A cleaning method and system based on a topcon cell manufacturing process

Abstract

This invention relates to a cleaning method and system based on TOPCon cell manufacturing process, belonging to the technical field of solar cells. The method includes: acquiring the manufacturing process specifications of TOPCon cells; performing texturing, boronizing, oxidation, and laser cutting on silicon wafers according to the manufacturing process specifications to form inspection silicon wafers; performing borosilicate glass removal, alkaline polishing, and tunneling on the inspection silicon wafers and acquiring inspection information for each inspection silicon wafer; adding additives to a cleaning solution and obtaining baseline information based on the manufacturing process specifications; comparing the consistency between the inspection information and the baseline information to obtain anomaly information; obtaining a cleaning control signal based on the anomaly information and outputting the cleaning control signal to place the inspection silicon wafers in a cleaning solution containing additives for cleaning; after cleaning, continuing to perform atomic layer deposition, positive film, back film, screen printing, and testing and sorting on each inspection silicon wafer to obtain the cell. This application has the effect of improving the production efficiency of TOPCon cells.

Description

Technical Field

[0001] This invention relates to the technical field of solar cells, and in particular to a cleaning method and system based on the TOPCon cell manufacturing process. Background Technology

[0002] TOPCon cells refer to tunnel oxide passivated contact cells, which are widely used in the modern photovoltaic industry due to their high efficiency and good electrical performance.

[0003] In the manufacturing process of TOPCon cells, in order to optimize the output power of the module, reduce shading loss and hot spot effect, laser cutting technology is usually used to divide the whole cell into small units such as half cells and shingled cells. Laser cutting will generate surface defects, lattice distortion and dangling bonds in the cutting area. These defects become carrier recombination centers, resulting in a decrease in the photoelectric conversion efficiency of the cell. Independent passivation equipment is usually used to perform edge passivation treatment to repair the edge defects caused by laser cutting.

[0004] In the manufacturing process of TOPCon batteries, when TOPCon batteries need to be mass-produced, the speed of using independent passivation equipment to passivate the edges of the laser-cut batteries is slow, which reduces the production efficiency of TOPCon batteries. Summary of the Invention

[0005] To improve the production efficiency of TOPCon batteries, this invention provides a cleaning method and system based on the TOPCon battery manufacturing process.

[0006] In a first aspect, the present invention provides a cleaning method based on the TOPCon battery manufacturing process, employing the following technical solution: A cleaning method based on TOPCon battery manufacturing process, comprising: Collect the manufacturing process specifications of TOPCon batteries; According to the manufacturing process specifications, silicon wafers are texturized, boron expanded, oxidized, and laser-cut to form various test silicon wafers; Each silicon wafer undergoes borosilicate glass removal, alkaline polishing, and tunneling, and the test information for each wafer is collected. Pre-set additives are added to the cleaning solution, and the benchmark information corresponding to the test information is obtained based on the manufacturing process specifications. Anomaly information is obtained by comparing the consistency between the detection information and the baseline information; Based on the abnormal information, a cleaning control signal is obtained, and the cleaning control signal is output to place the detected silicon wafer into a cleaning solution containing additives for cleaning. After cleaning, each silicon wafer undergoes atomic layer deposition, positive film deposition, back film deposition, screen printing, and testing and sorting to produce a battery.

[0007] By adopting the above technical solution, abnormal information is obtained by comparing and analyzing the detection information after the previous process with the benchmark information. Based on the abnormal information, a cleaning control signal is generated, which enables silicon wafers to be placed in a cleaning solution containing additives for differentiated cleaning as needed. There is no need for additional independent passivation equipment. The cleaning effect of each silicon wafer is accurately matched by the cleaning control signal directly reflected by the abnormal information, which shortens the process cycle loss caused by equipment passivation when there is high demand for cells and improves the production efficiency of TOPCon cells. At the same time, by using preset additives, polycrystalline silicon winding plating removal, laser cutting area texturing and surface passivation are integrated in the cleaning process, which solves the problem that existing RCA cleaning cannot complete multiple functions at the same time, and effectively compensates for the risk of uneven cleaning caused by abnormalities such as uneven thickness and damage differences in the previous process.

[0008] Optionally, methods for obtaining cleaning control signals include: The installation position of the megasonic vibration plate in the cleaning tank was measured. Retrieve reference acoustic parameters from manufacturing process specifications; The acoustic coverage range is obtained based on the reference acoustic parameters and the installation location. Compare the size of each abnormal information and sort them to obtain the arrangement parameters from largest to smallest; Based on the arrangement parameters, the detection arrangement parameters are obtained through preset flower basket specifications, sound wave coverage range, and abnormal information; The maximum abnormal information is retrieved from the permutation parameters and used as the marked abnormal information. Based on the anomaly information and reference acoustic parameters, the actual installation position of each detected silicon wafer in the basket is obtained, and the actual installation position is used as the cleaning control signal.

[0009] By adopting the above technical solution, the acoustic coverage range and detection arrangement parameters are obtained by analyzing the installation position, reference acoustic parameters and abnormal information. Then, the largest abnormal information is used as the marker abnormal information to infer the actual installation position of each silicon wafer. This allows silicon wafers with high degree of abnormality to be preferentially arranged in the area with the strongest energy within the acoustic coverage range, reducing insufficient cleaning or over-corrosion caused by fixed positions, and improving cleaning uniformity and batch consistency.

[0010] Optionally, methods for obtaining the actual installation location include: The maximum cleaning time is obtained based on the marked anomaly information and the reference acoustic parameters; The acoustic parameters corresponding to each abnormality are obtained based on the detection arrangement parameters and the acoustic coverage range. Baseline cleaning information is obtained by using acoustic parameters and maximum cleaning time; Compare the abnormal information with the baseline cleanup information and identify any abnormal information that does not exceed the baseline cleanup information as adjustment abnormal information; The actual installation location is obtained based on the adjustment anomaly information and the baseline cleanup information.

[0011] By adopting the above technical solution, the maximum cleaning time is determined by marking abnormal information and reference acoustic parameters. Then, the reference cleaning information is calculated based on the acoustic parameters corresponding to each abnormal information. Abnormal information that does not exceed the reference cleaning information is used as adjustment abnormal information and its actual installation position is obtained. Thus, under the constraint of fixed time, the most suitable vertical position can be found for each silicon wafer, so that the cleaning capacity and cleaning requirements are accurately matched, reducing process fluctuations caused by blind adjustments and improving the robustness of the cleaning process.

[0012] Optionally, methods for obtaining the actual installation location include: The difference between the adjustment anomaly information and the baseline cleanup information is calculated as the adjustment excess information; The acoustic parameters can be adjusted by inversely deriving the information beyond the maximum cleaning time. The vertical adjustment distance is obtained by adjusting the acoustic wave parameters; The actual installation location is determined based on the flower basket specifications, vertical adjustment distance, and any abnormal adjustment information.

