Rapid annealing simulation parameter tuning method and device, computer equipment and storage medium

Abstract

The invention relates to a rapid annealing simulation parameter tuning method and device. The method comprises the following steps: acquiring a silicon carbide parameter file; analyzing the parameter file, and reading and sorting the system parameter information in the parameter file; inputting the system parameter information into a graph attention network model GAT to form a preliminary temperature rise scheme; a final heating scheme is generated in combination with historical data and process requirements; and outputting a silicon carbide annealing result. The problems that an existing annealing process depends on empirical setting, an optimal solution is difficult to find under different materials and process conditions, modern production requirements cannot be met, and improvement of material performance and efficiency optimization are limited are solved, the precision and the efficiency of the annealing process can be improved, and according to actual process requirements and historical data, the optimal solution can be found out easily. And a customized optimization scheme is provided to meet the requirements of different users.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, device, computer equipment, storage medium and computer program product for tuning parameters of a rapid annealing simulation. Background Art

[0002] In modern materials science and semiconductor manufacturing, ion implantation has become one of the important technologies for changing material properties and improving microstructure. Ion implantation changes the physical and chemical properties of materials by implanting high-energy ions into the surface or interior of materials, thereby optimizing the electrical, thermal and mechanical properties of materials. However, lattice defects and internal stresses are inevitably introduced during ion implantation, which must be repaired through a subsequent rapid annealing process. Rapid annealing can not only repair the lattice structure, but also promote the uniform distribution of ions and improve material properties. Traditional annealing processes are prone to excessive energy consumption or long process times due to the lack of targeting and precise control. At the same time, the parameters of each stage of the annealing process, such as heating rate, holding time, cooling rate, etc., often rely on empirical settings, and it is difficult to find the optimal solution under different material and process conditions. With the increase in process complexity and material diversity, traditional annealing methods cannot fully adapt to modern production needs, which in turn limits the improvement of material performance and the optimization of process efficiency. Summary of the invention

[0003] Based on this, it is necessary to provide a rapid annealing simulation parameter tuning method, device, computer equipment, storage medium and computer program product to address the above technical problems.

[0004] In a first aspect, the present application provides a method for tuning parameters of a rapid annealing simulation, the method comprising:

[0005] Get silicon carbide parameter file;

[0006] Parsing the parameter file, reading and arranging the system parameter information in the parameter file;

[0007] Input the system parameter information into the graph attention network model GAT to form a preliminary heating plan;

[0008] Combine historical data and process requirements to generate the final heating plan;

[0009] Heating the silicon carbide material according to the final heating scheme;

[0010] Predict and optimize the holding time based on the silicon carbide material properties, process requirements and industrial data;

[0011] Predict and optimize cooling paths according to silicon carbide system status, user requirements, and real data, the cooling paths including cooling temperature and cooling speed;

[0012] Output silicon carbide annealing results.

[0013] In one embodiment, parsing the parameter file, reading and arranging the system parameter information in the parameter file, includes:

[0014] Use the parser function to read the parameter file, perform bytecode conversion, and save the system parameter information into a list;

[0015] Read the system parameter information in the list and perform initialization assignment.

[0016] In one embodiment, the inputting the system parameter information into a graph attention network model to form a preliminary heating plan includes:

[0017] Performing standardization processing on the system parameter information;

[0018] Using silicon carbide particles as nodes of a graph in the graph attention network model and forming an adjacency matrix;

[0019] The standardized system parameter information is input into the graph attention network model as the silicon carbide material state data at multiple moments during the annealing process.

[0020] In one embodiment, combining historical data and process requirements to generate a final heating plan includes:

[0021] Input node features and adjacency matrix into the GAT model;

[0022] Calculate the loss based on the model prediction results and the actual annealing results;

[0023] Update the model parameters by minimizing the loss function and iterate multiple training cycles until the loss function converges or reaches a predetermined accuracy;

[0024] The annealing data is input into the GAT model, and a final heating scheme is generated; the final heating scheme includes an optimal heating temperature range, a heating rate, and a heating curve.

