A novel substrate cutting method

Abstract

This invention relates to the field of laser cutting control technology, and provides a novel substrate cutting method. This invention acquires substrate cutting task data and equipment data, generates a real-time image of the substrate to be cut using a digital twin model, and simulates cutting using a substrate cutting simulation model, comparing the simulation with the target cutting data to improve quality. If the cutting is satisfactory, the cost is predicted; if the cost does not meet the target, the simulation model is optimized. Cluster analysis is performed on the cutting steps and optimized, and the optimized model is used to simulate and select the optimal cutting steps. This invention effectively improves the efficiency and yield of semiconductor substrate cutting, reduces manufacturing costs, and solves the problem of low substrate yield leading to increased costs in existing technologies.

Description

Technical Field

[0001] This invention relates to the field of laser cutting control technology, and in particular to a novel substrate cutting method. Background Technology

[0002] Due to their small size and low single-wafer processing efficiency, semiconductor substrates are typically processed in large sheets, which are then cut into smaller pieces to create the final product. To illustrate this, a practical example is provided below. The most common semiconductor substrate product is rectangular. In practice, it is first cut horizontally, then vertically, resulting in small rectangular pieces. However, due to product design requirements, sometimes it is necessary to cut it into non-rectangular shapes, such as equilateral triangles. The conventional approach is to still first cut it into rectangles, and then polish it to form the desired triangular shape.

[0003] The existing semiconductor substrate cutting method involves first cutting the semiconductor substrate into quadrilaterals, including rectangles or squares, and then performing secondary processing on the semiconductor substrate. This results in low semiconductor substrate production efficiency and low yield, which significantly increases the overall manufacturing cost of semiconductor substrates and affects the mass production of semiconductor substrates. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a novel substrate cutting method that solves the problem of significantly increased overall manufacturing costs of semiconductor substrates due to low yield rates.

[0005] To solve the above-mentioned technical problems, the specific technical solution of the present invention is as follows:

[0006] This invention provides a novel substrate cutting method, comprising:

[0007] Step S101: Obtain substrate cutting task data and substrate cutting equipment data. The substrate cutting task data includes substrate cutting target data and substrate cutting parameter standards. The substrate cutting equipment data includes substrate cutting equipment operating status, cutting equipment accuracy, and cutting equipment environmental information.

[0008] Step S102: Based on the substrate cutting task data, collect the size data and position data of the substrate to be cut, and substitute the size data and position data of the substrate to be cut into the preset digital twin model to generate real-time image data of the substrate to be cut.

[0009] Step S103: Substitute the real-time image data of the substrate to be cut into the preset substrate cutting simulation model to obtain the substrate cutting simulation result. Compare the substrate cutting simulation result with the substrate cutting target data to obtain the substrate cutting comparison result.

[0010] Step S104: If the substrate cutting comparison result is qualified, the substrate cutting comparison result is substituted into the preset substrate cutting cost model to obtain the substrate cutting cost prediction result. If the substrate cutting cost prediction result does not meet the substrate cutting target data, the substrate cutting simulation model is optimized to obtain the optimized substrate cutting simulation model.

[0011] Step S105: Process the real-time substrate image data to be cut using the optimized substrate cutting simulation model to obtain a real-time substrate data set. Perform cluster analysis on the data in the real-time substrate data set to obtain the cluster analysis results. If the cluster analysis results of the real-time substrate data to be cut are lower than the substrate cutting target data, optimize the substrate cutting steps. Receive the substrate cutting step data set. Use the optimized substrate cutting simulation model to sequentially apply the cutting steps in the substrate cutting step data set to obtain multiple sets of cutting step simulation results. Substitute the multiple sets of cutting step simulation results into the preset substrate cutting cost model to obtain multiple sets of cutting step cost prediction results. Select the cutting steps that meet the substrate cutting target data from the multiple sets of cutting step cost prediction results as the real-time substrate cutting steps.