[0013] By adopting the above technical solution, the vertical adjustment distance is obtained by calculating and adjusting the excess information, thereby converting the additional energy compensation required by the silicon wafer into a specific vertical displacement command, realizing continuous adjustable position control, and reducing energy waste or insufficient compensation caused by discrete grading of sound waves.

[0014] Optionally, other methods for obtaining the actual installation location include: Retrieve the thickness distribution of a single silicon wafer from the anomaly information; The average thickness value is calculated based on the detected thickness distribution, and the detected thickness distribution is divided into areas with excessive thickness and areas with insufficient thickness based on the average thickness value. The placement orientation of the silicon wafer within the basket is determined based on the relative orientation between the thickness-exceeding and thickness-deficient areas. The processing time is obtained based on the thickness exceeding the region, the thickness thinning region, and the reference acoustic wave parameters. The total processing time is obtained by adjusting the distance based on the processing time of the two regions; By comparing the total processing time with the maximum cleaning time, the target adjustment distance and its corresponding target adjustment time point are obtained, and the actual installation position is updated in real time based on the target adjustment distance and its corresponding target adjustment time point.

[0015] By adopting the above technical solution, by analyzing the thickness distribution of a single silicon wafer, dividing the thickness excess area and the weak area, and determining the placement direction of the silicon wafer in the basket, and then dynamically adjusting the vertical distance to ensure that the two areas are cleaned simultaneously, the problem of unevenness within the same silicon wafer surface can be solved, and the consistency of in-plane cleaning at the single wafer level can be achieved.

[0016] Optional, also includes: By comparing the abnormal information with the baseline cleanup information, the abnormal information that exceeds the baseline cleanup information is taken as the target abnormal information. The difference between the target anomaly information and the baseline cleanup information is calculated as the target remaining amount of the target anomaly information; Supplementary acoustic parameters are obtained based on the target remaining amount and the maximum cleaning time; The sweep frequency parameters are obtained by supplementing the acoustic wave parameters and the reference acoustic wave parameters, and then the sweep frequency parameters are added to the cleaning control information.

[0017] By adopting the above technical solution, by identifying target anomalies that exceed the baseline cleaning information and calculating their remaining quantity to back-calculate supplementary acoustic parameters to generate sweep frequency parameters, it is possible to use sweep frequency modulation to change the spatial distribution and time-averaged energy of the sound field under vertical position adjustment, thereby providing additional energy compensation for other abnormal silicon wafers.

[0018] Optionally, methods for obtaining cleanup control information include: The reference energy distribution at each position of the detection silicon wafer in the basket is obtained based on the reference acoustic wave parameters and the frequency sweep parameters. Collect the actual energy distribution at various locations within the flower basket; The energy distribution difference is calculated by comparing the actual energy distribution with the reference energy distribution. The energy distribution gradient can be obtained by the energy distribution difference; The control information for the flower basket is obtained based on the energy distribution gradient, and the cleaning control information is updated accordingly.

[0019] By adopting the above technical solution, the energy distribution difference is obtained by comparing the actual energy distribution with the reference energy distribution, and then the energy distribution gradient is calculated. This enables online sensing of sound field distortion caused by external factors such as bubbles, temperature, and liquid concentration, providing a quantitative basis for subsequent dynamic adjustment of the basket movement, realizing closed-loop control, reducing process drift between different batches of silicon wafers, and improving the consistency of large-scale mass production.

[0020] Optionally, methods for obtaining and updating the flower basket control information include: When the energy distribution gradient is the preset vertical gradient, the vertical stroke is obtained based on the energy distribution gradient and the specifications of the flower basket. The baseline spacing is retrieved from the basket specifications, and the variation gradient is retrieved from the energy distribution gradient by comparing the vertical direction with the baseline spacing. The vertical reciprocating velocity is obtained based on the changing gradient; The standard position range of a normal silicon wafer is obtained by the actual installation location and energy distribution gradient, and the sweep frequency parameters are updated by the vertical reciprocating speed and the standard position range. The cleaning control information is updated using the updated sweep frequency parameters, and the vertical reciprocating speed and vertical stroke are added to the cleaning control information as basket control information.

[0021] By adopting the above technical solution, when the energy distribution gradient is a vertical gradient, the vertical stroke and vertical reciprocating speed are determined according to the energy distribution gradient, and the sweep frequency parameters are updated according to the standard position range of normal silicon wafers. In this way, the up-and-down reciprocating motion of the basket can be used to actively compensate for the energy difference between the upper and lower layers. At the same time, by adjusting the sweep frequency parameters, qualified silicon wafers can be protected, realizing the redistribution of energy in the vertical direction and reducing the unevenness caused by sound wave attenuation.

[0022] Optionally, methods for obtaining flower basket control information and updating cleanup control information also include: When the energy distribution gradient is a preset horizontal gradient, the largest energy distribution difference is retrieved from the energy distribution gradient as the horizontal energy difference. The rotational speed is obtained based on the horizontal energy difference. The rotational excess region and its corresponding rotational excess energy are obtained by using the rotational speed and energy distribution gradient. The sweep frequency parameters are updated based on the rotation out-of-region and rotation out-of-energy. The cleaning control information is updated based on the updated sweep frequency parameters, and the rotation speed is used as the basket control information and added to the cleaning control information.

[0023] By adopting the above technical solution, when the energy distribution gradient is horizontal, the rotation speed is determined based on the horizontal energy difference, and the rotation excess area and its rotation excess energy are identified to update the sweep frequency parameters. In this way, the horizontal rotation motion of the basket can be used to make the silicon wafer pass through the high and low energy areas in turn, eliminating the horizontal non-uniformity in time. At the same time, the energy peak value is reduced for local over-strong areas by sweep frequency parameters, reducing the occurrence of new over-etching points during rotation and improving the overall cleaning uniformity.

[0024] Secondly, this application provides a cleaning system based on the TOPCon battery manufacturing process, which adopts the following technical solution: A cleaning system based on TOPCon battery manufacturing process, comprising: The acquisition module is used to acquire manufacturing process specifications; The memory is used to store a program for a cleaning method based on the TOPCon battery manufacturing process; The processor is used to load and execute programs stored in memory.