[0025] In one embodiment, the outputting of the silicon carbide annealing result includes:

[0026] Obtain key data during annealing;

[0027] The annealing results are evaluated based on the key data.

[0028] In one embodiment, the key data include the repair efficiency of silicon carbide lattice defects, ion energy distribution, residual stress and strain distribution.

[0029] In a second aspect, the present application also provides a rapid annealing simulation parameter tuning device, the device comprising:

[0030] A file acquisition module, used to acquire silicon carbide parameter files;

[0031] A parsing module, used to parse the parameter file, read and organize the system parameter information in the parameter file;

[0032] A first generation module is used to input the system parameter information into a graph attention network model GAT to form a preliminary heating plan;

[0033] The second generation module is used to generate the final heating plan by combining historical data and process requirements;

[0034] A heating module, used to heat the silicon carbide material according to the final heating scheme;

[0035] A heat preservation module, used for predicting and optimizing the heat preservation time according to the silicon carbide material characteristics, process requirements and industrial data;

[0036] A cooling module, used to predict and optimize a cooling path according to the silicon carbide system status, user requirements and real data, wherein the cooling path includes a cooling temperature and a cooling speed;

[0037] The output module is used to output the silicon carbide annealing results.

[0038] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0039] Get silicon carbide parameter file;

[0040] Parsing the parameter file, reading and arranging the system parameter information in the parameter file;

[0041] Input the system parameter information into the graph attention network model GAT to form a preliminary heating plan;

[0042] Combine historical data and process requirements to generate the final heating plan;

[0043] Heating the silicon carbide material according to the final heating scheme;

[0044] Predict and optimize the holding time based on the silicon carbide material properties, process requirements and industrial data;

[0045] Predict and optimize cooling paths according to silicon carbide system status, user requirements and real data, the cooling paths including cooling temperature and cooling speed;

[0046] Output silicon carbide annealing results.

[0047] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0048] Get silicon carbide parameter file;

[0049] Parsing the parameter file, reading and arranging the system parameter information in the parameter file;

[0050] Input the system parameter information into the graph attention network model GAT to form a preliminary heating plan;

[0051] Combine historical data and process requirements to generate the final heating plan;

[0052] Heating the silicon carbide material according to the final heating scheme;

[0053] Predict and optimize the holding time based on the silicon carbide material properties, process requirements and industrial data;

[0054] Predict and optimize cooling paths according to silicon carbide system status, user requirements, and real data, the cooling paths including cooling temperature and cooling speed;

[0055] Output silicon carbide annealing results.

[0056] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0057] Get silicon carbide parameter file;

[0058] Parsing the parameter file, reading and arranging the system parameter information in the parameter file;

[0059] Input the system parameter information into the graph attention network model GAT to form a preliminary heating plan;

[0060] Combine historical data and process requirements to generate the final heating plan;

[0061] Heating the silicon carbide material according to the final heating scheme;

[0062] Predict and optimize the holding time based on the silicon carbide material properties, process requirements and industrial data;

[0063] Predict and optimize cooling paths according to silicon carbide system status, user requirements, and real data, the cooling paths including cooling temperature and cooling speed;

[0064] Output silicon carbide annealing results.

[0065] The above-mentioned rapid annealing simulation parameter tuning method, device, computer equipment, storage medium and computer program product obtain silicon carbide parameter files; parse the parameter files, read and organize the system parameter information in the parameter files; input the system parameter information into the graph attention network model GAT to form a preliminary heating plan; combine historical data and process requirements to generate a final heating plan; heat the silicon carbide material according to the final heating plan; predict and optimize the insulation time according to the silicon carbide material characteristics, process requirements and industrial data; predict and tune the cooling path according to the silicon carbide system status, user needs and real data, the cooling path includes cooling temperature and cooling speed; output silicon carbide annealing results. It solves the problem that the existing annealing process relies on empirical settings, it is difficult to find the optimal solution under different material and process conditions, it cannot adapt to modern production needs, and it limits the improvement of material performance and efficiency optimization. It can not only improve the accuracy and efficiency of the annealing process, but also provide customized optimization solutions according to actual process requirements and historical data to meet the needs of different users. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A schematic diagram of a process flow of a method for tuning parameters of a rapid annealing simulation in one embodiment;