[0012] Furthermore, in the novel substrate cutting method provided by the present invention, step S102 includes:

[0013] Information extracted from substrate cutting task data includes the specific requirements, target size, shape, and precision of substrate cutting;

[0014] Based on the substrate material, size, and cutting requirements, determine the measuring tools to be used to measure the semiconductor substrate, determine the measurement method, and use the determined measuring tools to measure the length, width, and thickness of the substrate.

[0015] Accurately mark the cutting positions and lines on the substrate, and record the coordinates or relative positions of the marked points.

[0016] Furthermore, in the novel substrate cutting method provided by the present invention, step S103 includes:

[0017] The cutting simulation results obtained by processing the real-time image data of the substrate to be cut through the preset substrate cutting simulation model are loaded into the comparison system.

[0018] The cutting target data in the substrate cutting task data includes cutting size, shape, and precision indicators;

[0019] Based on the requirements of the cutting task and the results of the cutting simulation, the comparison indicators include cutting accuracy, dimensional deviation, and shape fit.

[0020] The cutting simulation results are compared with the cutting target data one by one using comparison software to calculate the difference or ratio. For graphic indicators, image recognition contour matching technology is used for comparison.

[0021] Based on the results of the comparison operation, analyze the differences between the cutting simulation results and the cutting target data, and generate a comparison result report.

[0022] Furthermore, in the novel substrate cutting method provided by the present invention, step S104 includes:

[0023] If the substrate cutting comparison results show that the substrate cutting is unqualified, the substrate cutting simulation model will be optimized to obtain the optimized substrate cutting simulation model.

[0024] Furthermore, in the novel substrate cutting method provided by the present invention, step S104 includes:

[0025] Based on the parameters and variables of the substrate cutting simulation model, a suitable encoding method is selected to represent each individual, and each individual has a set of cutting parameters;

[0026] Initialize a population containing multiple individuals, each representing a cutting scheme;

[0027] For each individual in the population, it is substituted into the substrate cutting simulation model to obtain the cutting simulation results;

[0028] The fitness value of each individual is calculated by comparing the cutting simulation results with the substrate cutting target data.

[0029] Based on fitness values, a tournament selection method is used to select individuals with higher fitness from the population as parents.

[0030] The selected parent individuals undergo crossover operations, exchanging gene fragments or recombining chromosomes to generate new offspring individuals. These offspring individuals are then subjected to mutation operations.

[0031] Repeat the fitness evaluation, selection, crossover, and mutation operations until the preset number of iterations is reached or other stopping conditions are met.

[0032] Furthermore, in the novel substrate cutting method provided by the present invention, step S105 includes:

[0033] Receive a substrate cutting step data set from the database. The substrate cutting step data set includes the cutting sequence, cutting depth, and cutting speed of multiple cutting steps.

[0034] In the optimized substrate cutting simulation model, the simulation environment is initialized based on the real-time image data of the substrate to be cut. Initializing the simulation environment includes setting the substrate's size parameters, shape parameters, material parameters, as well as the type and parameters of the cutting tool.

[0035] Furthermore, in the novel substrate cutting method provided by the present invention, step S105 includes:

[0036] Following the order of data combination for the substrate cutting steps, the simulation of each cutting step is executed sequentially. During the simulation, cutting effect, cutting force, and cutting temperature parameters are collected.

[0037] For each cutting step, the simulation results are recorded, including the shape, size, and cutting quality of the substrate after cutting.

[0038] After simulating all cutting steps, multiple sets of cutting step simulation results are generated. These results include a record of the simulation status of each cutting step and the state of the substrate after cutting.

[0039] The beneficial effects of this invention are mainly reflected in the following aspects:

[0040] This invention, by acquiring accurate substrate cutting task data and equipment data, enables the cutting process to be based on the latest and most accurate information, thereby improving cutting efficiency and yield. Utilizing digital twin models and substrate cutting simulation models, this invention can simulate and optimize the cutting process before actual cutting, reducing errors and waste during the cutting process and further improving cutting efficiency and yield.