[0025] In summary, this application includes at least one of the following beneficial technical effects: 1. By comparing and analyzing the detection information and benchmark information after the previous process to obtain abnormal information, and generating cleaning control signals based on the abnormal information, silicon wafers can be placed in cleaning solutions containing additives for differentiated cleaning as needed. There is no need for additional independent passivation equipment. The cleaning effect of each silicon wafer is accurately matched by the cleaning control signals directly reflected by the abnormal information, which shortens the process cycle loss caused by equipment passivation when high-demand cells are used, and improves the production efficiency of TOPCon cells. At the same time, by using preset additives, polysilicon winding plating removal, laser cutting area texturing and surface passivation are integrated in the cleaning process, solving the problem that existing RCA cleaning cannot complete multiple functions at the same time, and effectively compensating for the risk of uneven cleaning caused by abnormalities such as uneven thickness and damage differences in the previous process. 2. By analyzing the installation location, reference acoustic parameters, and abnormal information, the acoustic coverage range and detection arrangement parameters are obtained. Then, the largest abnormal information is used as the marker abnormal information to infer the actual installation location of each silicon wafer. This allows silicon wafers with high abnormality to be preferentially arranged in the area with the strongest energy within the acoustic coverage range, reducing insufficient cleaning or over-corrosion caused by fixed positions, and improving cleaning uniformity and batch consistency. 3. When the energy distribution gradient is vertical, the vertical stroke and vertical reciprocating speed are determined based on the energy distribution gradient. The sweep frequency parameters are updated according to the standard position range of a normal silicon wafer. This allows the basket's up-and-down reciprocating motion to actively compensate for energy differences between the upper and lower layers. Simultaneously, by adjusting the sweep frequency parameters, qualified silicon wafers can be protected, achieving vertical energy redistribution and reducing unevenness caused by acoustic wave attenuation. Attached Figure Description

[0026] Figure 1 This is a flowchart of a cleaning method based on the TOPCon battery manufacturing process according to an embodiment of the present invention; Figure 2 This is a flowchart of a method for obtaining a cleaning control signal according to an embodiment of the present invention. Detailed Implementation

[0027] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0028] Reference Figure 1 This application discloses a cleaning method based on the TOPCon battery manufacturing process, including the following steps: S10: Collect the manufacturing process specifications of TOPCon batteries.

[0029] Manufacturing process specifications refer to the process steps and standard manufacturing parameters required for each step in the manufacture of TOPCon batteries, including temperature, time, and chemical concentration for each step. These manufacturing process specifications are obtained by the operator pre-inputting them into the system.

[0030] S11: According to the manufacturing process specifications, silicon wafers are texturized, boron expanded, oxidized, and laser-cut to form various test silicon wafers.

[0031] Test silicon wafers refer to silicon wafers that have completed texturing, boron diffusion, oxidation, and laser cutting processes and are awaiting subsequent cleaning. The entire silicon wafer is texturized, boron diffused, oxidized, and laser cut according to the parameters in the manufacturing process specifications, and each silicon wafer after completing this step is used as a test silicon wafer.

[0032] S12: Perform borosilicate glass removal, alkaline polishing, and tunneling on each silicon wafer and collect the test information for each silicon wafer.

[0033] The detection information refers to the thickness information of polycrystalline silicon coating on the surface of the silicon wafer. By performing borosilicate glass removal, alkaline polishing, and tunneling processes on each silicon wafer, the surface of the silicon wafer is detected by an online ellipsometry after the process is completed. The detected data is converted into thickness information as the detection information.

[0034] S13: Add the pre-set additives to the cleaning solution and obtain the benchmark information corresponding to the test information based on the manufacturing process specifications.

[0035] Additives are agents used to repair surface defects, dangling bonds, polysilicon winding, and lattice defects caused by the laser cutting process.

[0036] The additives include: textured nucleating agents (0.5% to 3%), including polyaspartic acid, polyglutamic acid, and polyvinylpyrrolidone. These polymers contain amino or carboxyl groups, which can bind to the hydroxyl groups on the silicon wafer surface via hydrogen bonds to form a uniform adsorption layer. This layer serves as nucleation sites for the etching reaction, promoting the orderly growth of the pyramid structure formed on the silicon wafer surface by the aforementioned processes. Taking polyaspartic acid as an example, it can reduce the electrode potential difference between metal ions and the silicon wafer, thereby slowing down the etching reaction rate, effectively controlling the size of the pyramid structure, and making the pyramid formation more uniform.

[0037] Textured surface corrosion inhibitors (0.1% to 2%) include surfactants such as sodium dodecylbenzenesulfonate, betaine, and imidazoline. Strongly polar groups such as sulfonate groups are adsorbed onto the silicon wafer surface via hydrogen bonds, regulating the corrosion rate differences between different crystal orientations [e.g., (100) and (111) crystal orientations], thereby achieving anisotropic corrosion and optimizing the formation of the pyramid structure. Betaine not only regulates surface tension and improves degassing, but also effectively enhances the corrosion uniformity of the silicon wafer surface, ensuring the stability of the reaction during texturing.

[0038] Microstructure modifiers (0.3% to 5%), including high-molecular-weight substances such as sodium carboxymethyl cellulose, polyethylene glycol, and glycol condensate, can bind to the silicon wafer surface through hydroxyl or ether bonds, generating a steric hindrance effect. This promotes the formation of nanoscale cracks or microstructures on the sides of the pyramid, enhancing the light-trapping effect. These modifiers can further refine the pyramid structure (e.g., to 0.5 μm to 2.0 μm), effectively reducing reflectivity and improving photoelectric conversion efficiency.

[0039] Nucleation aids (0.1% to 1%), including organic acid compounds such as acetic acid and citric acid, can further regulate the nucleation and growth rates, thereby optimizing the size of the pyramid structure. Their role is to promote the uniformity of the nucleation process, effectively refine the pyramid structure, and improve the stability of the flocking effect.

[0040] Defoaming agents (0.2% to 5%), such as polyvinyl alcohol, can quickly remove hydrogen gas generated during alkaline etching from the silicon wafer surface, preventing bubbles from adhering and thus avoiding uneven cleaning caused by bubbles. The addition of defoaming agents helps ensure the uniformity of the reaction and further stabilizes the texturing process, ensuring the uniformity of the silicon wafer surface quality.

[0041] Accelerators (2% to 20%), such as ammonium persulfate, can accelerate the removal process of poly-Si (polycrystalline silicon) and improve removal efficiency. Commonly used accelerators include tannic acid, gallic acid, nicotinic acid, and tartaric acid, which can effectively increase the dissolution rate of poly-Si, shorten the time required to remove poly-Si, and ensure that the poly-Si layer is completely removed.

[0042] Composite surfactants (1% to 5%) include anionic, nonionic, and gemini surfactants (such as lauryl ethoxylate carboxylate). These composite surfactants can form a dense protective film on the silicon wafer surface, promote the wetting and penetration of the cleaning solution, prevent excessive corrosion, and effectively protect the silicon wafer surface during poly-Si removal, reducing surface damage and ensuring high-quality silicon wafer surface.

[0043] Complexing agents (0.2% to 5%), such as ethylenediaminetetramethylenephosphonic acid, can bind with metal ions in the cleaning solution, preventing these metal ions from being re-adsorbed onto the silicon wafer surface. This avoids localized contamination or electrochemical corrosion, effectively improving the cleanliness of the silicon wafer surface after cleaning and providing a clean and stable surface for subsequent passivation processes.