[0067] Figure 2 It is a structural block diagram of a rapid annealing simulation parameter tuning device in one embodiment;

[0068] Figure 3 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0069] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0070] The present application provides a parameter tuning method and system for rapid annealing simulation after silicon carbide ion implantation, the purpose of which is to perform parameter tuning on the rapid annealing process after silicon carbide ion implantation through a graph attention network (GAT) to achieve optimal control of the heating, holding and cooling stages. By obtaining the relevant parameters of the system and inputting them into the trained network model, the heating temperature, holding time and cooling path are predicted and optimized to ensure that the silicon carbide material maintains the best physical state during the annealing process, so as to achieve the final effect of reasonable ion distribution, lattice repair and stable material performance. In addition, the method can be further optimized based on the trained model in combination with actual industrial data to ensure the scientific nature of the simulation results and close to the actual process requirements; and customized tuning is performed according to user needs to achieve energy minimization or the shortest annealing scheme.

[0071] In one embodiment of the present application, Figure 1 As shown, a method for tuning parameters of a rapid annealing simulation is provided, the method comprising:

[0072] S100, obtaining a silicon carbide parameter file.

[0073] Get the user-defined SiC input parameter file imported by the user, and read the relevant parameters of the system to be simulated, including the distribution of Si and C elements and other system-related information.

[0074] S200, parsing the parameter file, reading and organizing the system parameter information in the parameter file.

[0075] In one embodiment of the present application, S200 further includes the following steps:

[0076] (1) Use the parser function to read the parameter file, perform bytecode conversion, and save the system parameter information into a list.

[0077] Use the parser syntax analyzer function in the initialization module to read the input file of the system parameter information set by the user, perform bytecode conversion, and save the processed system parameter information into a list.

[0078] (2) Read the system parameter information in the list and perform initialization assignment.

[0079] The system parameter information in the generated list is read using the read_parameters() function in the input module, and initialized and assigned using the init_parameters() function. The system parameter information not set in the input will be assigned the default value.

[0080] The main system parameter information includes the concentration and position of different elements in the silicon carbide system, related information of different isomers in the material, the initial temperature of the material, the system's collective size and boundary conditions, as well as some other simulation-related parameters.

[0081] S300, inputting the system parameter information into the graph attention network model GAT to form a preliminary heating plan.

[0082] Specifically, the system parameter information read in S200 is input into a trained graph attention network (GAT) model. The GAT model performs a preliminary analysis based on the system parameter information, generates a preliminary heating plan, and predicts the optimal heating temperature range.

[0083] In one embodiment, S300 includes the following steps:

[0084] (1) Standardizing the system parameter information.

[0085] Before inputting the system parameter information into the GAT model, the data needs to be standardized through the standard() function to ensure that parameters of different dimensions can be correctly processed by the model and improve the accuracy of training and prediction.

[0086] (2) The silicon carbide particles are used as nodes of the graph in the graph attention network model and an adjacency matrix is ​​formed.

[0087] The silicon carbide ions are used as nodes in the graph of the GAT model, and edges are established based on the inter-ion forces to form an adjacency matrix. The node characteristics include the physical properties of each silicon carbide ion, such as temperature, and the edge weights reflect the strength of the interaction between nodes, such as distance and energy exchange.

[0088] (3) The standardized system parameter information is input into the graph attention network model as the silicon carbide material state data at multiple moments during the annealing process.

[0089] The standardized system parameter information is then input into the model through the gat_layer() function as the silicon carbide material state data at multiple moments during the annealing process.