[0041] This invention performs cost prediction after the cutting is qualified, and optimizes the substrate cutting simulation model based on the cost prediction results, thereby improving the ability to reduce manufacturing costs while meeting cutting quality requirements. By optimizing and selecting cutting steps, this invention can find the cutting steps that best match the target data, further reducing production costs.

[0042] This invention determines suitable measuring tools and methods based on the substrate material, size, and cutting requirements, improving the accuracy of substrate size measurement and accurately marking cutting positions and lines on the substrate, thereby enhancing cutting precision. Comparison software is used to compare the cutting simulation results with the target cutting data point by point, including indicators such as cutting precision, dimensional deviation, and shape fit, improving the cutting quality to meet requirements.

[0043] This invention, through a pre-set digital twin model and a substrate cutting simulation model, can quickly adapt to different substrate cutting tasks and equipment states, improving the system's flexibility and adaptability. During the cutting step optimization process, this invention can receive multiple cutting step data sets from a database and use the optimized model for simulation and filtering to find the most suitable cutting steps for the current task.

[0044] This invention reduces human intervention and errors by automating the data acquisition, simulation, comparison, and optimization processes, thereby improving the system's automation and intelligence levels. Furthermore, the use of intelligent optimization algorithms, such as genetic algorithms, to optimize the substrate cutting simulation model further enhances the system's intelligence and optimization efficiency.

[0045] In summary, this invention effectively solves the problem of low semiconductor substrate yield leading to a significant increase in manufacturing costs through intelligent steps, improves cutting efficiency and product quality, reduces overall manufacturing costs, and enhances the flexibility and adaptability of the system. Attached Figure Description

[0046] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a schematic diagram of a novel substrate cutting method provided in an embodiment of the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention. The technical solutions provided by various embodiments of this invention will be described in detail below with reference to the accompanying drawings.

[0049] To better understand the purpose of this invention, the invention will now be described in further detail.

[0050] This invention provides a novel substrate cutting method, comprising:

[0051] Step S101: Obtain substrate cutting task data and substrate cutting equipment data. The substrate cutting task data includes substrate cutting target data and substrate cutting parameter standards. The substrate cutting equipment data includes substrate cutting equipment operating status, cutting equipment accuracy, and cutting equipment environmental information.

[0052] In step S101, it is first necessary to obtain substrate cutting task data and substrate cutting equipment data.

[0053] The substrate cutting task data includes substrate cutting target data and substrate cutting parameter standards, which define the expected cutting results and required accuracy.

[0054] Data on substrate cutting equipment includes the operating status of the substrate cutting equipment, the cutting accuracy of the equipment, and information about the environment in which the equipment is located. This data helps to understand the current performance of the equipment and the environmental factors that affect the cutting results.

[0055] The target data for substrate cutting specifically includes the required dimensions, shape, and precision of the substrate after cutting.

[0056] Standards for substrate cutting parameters include recommended values ​​or allowable ranges for parameters such as cutting speed, cutting depth, and cutting angle.

[0057] Equipment operating status includes whether the equipment is running normally, whether there are any malfunctions, or whether maintenance is required. Cutting equipment precision reflects its cutting capability, i.e., the achievable cutting accuracy. Environmental information for the cutting equipment includes temperature, humidity, dust concentration, etc., as these factors affect cutting quality and equipment performance.

[0058] Obtaining accurate substrate cutting task data and substrate cutting equipment data is a prerequisite for improving cutting quality. Data is acquired through sensors, measuring tools, and manual input. After acquisition, the data needs to be preprocessed, such as cleaning and format conversion, to meet the requirements of subsequent steps.

[0059] As the cutting task progresses and the equipment operates, both task and equipment data will change. Therefore, these data need to be updated regularly to improve the accuracy and reliability of the cutting process. When acquiring, storing, and processing this data, it is essential to enhance data security and privacy protection, including preventing risks such as data leakage, corruption, or unauthorized access.

[0060] Step S102: Based on the substrate cutting task data, collect the size data and position data of the substrate to be cut, and substitute the size data and position data of the substrate to be cut into the preset digital twin model to generate real-time image data of the substrate to be cut.