[0044] Adding additives to the cleaning solution enables the simultaneous achievement of multiple objectives—selective removal, uniform etching, and efficient protection—during both poly-Si removal and texturing processes in RCA (Removal Processing), a fundamental wet process for silicon wafer fabrication. The various components in the additive work synergistically through different mechanisms to ensure effective protection of the silicon wafer surface throughout the process, providing an ideal surface structure and stability for subsequent texturing after poly-Si removal. This additive formulation significantly improves photoelectric conversion efficiency and enhances the uniformity and stability of the texturing process, providing strong support for the efficient manufacturing of TOPCon cells.

[0045] Reference information refers to the standard thickness information that appears on the surface of a silicon wafer after the texturing to tunneling process steps. The reference information is obtained by retrieving the standard thickness information that appears on the surface of a silicon wafer after the texturing to tunneling process steps from the manufacturing process specifications.

[0046] S14: Obtain anomaly information by comparing the consistency between the detection information and the baseline information.

[0047] Abnormal information refers to the deviation value of the polysilicon coating thickness on the silicon wafer surface based on the standard. By comparing the detection information with the reference information, the analysis is carried out. When the detection information is consistent with the reference information, it means that there is no abnormal deviation in the polysilicon coating thickness on the silicon wafer surface, and no adjustment is made.

[0048] When the detection information is inconsistent with the reference information, it indicates that there is an abnormal deviation in the polysilicon coating thickness on the silicon wafer surface. The difference between the detection information and the reference information is then calculated as the abnormal information.

[0049] S15: Based on the abnormal information, a cleaning control signal is obtained, and a cleaning control signal is output to place the detected silicon wafer into a cleaning solution containing additives for cleaning.

[0050] A cleaning tank is a tank used to hold cleaning solutions containing additives.

[0051] The cleaning control signal refers to the control signal command used to control the movement of the basket in the cleaning tank and the operation of the megasonic wave vibrator in the cleaning tank. The cleaning control signal is obtained by analyzing abnormal information and then output to place the silicon wafer in a cleaning solution containing additives for cleaning. The specific analysis method for the cleaning control signal is described below.

[0052] S16: After cleaning, each silicon wafer will undergo atomic layer deposition, positive film, back film, screen printing, and testing and sorting to obtain the battery.

[0053] After cleaning, each silicon wafer is subjected to atomic layer deposition, positive film, back film, screen printing, and testing and sorting processes according to the parameters in the manufacturing process specifications to obtain TOPCon cells.

[0054] Reference Figure 2 Methods for obtaining cleaning control signals include: S20: Collect the installation position of the megasonic vibration plate in the cleaning tank.

[0055] The installation position refers to the coordinate position of the megasonic wave vibrator plate installed in the cleaning tank. It is obtained through mechanical calibration during the installation of the megasonic wave vibrator plate and stored in the system's parameter table for retrieval.

[0056] S21: Retrieve reference acoustic parameters from the manufacturing process specifications.

[0057] Reference acoustic parameters refer to the basic mega-sonic settings required for the cleaning effect to reach the reference information, including the center frequency (e.g., 1.0MHz) and output power density (e.g., 2W / cm²). 2 (This is achieved by retrieving reference acoustic parameters from the manufacturing process specifications).

[0058] S22: Obtain the sound wave coverage range based on the reference sound wave parameters and installation location.

[0059] The acoustic wave coverage range refers to the effective spatial area of ​​mega-sound waves in the cleaning fluid. The reference acoustic wave parameters and installation position are input into the sound field simulation software (COMSOL Multiphysics, ANSYS) to calculate the effective spatial area of ​​mega-sound waves in the cleaning tank as the acoustic wave coverage range.

[0060] S23: Compare the size of each abnormal information and sort them to obtain the arrangement parameters from largest to smallest.

[0061] The sorting parameters refer to the sequence parameters after sorting the anomaly information according to their values. For example, the silicon wafer with the largest deviation is numbered 1st, the second largest is numbered 2nd, and so on. The sorting parameters are obtained by the system executing a sorting algorithm on each anomaly information.

[0062] S24: Based on the arrangement parameters, the detection arrangement parameters are obtained through preset basket specifications, sound wave coverage range, and abnormal information.

[0063] The specifications for the flower basket are the structural parameters set by the technicians, including the total height of the flower basket (e.g., 150mm), the number of slots that can hold silicon wafers (e.g., 100 slots, 4×25 structure), and the vertical position coordinates of each slot.

[0064] The detection arrangement parameters refer to the result of mapping the sorted abnormal information to the vertical positions of the basket slots. According to the law that the sound wave intensity decreases with increasing distance, each detection silicon wafer is placed in the basket according to the sequence. For example, the silicon wafer with the largest abnormal information is placed in the vertical position with the highest sound intensity within the sound wave coverage area (the slot with the smallest straight-line distance between the slot coordinates on the basket and the installation position). The next largest abnormal value is placed in the position with the second smallest spacing, and so on. The information of the corresponding slot position of each abnormal information in the basket is used as the detection arrangement parameters.

[0065] S25: The maximum abnormal information is retrieved by detecting the arrangement parameters and used as the marked abnormal information.

[0066] Labeling anomalies refers to detecting the anomaly with the largest value among the permutation parameters, and using the anomaly with the largest value among the permutation parameters as the labeling anomaly.

[0067] S26: Based on the anomaly information and reference acoustic parameters, the actual installation position of each detected silicon wafer in the basket is obtained, and the actual installation position is used as the cleaning control signal.

[0068] The actual installation position refers to the coordinate position of the silicon wafer actually installed in the slot of the basket. In this embodiment, the coordinates of the slot are the center coordinates of the bottom of the slot where the silicon wafer is placed. The actual installation position is obtained by analyzing the marked abnormal information and the reference acoustic parameters, and is used as the cleaning control signal. The specific analysis method for the actual installation position is described below.

[0069] Methods for obtaining the actual installation location include: S30: The maximum cleaning time is obtained based on the marked anomaly information and the reference acoustic parameters.

[0070] The maximum cleaning time refers to the time required for the silicon wafer corresponding to the marked abnormal information (the worst silicon wafer) to be cleaned to the reference information under the reference acoustic parameters. By establishing an experimental model in advance, the cleaning time relationship of silicon wafers with different degrees of abnormality under the reference acoustic wave is measured and fitted into a function: time = f(abnormal information). Then, the time calculated by substituting the marked abnormal information into the function model is taken as the maximum cleaning time.

[0071] S31: Obtain the acoustic parameters corresponding to each abnormality based on the detection arrangement parameters and acoustic coverage range.

[0072] Acoustic parameters refer to the megaacoustic power density (unit: W / cm²) of the acoustic waves covering each silicon wafer when the basket is immersed in the cleaning solution at a fixed position after the detection and arrangement parameters are set. 2(or equivalent sound pressure) is obtained by using a basket to obtain the coordinate position of each silicon wafer in the cleaning tank according to the detection arrangement parameters, and then the parameters of the mega-sound waves received by each silicon wafer are retrieved from the sound wave coverage range as sound wave parameters.