[0090] S400, combines historical data and process requirements to generate the final heating plan.

[0091] In this application, the GAT model will combine historical data and actual process requirements to further optimize the physical changes of the material that may occur during the heating process to ensure the thermodynamic stability and process accuracy of the annealing process. Specifically, S400 also includes the following steps:

[0092] (1) Input node features and adjacency matrix into the GAT model.

[0093] Specifically, the node features and adjacency matrix are input into the GAT model through the heating_in() function, and the attention mechanism is used to calculate the weighted sum between each node and its neighboring nodes, and update the feature representation of each node.

[0094] (2) Calculate the loss based on the model prediction results and the actual annealing results.

[0095] Specifically, the loss is calculated based on the model prediction results and the actual annealing results, and the Adam optimization algorithm is used for back propagation.

[0096] (3) Update the model parameters by minimizing the loss function and iterate multiple training cycles until the loss function converges or reaches the predetermined accuracy.

[0097] (4) Inputting the annealing data into the GAT model and generating a final heating plan; the final heating plan includes an optimal heating temperature range, a heating rate, and a heating curve.

[0098] Specifically, the annealing data collected by sensors in the actual process are input into the GAT model through the heating_fact() function for fine-tuning to make the model closer to the actual operating conditions.

[0099] Furthermore, based on the trained GAT model, a plan including the optimal heating temperature range, heating rate and heating curve is generated through the heating_plan() function for use in the next step.

[0100] S500, heating the silicon carbide material according to the final heating scheme.

[0101] S600: predict and optimize the holding time according to the silicon carbide material characteristics, process requirements and industrial data.

[0102] After the heating is completed, the temperature is maintained within the optimal range for a period of time. At this time, the GAT model will predict and adjust the holding time based on the material properties and process requirements to ensure that the ion distribution in the system reaches the optimal state, reduce lattice defects and repair structural damage inside the material.

[0103] Specifically, S600 includes the following steps:

[0104] (1) After the heating process is completed, the GAT model determines the optimal temperature range for the holding phase based on the generated heating curve.

[0105] (2) Based on the preliminary prediction results of the holding time, the GAT model calls the attention mechanism through the GATConv() function to identify the key nodes that have the greatest impact on the internal ion distribution and lattice repair of the silicon carbide material (such as the ion concentration or lattice state in a specific area). Combined with the process requirements, the GAT model focuses on tuning these key nodes and dynamically adjusts the holding time. For example, under certain conditions, the GAT model may identify uneven ion distribution or insufficient repair in certain areas, requiring the holding time to be extended or shortened to ensure that the silicon carbide ion distribution in the system is balanced and the lattice defects are repaired to the optimal state. The entire tuning process relies on the GAT model's learning of historical data and integration of process experience to ensure the accuracy and effectiveness of the silicon carbide annealing process;

[0106] (3) After optimization, the GAT model continuously applies the attention mechanism during the hold time stage to monitor the ion distribution and lattice repair effect inside the silicon carbide material in real time. Through the attention mechanism, the GAT model can focus on the areas with the greatest changes inside the material and dynamically track the ion distribution and lattice repair of these nodes. If the model recognizes that the ion distribution of certain nodes is uneven or the lattice repair effect does not meet expectations, GAT will automatically adjust the attention level to these nodes and further adjust the overall hold time to ensure that the ion distribution inside the material is uniform and the lattice defects are effectively repaired, thereby optimizing the material performance.

[0107] S700, predicting and optimizing a cooling path according to a silicon carbide system state, user requirements, and real data, wherein the cooling path includes a cooling temperature and a cooling speed.

[0108] After the temperature holding time ends, the cooling phase begins. The GAT model predicts and optimizes the cooling temperature range and cooling rate based on the current state of the annealing system to ensure that the material properties of silicon carbide remain stable during the cooling process and avoid new defects. At the same time, the user's needs (such as minimizing energy or minimizing time) are input to optimize the cooling strategy and find the best cooling path and time.