[0061] Key information is extracted from the substrate cutting task data, including specific cutting requirements, target dimensions, shape, and precision. Based on the substrate material, dimensions, and cutting requirements, suitable measuring tools and methods are determined. This is to improve measurement accuracy, thereby ensuring cutting precision. Using the determined measuring tools, the length, width, and thickness of the substrate are measured. The accuracy of the measurement results directly affects the precision and quality of subsequent cutting.

[0062] Accurately marking the cutting positions and lines on the substrate is crucial for the cutting process. Simultaneously, recording the coordinates or relative positions of the marked points facilitates subsequent operations and record-keeping. Throughout step S102, it's essential to improve the accuracy of measurement and marking to avoid errors and waste during subsequent cutting. Furthermore, the measurement methods and marking techniques should be flexibly adjusted based on the specific substrate material and cutting requirements to achieve optimal cutting results.

[0063] After completing the above steps, the size and position data of the substrate to be cut are substituted into a preset digital twin model to generate real-time image data of the substrate to be cut. This step provides important reference and basis for subsequent cutting simulation and actual operation. By executing step S102 in detail, subsequent cutting simulation and actual operation can be made more accurate and efficient, thereby improving the precision and quality of substrate cutting.

[0064] Step S103: Substitute the real-time image data of the substrate to be cut into the preset substrate cutting simulation model to obtain the substrate cutting simulation result. Compare the substrate cutting simulation result with the substrate cutting target data to obtain the substrate cutting comparison result.

[0065] The cutting simulation results, obtained by processing real-time image data of the substrate to be cut through a preset substrate cutting simulation model, are loaded into the comparison system. Simultaneously, the cutting target data from the substrate cutting task data is loaded, including cutting dimensions, shape, and accuracy indicators. Based on the requirements of the cutting task and the cutting simulation results, the comparison indicators are determined, mainly including cutting accuracy, dimensional deviation, and shape fit.

[0066] The cutting simulation results are compared one by one with the target cutting data using comparison software, and the differences or ratios are calculated. For graphic indicators, image recognition contour matching technology is used for comparison to improve shape accuracy. Based on the comparison results, the differences between the cutting simulation results and the target cutting data are analyzed. A comparison result report is generated, recording in detail the differences between the cutting simulation results and the target cutting data.

[0067] If a significant discrepancy is found between the cutting simulation results and the target cutting data, the process can return to steps S102 or S104 for further optimization. Optimization includes adjusting measurement tools, measurement methods, cutting parameters, or the simulation model. Through these specific execution methods and related details, step S103 can improve the accuracy and reliability of the cutting simulation results, providing strong support for subsequent cost prediction and cutting step optimization.

[0068] Step S104: If the substrate cutting comparison result is qualified, the substrate cutting comparison result is substituted into the preset substrate cutting cost model to obtain the substrate cutting cost prediction result. If the substrate cutting cost prediction result does not meet the substrate cutting target data, the substrate cutting simulation model is optimized to obtain the optimized substrate cutting simulation model.

[0069] Step S104 is a key step in the technical solution of this invention, involving the prediction of substrate cutting costs and the optimization of the substrate cutting simulation model. The following is a detailed explanation of the specific execution method and related details of this step:

[0070] First, it's necessary to determine if the substrate cutting comparison result is acceptable. This step is fundamental for subsequent operations, ensuring that only acceptable cutting results are used for cost prediction. If the substrate cutting is acceptable: proceed to the next step, substituting the substrate cutting comparison result into the preset substrate cutting cost model to obtain the substrate cutting cost prediction result. If the substrate cutting is unacceptable: directly optimize the substrate cutting simulation model to obtain a more accurate cutting result.

[0071] If the substrate cutting is satisfactory, the substrate cutting comparison results are substituted into a preset substrate cutting cost model. This model, based on historical data and algorithms, can predict the cost of the current cutting scheme. Through cost prediction, companies can understand the economic benefits of the cutting process, providing a basis for subsequent cost control. The obtained substrate cutting cost prediction results are compared with the substrate cutting target data to determine whether they meet the cost requirements. If the cost prediction results do not match the target data, it indicates that the current cutting scheme is not the optimal solution and needs to be optimized.