[0073] S32: Obtain baseline cleaning information by using acoustic parameters and maximum cleaning time.

[0074] The baseline cleaning information refers to the thickness of polysilicon that can be processed on each silicon wafer surface under the conditions of acoustic parameters and maximum cleaning time. The final removal amount is calculated by integrating the product of acoustic power density and time using a cleaning kinetic model as the baseline cleaning information.

[0075] S33: Compare the abnormal information with the baseline cleanup information and take the abnormal information that does not exceed the baseline cleanup information as the adjustment abnormal information.

[0076] Adjustment anomaly information refers to the anomaly information corresponding to the silicon wafer that has been over-cleaned under the conditions of acoustic wave parameters and maximum cleaning time. Since there are other silicon wafers with large thickness deviations, the cleaning time must reach the maximum cleaning time. Therefore, adjustment anomaly information may be found when cleaning has been completed but the wafer is still subjected to mega-acoustic waves, which may easily lead to over-exertion and damage to the silicon wafer surface. Therefore, by analyzing the deviation of the anomaly information from the baseline cleaning information, the anomaly information that does not exceed the baseline cleaning information is regarded as adjustment anomaly information.

[0077] S34: The actual installation location is obtained based on the adjustment anomaly information and the baseline cleanup information.

[0078] The actual installation location of the silicon wafer corresponding to the adjustment anomaly information is obtained by analyzing the adjustment anomaly information and the baseline cleanup information. The specific analysis method for the actual installation location is described below.

[0079] Methods for obtaining the actual installation location include: S40: Calculate the difference between the adjustment anomaly information and the baseline cleanup information as the adjustment overrun information.

[0080] Adjustment excess information refers to the deviation between adjustment anomaly information and baseline cleanup information. The difference between adjustment anomaly information and baseline cleanup information is used as adjustment excess information.

[0081] S41: Adjust the acoustic parameters by adjusting the information beyond the limit and the maximum cleaning time.

[0082] Adjusting the acoustic parameters refers to adjusting the megaacoustic power density (W / cm²) required to ensure that the silicon wafer reaches the reference information precisely within the maximum cleaning time. 2(or equivalent sound pressure). Referring to S30, the sound wave parameters calculated by inputting the adjustment excess information and the maximum cleaning time into the function model are used as the adjustment sound wave parameters.

[0083] S42: The vertical adjustment distance is obtained based on the adjusted acoustic parameters.

[0084] The vertical adjustment distance refers to the distance that the silicon wafer needs to be adjusted in the vertical direction. It is obtained by substituting the adjustment acoustic wave parameters into the acoustic power density-vertical distance mapping curve, and the resulting distance is used as the vertical adjustment distance. The acoustic power density-vertical distance mapping curve is obtained by measuring or simulating different acoustic wave data with vertical distance using a hydrophone, and is expressed as the percentage increase in power density for every 1mm decrease in distance.

[0085] S43: The actual installation position is obtained based on the flower basket specifications, vertical adjustment distance, and adjustment anomaly information.

[0086] By retrieving the slot coordinates of the abnormal information from the basket specifications, and using the slot coordinates as a reference, the actual installation position is determined by moving the vertical adjustment distance away from the megasonite vibrating plate.

[0087] Other methods for obtaining the actual installation location include: S50: Retrieve the detection thickness distribution of a single silicon wafer from the anomaly information.

[0088] The thickness distribution refers to the polysilicon coating thickness values ​​at different locations on the same silicon wafer surface. By retrieving the thickness distribution of a single silicon wafer from the anomaly information, this distribution reflects the non-uniformity of the silicon wafer surface. In this embodiment, the thickness distribution is measured rapidly offline or online before cleaning, rather than being scanned in real-time during the cleaning process.

[0089] S51: Calculate the average thickness value according to the detected thickness distribution, and divide the detected thickness distribution into areas with excessive thickness and areas with weak thickness based on the average thickness value.

[0090] The average thickness value refers to the arithmetic mean of the thickness of all detection points on the silicon wafer being tested. It is obtained by extracting the thickness of each detection point from the thickness distribution and calculating the sum of their thicknesses. The quotient of the sum of the thicknesses and the total number of detection points is then used as the average thickness value.

[0091] The thickness excess area refers to a continuous area on the same silicon wafer surface where the thickness is greater than the average thickness value, while the thickness deficiency area refers to a continuous area on the same silicon wafer surface where the thickness is less than the average thickness value. The thickness excess area and the thickness deficiency area are obtained by dividing the detected thickness distribution into regions based on the average thickness value.

[0092] S52: The placement direction of the silicon wafer in the basket is obtained based on the relative direction between the thickness-exceeding area and the thickness-deficient area.

[0093] The placement direction refers to the direction in which the silicon wafer is placed in the slot of the basket. In this embodiment, the silicon wafer is a quasi-square with rounded corners, and its placement direction has only four directions (making the square side parallel to the bottom of the slot). A cross coordinate system is set for the four directions. The lines of the cross coordinate system are formed by connecting the center points of each parallel side of the silicon wafer. By taking the center of the thickness-exceeding area as the basis, connecting it with the center of the thickness-deficient area, and then mapping the connecting line to the center point of the silicon wafer, the angle between the connecting line and the cross line is calculated. The two directions corresponding to the smallest angle are selected as candidate directions, and the candidate direction of the thickness-exceeding area that is closer to the bottom of the slot is selected as the placement direction.

[0094] S53: Processing time is obtained based on the thickness exceeding the region, the thickness thin region, and the reference acoustic parameters.

[0095] The processing time refers to the time required to clean the area with excessive thickness or the area with thin thickness under the condition of the reference acoustic parameters. Referring to S30, the processing time is calculated by inputting the thickness value of the area with excessive thickness or the thickness value of the area with thin thickness into the function model along with the reference acoustic parameters.

[0096] S54: Adjust the distance based on the processing time of the two regions to obtain the total processing time.

[0097] The total processing time refers to the total time required for both the excessive and weak areas to complete normal cleaning simultaneously by dynamically changing the vertical distance between the silicon wafer and the vibrating plate. Adjusting the distance will simultaneously change the acoustic intensity of both areas (but the magnitude of the change differs because the distance change from the excessive and weak areas to the vibrating plate is the same, but the rate of change of acoustic intensity differs due to the difference in initial position). Through iterative calculation: during the cleaning process, the basket moves up and down at a certain speed curve, changing the distance between the silicon wafer and the vibrating plate in real time. Integrating the removal rate at each moment, a set of distance-time trajectories is found that results in the final removal amount of both the excessive and weak areas being equal to their initial thickness. This total processing time is then used as the total processing time. Alternatively, after dynamically changing the vertical distance between the silicon wafer and the vibrating plate, the weak area can be completely outside the acoustic wave coverage range; the minimum total processing time required to traverse the thickness area before and after the change is then used as the total processing time. In this embodiment, different schemes have different total processing times.