[0109] Specifically, S700 includes the following steps:

[0110] (1) After the holding time ends, the system enters the cooling phase. The GAT model first determines the starting temperature of the silicon carbide cooling process through the cooling_info() function, that is, the temperature of the material at the end of the holding phase. This temperature will be used as the starting condition for cooling to ensure that the temperature drops under appropriate conditions.

[0111] (2) The GAT model analyzes the performance of silicon carbide materials at different cooling rates through the cooling_performance() function based on the physical properties (such as thermal conductivity, thermal expansion coefficient, melting point, etc.) parameters and process requirements extracted in step (1). By analyzing the characteristics of silicon carbide, the model can evaluate whether new defects will be generated in the cooling process of silicon carbide, and on this basis, formulate a cooling strategy through the cooling_plan() function.

[0112] (3) After determining the cooling temperature range and cooling rate, the model will input the user's specific needs (such as pursuing energy minimization, shortest time, or optimization of material properties) into the model through the userneed_in() function, and then customize the cooling strategy: if the user prioritizes the shortest cooling time, the model will speed up the cooling speed; if minimizing energy consumption is the primary goal, the cooling rate will be reduced to reduce energy consumption.

[0113] (4) During steps (2) and (3), the model monitors the state of the material in real time, such as temperature changes, internal structural stress and strain of the material, etc. If unstable performance or inappropriate cooling speed occurs during the cooling process, the neural network will automatically adjust the cooling path and speed through the cooling_adjust() function to ensure the stability of the final material performance.

[0114] (5) During the cooling process, the model will monitor the state of the silicon carbide material in real time through the cooling_monitor() function, such as temperature changes, internal structural stress and strain of the material, etc. If the performance is unstable or the cooling speed is inappropriate during the cooling process, the model will automatically adjust the cooling path and speed to ensure the stability of the final material performance.

[0115] (6) After step (5) is completed, the neural network will output the final silicon carbide cooling plan through the cooling_plan() function. The plan includes the cooling starting temperature, cooling temperature range, cooling rate, and the test results of material properties during the cooling process to ensure that the material remains stable during the cooling stage and avoid new defects.

[0116] S800, outputting silicon carbide annealing results.

[0117] After cooling is completed, the system outputs the final results of silicon carbide annealing, including lattice defect repair efficiency, ion energy distribution, residual stress and strain distribution, etc. Through these data, the GAT model will further optimize and analyze the entire annealing process, evaluate whether the expected process goals are achieved, and make necessary optimizations and adjustments based on the results, providing feedback and guidance for subsequent process optimization.

[0118] Specifically, S800 includes the following steps:

[0119] (1) Obtain key data during the annealing process. Key data include the repair efficiency of SiC lattice defects, ion energy distribution, residual stress, and strain distribution.

[0120] After cooling is completed, the system will output key data of the entire annealing process through the sys_output() function. These data include the repair efficiency of silicon carbide lattice defects, ion energy distribution, residual stress and strain distribution, etc.

[0121] (2) Evaluating the annealing results based on the key data.

[0122] The GAT model uses the sys_analyze() function to perform a comprehensive analysis based on the data from the annealing process to evaluate whether the expected process goals have been achieved. The focus is on whether lattice defects are effectively repaired, whether ion energy is evenly distributed, and whether new residual stress and strain are generated inside the material. If it is found that some indicators fail to meet expectations, the system will identify possible reasons and record the problem points.

[0123] It can be seen that the present application provides a method for tuning parameters of rapid annealing simulation, which uses a graph attention network (GAT) to tune the parameters of the rapid annealing process after silicon carbide ion implantation, and achieve optimal control of the heating, holding and cooling stages. By obtaining the relevant parameters of the system and inputting them into the trained network model, the heating temperature, holding time and cooling path are predicted and optimized to ensure that the silicon carbide material maintains the best physical state during the annealing process, so as to achieve the final effect of reasonable ion distribution, lattice repair and stable material performance. In addition, the method can be further optimized based on the trained model in combination with actual industrial data to ensure the scientific nature of the simulation results and close to the actual process requirements; and customized tuning is performed according to user needs to achieve energy minimization or the shortest annealing scheme.