[0072] If the predicted substrate cutting cost does not match the target data, the substrate cutting simulation model needs to be optimized. The optimization process involves adjusting model parameters and improving algorithms to enhance the accuracy and reliability of the simulation results. By optimizing the model, more accurate cutting results and cost predictions can be obtained, thereby improving production efficiency and economic benefits.

[0073] When substrate cutting fails to meet standards, a genetic algorithm can be used to optimize the substrate cutting simulation model. A genetic algorithm is an optimization algorithm based on the principles of biological evolution, solving complex problems by simulating natural selection and genetic mechanisms.

[0074] Encoding method: Based on the parameters and variables of the substrate cutting simulation model, a suitable encoding method is selected to represent each individual. Each individual represents a set of cutting parameters, i.e., a cutting scheme.

[0075] Population initialization: Initialize a population containing multiple individuals, each representing a cutting scheme. Through iterative optimization using a genetic algorithm, the optimal cutting scheme can be gradually approximated.

[0076] Iterative optimization: The population is iteratively optimized using a genetic algorithm (including fitness evaluation, selection, crossover, and mutation operations) until a preset number of iterations is reached or other stopping conditions are met. During the iteration process, cutting schemes are continuously screened and optimized to improve cutting efficiency and yield.

[0077] The optimized substrate cutting simulation model will be used for subsequent substrate cutting simulations and cost predictions. The optimized model can more accurately reflect the actual cutting situation, improving cutting efficiency and yield. At the same time, the optimization results can also provide strong support for enterprises' production planning and cost control.

[0078] The entire step S104 is an iterative optimization process. Through continuous comparison, prediction, and optimization, it gradually approaches the optimal cutting scheme and cost target. This method can significantly improve the production efficiency and yield of semiconductor substrates and reduce overall manufacturing costs. At the same time, iterative optimization also helps companies maintain a competitive edge in the fierce market competition.

[0079] By predicting and optimizing substrate cutting costs, companies can better control production costs and improve economic efficiency. Cost control is a crucial aspect of business management; effective cost control strategies and methods can reduce production costs, improve product quality, and enhance market competitiveness.

[0080] In summary, step S104 improves the accuracy and economy of the substrate cutting process by judging the substrate cutting comparison results, predicting substrate cutting costs, optimizing the substrate cutting simulation model, applying genetic algorithms, applying optimization results, iterative optimization, and controlling costs. This step is an indispensable part of the technical solution of this invention, providing strong support for improving the production efficiency and yield of semiconductor substrates and reducing overall manufacturing costs.

[0081] Step S105: The optimized substrate cutting simulation model is used to process the real-time substrate image data to be cut, resulting in a real-time substrate data set. Cluster analysis is performed on the data in the real-time substrate data set to obtain the cluster analysis results. If the cluster analysis results are lower than the substrate cutting target data, the substrate cutting steps are optimized. The substrate cutting step data set is received, and the optimized substrate cutting simulation model is used sequentially to apply the cutting steps in the substrate cutting step data set to obtain multiple sets of cutting step simulation results. The multiple sets of cutting step simulation results are substituted into the preset substrate cutting cost model to obtain multiple sets of cutting step cost prediction results. The cutting steps that meet the substrate cutting target data are selected from the multiple sets of cutting step cost prediction results as real-time substrate cutting steps. The real-time substrate cutting steps are sent to the substrate cutting equipment, and the substrate cutting equipment executes the real-time substrate cutting steps to cut the substrate.

[0082] Receive a substrate cutting step data set from the database. This data set contains the cutting sequence, cutting depth, and cutting speed of multiple cutting steps, providing detailed cutting parameters for subsequent simulation processes.

[0083] In the optimized substrate cutting simulation model, the simulation environment is initialized based on real-time image data of the substrate to be cut. This includes setting the substrate's size, shape, and material parameters, as well as the type and parameters of the cutting tool, improving the consistency between the simulation environment and actual cutting conditions.