[0098] S55: By comparing the total processing time with the maximum cleaning time, the target adjustment distance and its corresponding target adjustment time point are obtained, and the actual installation position is updated in real time based on the target adjustment distance and its corresponding target adjustment time point.

[0099] The target adjustment distance refers to the distance the silicon wafer moves above the bottom of the tank when the total processing time is the same as the maximum cleaning time. The target adjustment time point refers to the time point at which the silicon wafer moves above the bottom of the tank when the total processing time is the same as the maximum cleaning time. By analyzing the consistency between the total processing time and the maximum cleaning time, the distance corresponding to the total processing time that is the same as the maximum cleaning time is taken as the target adjustment distance. The time point obtained by adding the total processing time to the set initial time point is taken as the target adjustment time point. The actual installation position is changed in real time using the target adjustment distance and its corresponding target adjustment time point.

[0100] In this embodiment, the flower basket can be configured as a layered, independently lifting flower basket or an adjustable groove depth.

[0101] Layered independent lifting flower basket: The flower basket is divided into multiple independent lifting layers along the vertical direction. Each layer can hold multiple silicon wafers. Each layer is driven by a motor to rise or fall independently, thereby changing the vertical distance between the silicon wafers in that layer and the megaacoustic wave plate.

[0102] Adjustable slot depth: Each slot in the basket has an adjustable depth support block at the bottom. The support block is raised and lowered by a servo motor, allowing for fine-tuning of the height of a single silicon wafer. Since the slot spacing is fixed (e.g., 10mm), the adjustment is limited to the spacing between adjacent slots. Therefore, the actual installation position can only be offset slightly from the preset slot coordinates.

[0103] The bottom of the tank is equipped with a clamping base and a telescopic rod to drive the silicon wafer to move vertically above the bottom of the tank.

[0104] Also includes: S60: Compare the abnormal information with the baseline cleanup information to obtain the abnormal information that exceeds the baseline cleanup information as the target abnormal information.

[0105] Target anomaly information refers to anomaly information that exceeds the baseline cleanup information. By analyzing the anomaly information and the baseline cleanup information, the anomaly information that exceeds the baseline cleanup information is regarded as target anomaly information.

[0106] S61: Calculate the difference between the target anomaly information and the baseline cleanup information as the target remaining amount of the target anomaly information.

[0107] The target remaining quantity refers to the deviation between the target anomaly information and the baseline cleanup information. The target remaining quantity is calculated by the difference between the target anomaly information and the baseline cleanup information.

[0108] S62: Supplementary acoustic parameters are obtained based on the target remaining amount and the maximum cleaning time.

[0109] The supplementary acoustic parameters refer to the additional mega-sound power density or equivalent sound pressure required to remove the target residue within the same maximum cleaning time, based on the baseline acoustic parameters. Referring to S30, the supplementary acoustic parameters are calculated by inputting the target residue and the maximum cleaning time into the function model.

[0110] S63: Obtain the sweep frequency parameters by supplementing the acoustic wave parameters and the reference acoustic wave parameters, and add the sweep frequency parameters to the cleaning control information.

[0111] Sweep parameters refer to the frequency modulation parameters that a megohmmeter generator needs to adjust in order to generate supplementary acoustic wave parameters. These parameters include: sweep range (e.g., center frequency), sweep period, sweep waveform (triangular wave or sine wave), and partitioned timing drive mode.

[0112] The sweep frequency parameters are matched by superimposing the reference acoustic wave parameters and the supplementary acoustic wave parameters and inputting them into a preset sweep frequency mapping table.

[0113] The frequency sweep mapping table stores the sweep parameters corresponding to the superposition of different sound waves. The function of the frequency sweep parameters is to change the spatial distribution of the sound field, so that the energy is more concentrated in the area of ​​the silicon wafer that needs to be compensated.

[0114] Methods after obtaining cleanup control information include: S70: Based on the reference acoustic wave parameters and the frequency sweep parameters, the reference energy distribution of each silicon wafer placed in the basket is obtained.

[0115] The reference energy distribution refers to the theoretical value of the acoustic power density distributed at the bottom coordinates of each slot where the silicon wafer is placed in the basket. By inputting the reference acoustic parameters and the sweep frequency parameters into the sound field simulation software, the sound intensity at each slot coordinate point of the basket is simulated and calculated to generate the reference energy distribution (e.g., stored in matrix form).

[0116] S71: Collect the actual energy distribution at various locations within the flower basket.

[0117] The actual energy distribution refers to the actual sound power density distribution obtained during the cleaning process by using an online hydrophone array or indirectly by back-calculating the cleaning state (reflectivity change) of the silicon wafers in each tank. This is achieved by installing hydrophones in fixed (or movable) positions within the tank, measuring the sound pressure in multiple tanks in real time, and interpolating to generate the actual energy distribution of the entire tank. Alternatively, the in-situ reflectance method (in-situ reflectance spectrometer) can be used to collect the real-time reflectivity change rate of each silicon wafer, and the sound power density it experiences can be back-calculated using a cleaning kinetic model. The cleaning kinetic model can be obtained by fitting several standard cleaning experiments (with known sound power).

[0118] When using an online hydrophone array to acquire actual energy distribution, 5 to 10 miniature hydrophones (such as Precision Acoustics' needle-type hydrophones, 0.5 mm in diameter) can be arranged vertically along the sidewall of the cleaning tank, spaced 20 mm apart, and a 3×3 grid array can be arranged horizontally at the bottom of the tank. The hydrophone signals are synchronously sampled by a multi-channel acquisition card, and the amplitude of the mega-sound frequency components is extracted using Fourier transform. Then, the absolute sound pressure value at each measurement point is obtained based on pre-calibrated voltage-sound pressure conversion coefficients. An interpolation algorithm (such as radial basis functions) is used to generate a sound pressure distribution cloud map of the entire tank.

[0119] S72: Calculate the energy distribution difference by comparing the actual energy distribution with the reference energy distribution.

[0120] The energy distribution difference refers to the data difference between the actual energy distribution and the reference energy distribution. By analyzing the consistency between the actual energy distribution and the reference energy distribution, when the actual energy distribution is consistent with the reference energy distribution, it indicates that the sound wave energy can be transmitted normally without being affected by external factors, and no adjustment is made.

[0121] When the actual energy distribution is inconsistent with the reference energy distribution, it indicates that the sound wave energy is affected by external factors and cannot be transmitted normally. The energy distribution difference is obtained by calculating the difference between the corresponding data between the actual energy distribution and the reference energy distribution.

[0122] S73: Obtain the energy distribution gradient through the energy distribution difference.