[0124] In addition, compared with the prior art, the present application has the following beneficial effects:

[0125] (1) Since this application precisely controls the heating stage during the rapid annealing simulation of silicon carbide, the optimal heating scheme is generated by inputting system parameters into GAT and combining the historical data of silicon carbide with the actual process requirements. The model accurately predicts and optimizes the heating process to ensure the thermodynamic stability of the material during the heating stage and avoid physical stress or structural damage to the material caused by improper temperature changes, thereby greatly improving the overall accuracy of the annealing process and making the process more controllable and stable.

[0126] (2) Since this application optimizes the holding time during the rapid annealing simulation, the model predicts and adjusts the holding time based on material properties and process requirements to ensure uniform ion distribution inside the silicon carbide material, reduce lattice defects and repair structural damage. This optimization greatly improves the lattice repair efficiency, allowing the material to be fully repaired at the holding temperature, ensuring significant performance improvements at the microstructural level;

[0127] (3) Since this application effectively controls the cooling stage during the simulation of rapid annealing of silicon carbide, the GAT model optimizes the cooling temperature range and cooling rate according to the current state of the annealing system to ensure that no new stress or defects are generated during the cooling process. By tuning the cooling strategy, the cooling path and time are optimally configured to ensure that the performance of the material is stable during the cooling stage, avoid unevenness or new defects in the material after annealing, and meet the energy or time requirements of the user.

[0128] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0129] Based on the same inventive concept, the embodiment of the present application also provides a device for implementing the above-mentioned rapid annealing simulation parameter tuning. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above-mentioned method, so the specific limitations in one or more embodiments of the rapid annealing simulation parameter tuning device provided below can be referred to the limitations of the rapid annealing simulation parameter tuning method above, and will not be repeated here.

[0130] In one embodiment, Figure 2 As shown, a rapid annealing simulation parameter tuning device is provided, comprising:

[0131] The file acquisition module 100 is used to acquire a silicon carbide parameter file.

[0132] The parsing module 200 is used to parse the parameter file, read and organize the system parameter information in the parameter file.

[0133] The first generation module 300 is used to input the system parameter information into the graph attention network model GAT to form a preliminary heating plan.

[0134] The second generation module 400 is used to generate a final temperature increase plan in combination with historical data and process requirements.

[0135] A heating module 500, used to heat the silicon carbide material according to the final heating scheme;

[0136] A heat preservation module 600, used to predict and optimize the heat preservation time according to the silicon carbide material characteristics, process requirements and industrial data;

[0137] A cooling module 700 is used to predict and optimize a cooling path according to the silicon carbide system status, user requirements and real data, wherein the cooling path includes a cooling temperature and a cooling speed;

[0138] The output module 800 is used to output the silicon carbide annealing result.

[0139] In one embodiment, the parsing module 200 is further configured to:

[0140] Use the parser function to read the parameter file, perform bytecode conversion, and save the system parameter information into a list;

[0141] Read the system parameter information in the list and perform initialization assignment.

[0142] In one embodiment, the first generating module 300 is further configured to:

[0143] Performing standardization processing on the system parameter information;

[0144] Using silicon carbide particles as nodes of a graph in the graph attention network model and forming an adjacency matrix;

[0145] The standardized system parameter information is input into the graph attention network model as the silicon carbide material state data at multiple moments during the annealing process.

[0146] In one embodiment, the second generating module 400 is further configured to:

[0147] Input node features and adjacency matrix into the GAT model;

[0148] Calculate the loss based on the model prediction results and the actual annealing results;

[0149] Update the model parameters by minimizing the loss function and iterate multiple training cycles until the loss function converges or reaches a predetermined accuracy;

[0150] The annealing data is input into the GAT model, and a final heating scheme is generated; the final heating scheme includes an optimal heating temperature range, a heating rate, and a heating curve.