[0084] The simulation of each cutting step is performed sequentially according to the order in the substrate cutting step data set. During the simulation, key indicators such as cutting effect, cutting force, and cutting temperature are collected to evaluate the feasibility and efficiency of the cutting steps.

[0085] For each cutting step, the simulation results are recorded, including the shape, size, and cutting quality of the substrate after cutting. These results will be used for subsequent analysis and optimization.

[0086] After simulating all cutting steps, multiple sets of simulation results were generated. These results recorded the simulation results for each cutting step, as well as the state of the substrate after cutting, providing data support for subsequent screening and optimization.

[0087] The simulation results of multiple cutting steps are substituted into a preset substrate cutting cost model to obtain multiple sets of cutting step cost prediction results. Then, the cutting steps that meet the substrate cutting target data are selected from these results and used as the real-time substrate cutting steps.

[0088] If the clustering analysis result of the real-time substrate data to be cut is lower than the target data for substrate cutting, the substrate cutting steps are optimized. This is achieved by adjusting cutting parameters, cutting sequence, or introducing new cutting technologies to improve cutting efficiency and substrate quality. Through these specific execution methods and related details, step S105 can improve the accuracy and efficiency of substrate cutting, while reducing production costs and increasing overall production efficiency.

[0089] Specifically, the novel substrate cutting method provided by the present invention includes step S102, which comprises:

[0090] Information extracted from substrate cutting task data includes the specific requirements, target size, shape, and precision of substrate cutting;

[0091] Based on the substrate material, size, and cutting requirements, determine the measuring tools to be used to measure the semiconductor substrate, determine the measurement method, and use the determined measuring tools to measure the length, width, and thickness of the substrate.

[0092] Accurately mark the cutting positions and lines on the substrate, and record the coordinates or relative positions of the marked points.

[0093] Specifically, the novel substrate cutting method provided by the present invention includes step S103, which comprises:

[0094] The cutting simulation results obtained by processing the real-time image data of the substrate to be cut through the preset substrate cutting simulation model are loaded into the comparison system.

[0095] The cutting target data in the substrate cutting task data includes cutting size, shape, and precision indicators;

[0096] Based on the requirements of the cutting task and the results of the cutting simulation, the comparison indicators include cutting accuracy, dimensional deviation, and shape fit.

[0097] The cutting simulation results are compared with the cutting target data one by one using comparison software to calculate the difference or ratio. For graphic indicators, image recognition contour matching technology is used for comparison.

[0098] Based on the results of the comparison operation, analyze the differences between the cutting simulation results and the cutting target data, and generate a comparison result report.

[0099] Specifically, the novel substrate cutting method provided by the present invention includes step S104, which comprises:

[0100] If the substrate cutting comparison results show that the substrate cutting is unqualified, the substrate cutting simulation model will be optimized to obtain the optimized substrate cutting simulation model.

[0101] Specifically, the novel substrate cutting method provided by the present invention includes step S104, which comprises:

[0102] Based on the parameters and variables of the substrate cutting simulation model, a suitable encoding method is selected to represent each individual, and each individual has a set of cutting parameters;

[0103] Initialize a population containing multiple individuals, each representing a cutting scheme;

[0104] For each individual in the population, it is substituted into the substrate cutting simulation model to obtain the cutting simulation results;

[0105] The fitness value of each individual is calculated by comparing the cutting simulation results with the substrate cutting target data.

[0106] Based on fitness values, a tournament selection method is used to select individuals with higher fitness from the population as parents.

[0107] The selected parent individuals undergo crossover operations, exchanging gene fragments or recombining chromosomes to generate new offspring individuals. These offspring individuals are then subjected to mutation operations.

[0108] Repeat the fitness evaluation, selection, crossover, and mutation operations until the preset number of iterations is reached or other stopping conditions are met.

[0109] Specifically, the novel substrate cutting method provided by the present invention includes step S105, which comprises:

[0110] Receive a substrate cutting step data set from the database. The substrate cutting step data set includes the cutting sequence, cutting depth, and cutting speed of multiple cutting steps.