[0123] The energy distribution gradient refers to the rate of change of the energy distribution difference in space. By performing numerical difference on the energy distribution difference matrix, the rate of change along the vertical axis and the horizontal axis is extracted as the energy distribution gradient.

[0124] S74: Obtain basket control information based on energy distribution gradient and update cleaning control information.

[0125] The flower basket control information refers to the control commands used to control the flower basket to perform vertical up-and-down reciprocating motion or directly rotate horizontally. This control information is obtained by analyzing the energy distribution gradient and then used to update the cleaning control information. The specific methods for obtaining and updating the flower basket control information are described below.

[0126] Methods for obtaining and updating the flower basket control information include: S80: When the energy distribution gradient is the preset vertical gradient, the vertical stroke is obtained according to the energy distribution gradient and the specifications of the basket.

[0127] The vertical gradient is the rate of change of the energy distribution difference along the vertical direction of the flower basket, set by the technicians, with a preset threshold of 0.1 W / cm². 2 / cm. When the absolute value of the vertical gradient exceeds this threshold, vertical reciprocating motion control is initiated.

[0128] Vertical travel refers to the total distance the flower basket moves back and forth in the vertical direction. It is determined by taking the maximum vertical distance with an energy distribution gradient from the flower basket specifications.

[0129] In this embodiment, if the vertical travel is the same as the total height of the flower basket, the height value after subtracting the preset safety distance between the top and bottom ends from the total height of the flower basket is taken as the vertical travel.

[0130] S81: Extract the reference spacing from the basket specifications, and extract the variation gradient from the energy distribution gradient by means of the vertical direction and the reference spacing.

[0131] The reference spacing refers to the spacing between the slots used to place silicon wafers on each layer, which is obtained by retrieving the reference spacing from the basket specification.

[0132] The gradient of change refers to the rate of change of energy distribution corresponding to each reference spacing in the vertical direction. The gradient of change is obtained by extracting the rate of change of energy distribution corresponding to the reference spacing in the vertical direction from the energy distribution gradient.

[0133] S82: Obtain the vertical reciprocating speed based on the changing gradient.

[0134] Vertical reciprocating speed refers to the rate at which the basket rises and falls vertically. The vertical reciprocating speed is matched from a preset basket control mapping table by changing the gradient.

[0135] The basket control mapping table stores the vertical reciprocating speeds corresponding to different gradient changes; the larger the gradient change, the slower the vertical reciprocating speed.

[0136] S83: Obtain the standard position range of a normal silicon wafer by the actual installation location and energy distribution gradient, and update the sweep frequency parameters by the vertical reciprocating speed and the standard position range.

[0137] The standard location range refers to the range within which a silicon wafer meets the cleaning standard. It is defined as the range formed by actual installation positions that are not within the range of the energy distribution gradient.

[0138] The energy concentration is matched from the frequency sweep map table based on the vertical reciprocating speed, and the energy concentration within the standard position range is reduced in the frequency sweep parameters to obtain new frequency sweep parameters. The greater the vertical reciprocating speed, the greater the energy concentration.

[0139] S84: Update the cleaning control information using the updated sweep frequency parameters, and add the vertical reciprocating speed and vertical stroke as basket control information to the cleaning control information.

[0140] The updated sweep frequency parameters are used to replace the original sweep frequency parameters in the cleaning control information to form new cleaning control information. The vertical reciprocating speed and vertical stroke are then used as basket control information and added to the new cleaning control information.

[0141] Methods for obtaining and updating the cleanup control information for flower baskets also include: S90: When the energy distribution gradient is the preset horizontal gradient, the largest energy distribution difference is retrieved from the energy distribution gradient as the horizontal energy difference.

[0142] The horizontal gradient is the rate of change of the energy distribution difference along the horizontal direction of the basket (e.g., the X-axis or Y-axis), set by the technician, with a preset threshold of 0.05 W / cm². 2 / cm.

[0143] Horizontal energy difference refers to the difference between the maximum and minimum values ​​of energy distribution in the horizontal direction within the same height layer (unit: W / cm²). 2 By scanning the energy distribution differences of all slots, the largest energy distribution difference is taken as the horizontal energy difference.

[0144] S91: The rotational speed is obtained based on the horizontal energy difference.

[0145] Rotational speed refers to the angular velocity at which the basket rotates about its vertical axis. The rotational speed is matched from the basket control mapping table by the horizontal energy difference.

[0146] The flower basket control mapping table stores the rotation speed corresponding to different horizontal energy differences. The larger the horizontal energy difference, the faster the rotation speed.

[0147] S92: Obtain the rotation excess region and its corresponding rotation excess energy by using the rotation speed and energy distribution gradient.

[0148] The rotation excess region refers to the region where energy replenishment is too high under certain rotational speed conditions. The rotation excess energy refers to the energy value of the rotation excess region exceeding the baseline.

[0149] By combining the energy distribution difference matrix with the rotation speed, the excess energy accumulated by the silicon wafer when passing through the region during rotation is calculated: for each angle interval, the product of the energy distribution difference and the dwell time is integrated. The dwell time is determined by the rotation speed (the slower the speed, the longer the dwell time). Angle intervals where the integration result exceeds a preset threshold are marked as rotation excess regions, and the integrated value corresponding to the rotation excess region is taken as the rotation excess energy.

[0150] S93: Update sweep frequency parameters based on rotation out-of-region and rotation out-of-energy.

[0151] For the rotational excess region, the sweep frequency parameters of the megasonic wave are dynamically adjusted to reduce the actual energy output in this region and avoid localized over-corrosion. When the rotational excess energy > 0.5 W / cm 2 At that time, by rotating beyond the magnitude of the energy, the range of the sweep frequency is matched and adjusted from the sweep frequency mapping table to increase the sweep frequency range in the original sweep frequency parameters (to average the sound field time, reduce the peak energy, for example, to expand the sweep frequency range to ±0.8MHz).

[0152] When the rotational excess energy exceeds the preset sweep frequency range threshold, the correction duration is matched from the sweep frequency mapping table based on the magnitude of the rotational excess energy to reduce the opening time of the corresponding vibrating plate in the rotational excess area, thereby adjusting the partition timing to compensate for the unevenness in the horizontal direction.

[0153] The frequency sweep mapping table stores the adjustment range or correction duration corresponding to different rotation excess energies. The greater the rotation excess energy, the greater the adjustment range or correction duration.

[0154] S94: Update the cleaning control information according to the updated sweep frequency parameters, and add the rotation speed as basket control information to the cleaning control information.

[0155] The updated sweep frequency parameters are used to replace the original sweep frequency parameters in the cleaning control information to form new cleaning control information. The rotation speed is then used as basket control information and added to the new cleaning control information.

