficiency light conversion photovoltaic glass, obtaining the performance parameters of the photovoltaic glass such as short-circuit current, open-circuit voltage, and conversion efficiency through an electrical performance testing device, inputting the performance parameters into a preset process parameter feedback algorithm model, wherein the process parameter feedback algorithm model analyzes the correlation between the performance parameters and the process parameters such as the composition of the glass enamel ink, printing thickness, tempering temperature, and hot pressing time during the preparation process, outputting adjustment suggestions for the process parameters of the current batch, and storing the adjustment suggestions in the process database to guide the optimization of the preparation process for subsequent batches.

[0141] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the method for preparing the novel high-efficiency light-conversion photovoltaic glass as described in the first aspect above.

[0142] The computer-readable storage medium can be an internal storage unit of the electrical control box described in the foregoing embodiments, such as the hard drive or memory of the electrical control box. Alternatively, the computer-readable storage medium can be an external storage device of the electrical control box, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the electrical control box.

[0143] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for preparing a novel high-efficiency light-conversion photovoltaic glass, characterized in that, The method includes: A novel high-efficiency light-conversion photovoltaic glass-based colored enamel ink is prepared by mixing glass powder, light-conversion materials, pearlescent powder, light-diffusing particles, and an organic carrier. The light-conversion materials include green, yellow, orange, and red light-conversion materials, with a content greater than 5% and less than 30%. The pearlescent powder has a particle size greater than 5 μm and less than 20 μm, with a content greater than 1% and less than 10%. The light-diffusing particles have a particle size greater than 0.1 μm and less than 0.5 μm and a refractive index greater than 1.7 and less than 2.7, and include one or more of titanium dioxide, zinc oxide, aluminum oxide, and yttrium oxide, with a content greater than 0.1% and less than 5%. The organic carrier, composed of organic solvent, thickener, and dispersant, accounts for greater than 20% and less than 40% of the total ink mass to impart printability to the ink. The prepared glass enamel ink is printed onto the surface of the front glass layer; the front glass layer printed with glass enamel ink is tempered to sinter the glass enamel ink to form a glass enamel layer. The solar cells are bonded to the front glass layer with a glass glaze layer by an upper adhesive film layer, and the back glass layer is bonded to the solar cells by a lower polymer adhesive film layer, thus completing the preparation of a new type of high-efficiency light conversion photovoltaic glass.

2. The method according to claim 1, characterized in that, The green light conversion material includes Lu3(Al,Ga)5O 12 :Ce 3+ (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ β-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+ The yellow light-converting material comprises (Y,Ga)3(Al,Ga)5O 12 :Ce 3+ Orange light conversion materials include α-sialon:Eu 2+ and (Sr,Ca)2SiO4:Eu 2+ Red light conversion materials include (Sr,Ca)AlSiN3:Eu 2+ and (Sr,Ca)2Si5N8:Eu 2+ .

3. The method according to claim 1, characterized in that, The glass enamel ink for preparing novel high-efficiency light-conversion photovoltaic glass is prepared by mixing glass powder, light-conversion material, pearlescent powder, light-diffusing particles, and organic carrier, including: The mixing ratio of glass powder, light conversion material, pearlescent powder, light diffusing particles and organic carrier is determined using a pre-set machine learning model, which is trained based on historical mixing data and corresponding glass enamel ink performance parameters. According to the determined mixing ratio, glass powder, light conversion material, pearl powder and light diffusion particles are first added to the mixing container and stirred at a first preset speed for a first preset time. Then, an organic carrier is added and stirred at a second preset speed for a second preset time to form a uniform glass enamel ink.

4. The method according to claim 1, characterized in that, The step of printing the prepared glass enamel ink onto the surface of the front glass layer includes: The surface contour data of the front glass layer is obtained through an image recognition system, and a printing path is generated based on the surface contour data using a path planning algorithm. The screen printing equipment is used to print along the generated printing path. During the printing process, the printing pressure is monitored in real time by a pressure sensor, and the pressure of the printing squeegee is automatically adjusted by the control system according to the preset pressure threshold range.

5. The method according to claim 1, characterized in that, The process of tempering the front glass layer printed with glass enamel ink and sintering the glass enamel ink to form a glass enamel layer includes: The printed front glass layer is placed in a tempering furnace, and the temperature data inside the furnace is collected in real time by a temperature sensor. The heating rate and holding time at the current temperature are calculated using a preset neural network model. The input parameters of the neural network model include the composition ratio of the glass enamel ink, the thickness and size of the front glass layer. The tempering furnace heats the glass to a preset temperature range according to the calculated heating rate, holds it at that temperature for a third preset time, and then cools it to room temperature at a preset cooling rate, so that the glass enamel ink is sintered and forms a bonding layer with the front glass layer.

6. The method according to claim 1, characterized in that, The process of bonding the battery cells to the front glass layer with the glass enamel layer via an adhesive film layer includes: The machine vision system is used to obtain the position coordinates of the front glass layer with the glass glaze layer, and the bonding position of the battery cell is determined by the positioning algorithm. The adhesive film layer is cut to the size that matches the battery cell. The adhesive film layer is then laid on the surface of the glass enamel layer of the front glass layer using a robotic arm. The battery cell is then placed in the preset position of the adhesive film layer and fixed in position by a vacuum adsorption device, so that the battery cell and the adhesive film layer are completely bonded. During the bonding process, the bonding pressure is monitored by a pressure sensor to ensure that the pressure is evenly distributed.

7. The method according to claim 1, characterized in that, The process of bonding the backsheet glass layer to the battery cell via a lower polymer adhesive film layer includes: A lower polymer adhesive film layer is laid on the surface of the solar cell away from the upper adhesive film layer. A laser rangefinder is used to measure the distance between the backsheet glass layer and the solar cell. The descent speed and angle of the backsheet glass layer are adjusted through a closed-loop control system to align the backsheet glass layer with the lower polymer adhesive film layer. The back glass layer, lower polymer film layer, battery cell and front glass layer are pressed together by hot pressing process. During hot pressing, temperature and pressure data are collected in real time by temperature sensor and pressure sensor. When the temperature and pressure reach the preset temperature condition and the pressure reach the preset pressure condition, the conditions are maintained for a fourth preset time to complete the bonding.

8. The method according to claim 1, characterized in that, The method further includes: Before preparing the glass enamel ink, the spectral distribution data of the target light source is obtained using a spectral analysis device. The spectral distribution data is then input into a pre-trained light conversion material ratio optimization algorithm model. This light conversion material ratio optimization algorithm model is constructed based on the response spectrum of photovoltaic cells and the excitation-emission spectrum data of light conversion materials, and outputs the optimal ratio combination of green, yellow, orange, and red light conversion materials. Adjust the content of the light conversion material according to the optimal ratio combination.

9. The method according to claim 1, characterized in that, The method further includes: After the fabrication of the novel high-efficiency photovoltaic glass is completed, the short-circuit current, open-circuit voltage, and conversion efficiency of the photovoltaic glass are obtained through electrical performance testing equipment. The performance parameters are then input into a preset process parameter feedback algorithm model. The process parameter feedback algorithm model analyzes the correlation between the performance parameters and the process parameters such as the composition of the glass enamel ink, printing thickness, tempering temperature, and hot pressing time during the fabrication process. It outputs adjustment suggestions for the process parameters of the current batch and stores the adjustment suggestions in the process database to guide the optimization of the fabrication process for subsequent batches.

10. A novel high-efficiency light-conversion photovoltaic glass, characterized in that, Prepared by the method according to any one of claims 1-9.

Images