[0111] In the optimized substrate cutting simulation model, the simulation environment is initialized based on the real-time image data of the substrate to be cut. Initializing the simulation environment includes setting the substrate's size parameters, shape parameters, material parameters, as well as the type and parameters of the cutting tool.

[0112] Specifically, the novel substrate cutting method provided by the present invention includes step S105, which comprises:

[0113] Following the order of data combination for the substrate cutting steps, the simulation of each cutting step is executed sequentially. During the simulation, cutting effect, cutting force, and cutting temperature parameters are collected.

[0114] For each cutting step, the simulation results are recorded, including the shape, size, and cutting quality of the substrate after cutting.

[0115] After simulating all cutting steps, multiple sets of cutting step simulation results are generated. These results include a record of the simulation status of each cutting step and the state of the substrate after cutting.

[0116] The technical solution of this invention solves the problem of low semiconductor substrate yield leading to a significant increase in manufacturing costs through the following methods:

[0117] Acquire substrate cutting task data and substrate cutting equipment data, including substrate cutting target data, substrate cutting parameter standards, equipment operating status, cutting equipment accuracy, and environmental information, to provide a foundation for subsequent precise cutting. Based on the substrate cutting task data, collect the size and position data of the substrate to be cut, input them into a preset digital twin model, and generate real-time image data of the substrate to be cut, improving the accuracy of data before cutting.

[0118] The real-time image data of the substrate to be cut is substituted into the preset substrate cutting simulation model to obtain the substrate cutting simulation results, which are then compared with the substrate cutting target data to improve the cutting quality to meet the requirements.

[0119] If the cutting is satisfactory, the cost is predicted; if the cost prediction does not meet the target data, the substrate cutting simulation model is optimized to reduce the manufacturing cost and improve the cutting quality.

[0120] The optimized model is used to process the data and perform cluster analysis. If the analysis results are lower than the target data, the cutting steps are optimized. The cutting step data group is received, the optimized model is used for simulation, and the cutting steps that meet the target data are selected as the real-time substrate cutting steps.

[0121] Based on the substrate material, size, and cutting requirements, determine the measuring tools and methods to accurately measure the substrate dimensions and mark the cutting positions and lines on the substrate to improve cutting precision. Use comparison software to compare the cutting simulation results with the target cutting data one by one, including cutting precision, dimensional deviation, and shape fit, to improve cutting quality.

[0122] Through this series of refined and intelligent steps, the technical solution of this invention effectively solves the problem of low semiconductor substrate yield leading to a significant increase in manufacturing costs, improves cutting efficiency and product quality, and reduces overall manufacturing costs.

[0123] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations. The above-described embodiments of this invention do not constitute a limitation on the scope of protection of this invention.