[0156] Based on the same inventive concept, embodiments of the present invention provide a cleaning system based on the TOPCon battery manufacturing process, comprising: The acquisition module is used to acquire manufacturing process specifications; The memory is used to store a program for a cleaning method based on the TOPCon battery manufacturing process; The processor is used to load and execute programs stored in memory.

[0157] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0158] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A cleaning method based on TOPCon battery manufacturing process, characterized in that, include: Collect the manufacturing process specifications of TOPCon batteries; According to the manufacturing process specifications, silicon wafers are texturized, boron expanded, oxidized, and laser-cut to form various test silicon wafers; Each silicon wafer undergoes borosilicate glass removal, alkaline polishing, and tunneling, and the test information for each wafer is collected. Pre-set additives are added to the cleaning solution, and the benchmark information corresponding to the test information is obtained based on the manufacturing process specifications. Anomaly information is obtained by comparing the consistency between the detection information and the baseline information; Based on the abnormal information, a cleaning control signal is obtained, and the cleaning control signal is output to place the detected silicon wafer into a cleaning solution containing additives for cleaning. After cleaning, each silicon wafer undergoes atomic layer deposition, positive film deposition, back film deposition, screen printing, and testing and sorting to produce a battery.

2. The cleaning method based on the TOPCon battery manufacturing process according to claim 1, characterized in that, Methods for obtaining cleaning control signals include: The installation position of the megasonic vibration plate in the cleaning tank was measured. Retrieve reference acoustic parameters from manufacturing process specifications; The acoustic coverage range is obtained based on the reference acoustic parameters and the installation location. Compare the size of each abnormal information and sort them to obtain the arrangement parameters from largest to smallest; Based on the arrangement parameters, the detection arrangement parameters are obtained through preset flower basket specifications, sound wave coverage range, and abnormal information; The maximum abnormal information is retrieved from the permutation parameters and used as the marked abnormal information. Based on the anomaly information and reference acoustic parameters, the actual installation position of each detected silicon wafer in the basket is obtained, and the actual installation position is used as the cleaning control signal.

3. The cleaning method based on the TOPCon battery manufacturing process according to claim 2, characterized in that, Methods for obtaining the actual installation location include: The maximum cleaning time is obtained based on the marked anomaly information and the reference acoustic parameters; The acoustic parameters corresponding to each abnormality are obtained based on the detection arrangement parameters and the acoustic coverage range. Baseline cleaning information is obtained by using acoustic parameters and maximum cleaning time; Compare the abnormal information with the baseline cleanup information and identify any abnormal information that does not exceed the baseline cleanup information as adjustment abnormal information; The actual installation location is obtained based on the adjustment anomaly information and the baseline cleanup information.

4. The cleaning method based on the TOPCon battery manufacturing process according to claim 3, characterized in that, Methods for obtaining the actual installation location include: The difference between the adjustment anomaly information and the baseline cleanup information is calculated as the adjustment excess information; The acoustic parameters can be adjusted by inversely deriving the information beyond the maximum cleaning time. The vertical adjustment distance is obtained by adjusting the acoustic wave parameters; The actual installation location is determined based on the flower basket specifications, vertical adjustment distance, and any abnormal adjustment information.

5. A cleaning method based on TOPCon battery manufacturing process according to claim 4, characterized in that, Other methods for obtaining the actual installation location include: Retrieve the thickness distribution of a single silicon wafer from the anomaly information; The average thickness value is calculated based on the detected thickness distribution, and the detected thickness distribution is divided into areas with excessive thickness and areas with insufficient thickness based on the average thickness value. The placement orientation of the silicon wafer within the basket is determined based on the relative orientation between the thickness-exceeding and thickness-deficient areas. The processing time is determined based on the thickness-exceeding region, the thickness-weak region, and the reference acoustic parameters. The total processing time is obtained by adjusting the distance based on the processing time of the two regions; By comparing the total processing time with the maximum cleaning time, the target adjustment distance and its corresponding target adjustment time point are obtained, and the actual installation position is updated in real time based on the target adjustment distance and its corresponding target adjustment time point.

6. The cleaning method based on the TOPCon battery manufacturing process according to claim 5, characterized in that, Also includes: By comparing the abnormal information with the baseline cleanup information, the abnormal information that exceeds the baseline cleanup information is taken as the target abnormal information. The difference between the target anomaly information and the baseline cleanup information is calculated as the target remaining amount of the target anomaly information; Supplementary acoustic parameters are obtained based on the target remaining amount and the maximum cleaning time; The sweep frequency parameters are obtained by supplementing the acoustic wave parameters and the reference acoustic wave parameters, and then the sweep frequency parameters are added to the cleaning control information.

7. A cleaning method based on TOPCon battery manufacturing process according to claim 6, characterized in that, Methods after obtaining cleanup control information include: The reference energy distribution at each position of the detection silicon wafer in the basket is obtained based on the reference acoustic wave parameters and the frequency sweep parameters. Collect the actual energy distribution at various locations within the flower basket; The energy distribution difference is calculated by comparing the actual energy distribution with the reference energy distribution. The energy distribution gradient can be obtained by the energy distribution difference; The control information for the flower basket is obtained based on the energy distribution gradient, and the cleaning control information is updated accordingly.

8. A cleaning method based on TOPCon battery manufacturing process according to claim 7, characterized in that, Methods for obtaining and updating the flower basket control information include: When the energy distribution gradient is the preset vertical gradient, the vertical stroke is obtained based on the energy distribution gradient and the specifications of the flower basket. The baseline spacing is retrieved from the basket specifications, and the variation gradient is retrieved from the energy distribution gradient by comparing the vertical direction with the baseline spacing. The vertical reciprocating velocity is obtained based on the changing gradient; The standard position range of a normal silicon wafer is obtained by the actual installation location and energy distribution gradient, and the sweep frequency parameters are updated by the vertical reciprocating speed and the standard position range. The cleaning control information is updated using the updated sweep frequency parameters, and the vertical reciprocating speed and vertical stroke are added to the cleaning control information as basket control information.

9. A cleaning method based on TOPCon battery manufacturing process according to claim 7, characterized in that, Methods for obtaining and updating the cleanup control information for flower baskets also include: When the energy distribution gradient is a preset horizontal gradient, the largest energy distribution difference is retrieved from the energy distribution gradient as the horizontal energy difference. The rotational speed is obtained based on the horizontal energy difference. The rotational excess region and its corresponding rotational excess energy are obtained by using the rotational speed and energy distribution gradient. The sweep frequency parameters are updated based on the rotation out-of-region and rotation out-of-energy. The cleaning control information is updated based on the updated sweep frequency parameters, and the rotation speed is used as basket control information and added to the cleaning control information.

10. A cleaning system based on TOPCon battery manufacturing process, characterized in that, include: The acquisition module is used to acquire manufacturing process specifications; A memory for storing a program that implements a cleaning method based on the TOPCon battery manufacturing process as described in any one of claims 1 to 9; The processor is used to load and execute programs stored in memory.

Images