[0151] In one embodiment, the output module 800 is further configured to:

[0152] Obtain key data during annealing;

[0153] The annealing results are evaluated based on the key data.

[0154] In one embodiment, the key data include the repair efficiency of silicon carbide lattice defects, ion energy distribution, residual stress and strain distribution.

[0155] Each module in the above-mentioned rapid annealing simulation parameter tuning device can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0156] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store preset data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for tuning parameters of a rapid annealing simulation is implemented.

[0157] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0158] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the above-mentioned rapid annealing simulation parameter tuning method when executing the computer program.

[0159] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned rapid annealing simulation parameter tuning method is implemented.

[0160] In one embodiment, a computer program product is provided, including a computer program, which implements the above-mentioned rapid annealing simulation parameter tuning method when executed by a processor.

[0161] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the above-mentioned computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited thereto. The processors involved in each embodiment provided in this application may be general-purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, data processing logic devices based on quantum computing, etc., but are not limited thereto.

[0162] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0163] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A rapid annealing simulation parameter tuning method, characterized in that: The method comprises: Get silicon carbide parameter file; Parsing the parameter file, reading and arranging the system parameter information in the parameter file; Input the system parameter information into the graph attention network model GAT to form a preliminary heating plan; Combine historical data and process requirements to generate the final heating plan; Heating the silicon carbide material according to the final heating scheme; Predict and optimize the holding time based on the silicon carbide material properties, process requirements and industrial data; Predict and optimize cooling paths according to silicon carbide system status, user requirements and real data, the cooling paths including cooling temperature and cooling speed; Output silicon carbide annealing results.

2. The method according to claim 1, characterized in that The parsing of the parameter file, reading and arranging the system parameter information in the parameter file, includes: Use the parser function to read the parameter file, perform bytecode conversion, and save the system parameter information into a list; Read the system parameter information in the list and perform initialization assignment.

3. The method according to claim 1, characterized in that The inputting the system parameter information into the graph attention network model to form a preliminary heating plan includes: Performing standardization processing on the system parameter information; Using silicon carbide particles as nodes of a graph in the graph attention network model and forming an adjacency matrix; The standardized system parameter information is input into the graph attention network model as the silicon carbide material state data at multiple moments during the annealing process.

4. The method according to claim 3, characterized in that The combination of historical data and process requirements to generate the final heating plan includes: Input node features and adjacency matrix into the GAT model; Calculate the loss based on the model prediction results and the actual annealing results; Update the model parameters by minimizing the loss function and iterate multiple training cycles until the loss function converges or reaches a predetermined accuracy; The annealing data is input into the GAT model, and a final heating scheme is generated; the final heating scheme includes an optimal heating temperature range, a heating rate, and a heating curve.

5. The method according to claim 1, characterized in that The outputting of the silicon carbide annealing result comprises: Obtain key data during annealing; The annealing results are evaluated based on the key data.

6. The method according to claim 5, characterized in that The key data include the repair efficiency of silicon carbide lattice defects, ion energy distribution, residual stress and strain distribution.

7. A rapid annealing simulation parameter tuning device, characterized in that: The device comprises: A file acquisition module, used to acquire silicon carbide parameter files; A parsing module, used to parse the parameter file, read and organize the system parameter information in the parameter file; A first generation module is used to input the system parameter information into a graph attention network model GAT to form a preliminary heating plan; The second generation module is used to generate the final heating plan by combining historical data and process requirements; A heating module, used to heat the silicon carbide material according to the final heating scheme; A heat preservation module, used for predicting and optimizing the heat preservation time according to the silicon carbide material characteristics, process requirements and industrial data; A cooling module, used to predict and optimize a cooling path according to the silicon carbide system status, user requirements and real data, wherein the cooling path includes a cooling temperature and a cooling speed; The output module is used to output the silicon carbide annealing results.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images