Claims

1. A novel substrate cutting method, characterized in that, include: Step S101: Obtain substrate cutting task data and substrate cutting equipment data. The substrate cutting task data includes substrate cutting target data and substrate cutting parameter standards. The substrate cutting equipment data includes the operating status of the substrate cutting equipment, the cutting equipment accuracy, and the environmental information of the cutting equipment. The environmental information of the cutting equipment includes temperature, humidity, and dust concentration. Step S102: Based on the substrate cutting task data, collect the size data and position data of the substrate to be cut, and substitute the size data and position data of the substrate to be cut into the preset digital twin model to generate real-time image data of the substrate to be cut. Step S103: Substitute the real-time image data of the substrate to be cut into the preset substrate cutting simulation model to obtain the substrate cutting simulation result. Compare the substrate cutting simulation result with the substrate cutting target data to obtain the substrate cutting comparison result. Step S104: If the substrate cutting comparison result is qualified, the substrate cutting comparison result is substituted into the preset substrate cutting cost model to obtain the substrate cutting cost prediction result. If the substrate cutting cost prediction result does not meet the substrate cutting target data, the substrate cutting simulation model is optimized using a genetic algorithm to obtain the optimized substrate cutting simulation model. Step S105: Process the real-time substrate image data to be cut using the optimized substrate cutting simulation model to obtain a real-time substrate data set. Perform cluster analysis on the data in the real-time substrate data set to obtain the cluster analysis result. If the cluster analysis result of the real-time substrate data is lower than the substrate cutting target data, optimize the substrate cutting steps. Receive the substrate cutting step data set. Use the optimized substrate cutting simulation model to sequentially apply the cutting steps in the substrate cutting step data set to obtain multiple sets of cutting step simulation results. Substitute the multiple sets of cutting step simulation results into the preset substrate cutting cost model to obtain multiple sets of cutting step cost prediction results. Select the cutting steps that meet the substrate cutting target data from the multiple sets of cutting step cost prediction results as real-time substrate cutting steps. Send the real-time substrate cutting steps to the substrate cutting equipment. Step S104 includes: Based on the parameters and variables of the substrate cutting simulation model, a suitable encoding method is selected to represent each individual, and each individual represents a set of cutting parameters; Initialize a population containing multiple individuals, each representing a cutting scheme; For each individual in the population, it is substituted into the substrate cutting simulation model to obtain the cutting simulation results; The fitness value of each individual is calculated by comparing the cutting simulation results with the substrate cutting target data. Based on fitness values, a tournament selection method is used to select individuals with higher fitness from the population as parents. The selected parent individuals undergo crossover operations, exchanging gene fragments or recombining chromosomes to generate new offspring individuals. These offspring individuals are then subjected to mutation operations. Repeat the fitness evaluation, selection, crossover, and mutation operations until the preset number of iterations is reached or other stopping conditions are met.

2. The novel substrate cutting method as described in claim 1, characterized in that, Step S102 includes: Information extracted from substrate cutting task data includes the specific requirements, target size, shape, and precision of substrate cutting; Based on the substrate material, size, and cutting requirements, determine the measuring tools to be used to measure the semiconductor substrate, determine the measurement method, and use the determined measuring tools to measure the length, width, and thickness of the substrate. Accurately mark the cutting positions and lines on the substrate, and record the coordinates or relative positions of the marked points.

3. The novel substrate cutting method as described in claim 1, characterized in that, Step S103 includes: The cutting simulation results obtained by processing the real-time image data of the substrate to be cut through the preset substrate cutting simulation model are loaded into the comparison system. The cutting target data in the substrate cutting task data includes cutting size, shape, and precision indicators; Based on the requirements of the cutting task and the results of the cutting simulation, the comparison indicators include cutting accuracy, dimensional deviation, and shape fit. The cutting simulation results are compared with the cutting target data one by one using comparison software to calculate the difference or ratio. For graphic indicators, image recognition contour matching technology is used for comparison. Based on the results of the comparison operation, analyze the differences between the cutting simulation results and the cutting target data, and generate a comparison result report.

4. The novel substrate cutting method as described in claim 1, characterized in that, Step S104 includes: If the substrate cutting comparison results show that the substrate cutting is unqualified, the substrate cutting simulation model will be optimized to obtain the optimized substrate cutting simulation model.

5. The novel substrate cutting method as described in claim 1, characterized in that, Step S105 includes: Receive a substrate cutting step data set from the database. The substrate cutting step data set includes the cutting sequence, cutting depth, and cutting speed of multiple cutting steps. In the optimized substrate cutting simulation model, the simulation environment is initialized based on the real-time image data of the substrate to be cut. Initializing the simulation environment includes setting the substrate's size parameters, shape parameters, material parameters, as well as the type and parameters of the cutting tool.

6. The novel substrate cutting method as described in claim 5, characterized in that, Step S105 includes: Following the order of data combination for the substrate cutting steps, the simulation of each cutting step is executed sequentially. During the simulation, cutting effect, cutting force, and cutting temperature parameters are collected. For each cutting step, the simulation results are recorded, including the shape, size, and cutting quality of the substrate after cutting. After simulating all cutting steps, multiple sets of cutting step simulation results are generated. These results include a record of the simulation status of each cutting step and the state of the substrate after cutting.

Images