Silicon carbide grinding process parameter optimization method and device

Abstract

The invention provides a silicon carbide grinding process parameter optimization method and device, and the method comprises the steps: carrying out the preprocessing of original parameters of a silicon carbide grinding process, obtaining grinding process parameters, carrying out the division of the preprocessed grinding process parameters according to a preset dimension classification rule, obtaining multi-dimensional data, responding to the target process parameters of a user, and carrying out the optimization of the multi-dimensional data. Whether historical process schemes matched with the target process parameters exist in a database or not is judged, if yes, the target process scheme is extracted from the matched historical process schemes according to a priority target set by a user, and the corresponding process parameters serve as an optimal parameter combination, and if not, parameter combination is conducted on the multi-dimensional data based on an NSGA-II genetic algorithm, and the target process scheme is extracted from the matched historical process schemes; and obtaining an optimal parameter combination. And according to the target process parameters, through a preset matching mechanism or an NSGA-II genetic algorithm, performing parameter combination on multi-dimensional data obtained through division according to a preset dimension classification rule, so as to optimize the silicon carbide grinding process parameters.

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor material processing, and in particular to a method and device for optimizing silicon carbide grinding process parameters. Background Art

[0002] With the rapid development of third-generation semiconductor materials (such as silicon carbide (SiC),) SiC is finding widespread application in new energy vehicles, high-voltage power supplies, optoelectronic devices, aerospace, and other fields. Its excellent electrothermal properties, wide bandgap, high breakdown field strength, and high thermal conductivity make it a core material for high-end manufacturing. However, as a highly hard and brittle material, SiC is extremely challenging to machine. Traditional mechanical grinding methods struggle to balance precision, surface quality, and processing efficiency.

[0003] In recent years, with the development of the Industrial Internet and intelligent manufacturing, data-driven intelligent process modeling and decision-making optimization have become key to upgrading grinding processes. In silicon carbide grinding, in particular, collecting and integrating multi-source heterogeneous data from equipment, sensors, and historical process records, and leveraging artificial intelligence algorithms to intelligently recommend and adaptively optimize grinding parameters, are key challenges that need to be addressed.

[0004] Although some studies or systems have attempted to apply big data technology to manufacturing process modeling and optimization scenarios, most of them are concentrated in the fields of general CNC machining, injection molding, or robot path planning. The dedicated grinding process data system for high-hardness and brittle materials (such as silicon carbide) is still in the exploratory stage. Existing silicon carbide grinding process data optimization solutions have certain limitations: (1) The data structure is not unified, and heterogeneous data is difficult to integrate; (2) There is a lack of an efficient data management architecture; (3) The reasoning method is single and there is a lack of an algorithm backup mechanism; (4) There is no closed-loop feedback and self-learning capabilities. Summary of the Invention

[0005] The present invention provides a method and device for optimizing silicon carbide grinding process parameters, which are used to solve technical problems such as complex data sources, scattered data structures, difficulty in optimizing process parameters, lack of systematic data management and intelligent calling mechanism, etc. in the existing silicon carbide grinding process, so as to optimize the silicon carbide grinding process parameters.

[0006] In a first aspect, the present invention provides a method for optimizing silicon carbide grinding process parameters, comprising: Preprocessing the obtained original silicon carbide grinding process parameters to obtain preprocessed grinding process parameters; Dividing the pre-processed grinding process parameters according to preset dimensional classification rules to obtain multi-dimensional data; In response to a parameter optimization request from a user, extracting target process parameters from the parameter optimization request; Determine whether there is a historical process plan in the database that matches the target process parameters; If so, extract the target process plan from the matching historical process plans according to the priority target set by the user, and determine the process parameters in the target process plan as the optimal parameter combination corresponding to the target process parameters; If not, based on the NSGA-II genetic algorithm, perform parameter combination on the multi-dimensional data to obtain the optimal parameter combination corresponding to the target process parameters.

[0007] Optionally, preprocess the original parameters of the silicon carbide grinding process to obtain the preprocessed grinding process parameters, including: Perform data cleaning on the original parameters of the silicon carbide grinding process obtained to remove noise, anomalies, invalid or duplicate data, and obtain the grinding process parameters after data cleaning; Perform unit standardization operation, timestamp standardization operation, encoding conversion operation and numerical reduction operation on the grinding process parameters after data cleaning to convert the grinding process parameters after data cleaning into data in the same format, and obtain the grinding process parameters after format conversion; Perform field name standardization operation, multi-source field mapping operation, data table association modeling operation and data label consistency operation on the grinding process parameters after format conversion to standardize the naming of the grinding process parameters after format conversion, and obtain the preprocessed grinding process parameters.

[0008] Optionally, determining whether there is a historical process plan in the database that matches the target process parameters includes: Convert the target process parameters into a standardized process vector in the same format as the multi-dimensional data; Based on a preset similarity matching mechanism, calculate the matching degrees between the standardized process vector and all historical process plans in the database to obtain multiple matching degree values; According to the magnitude relationship between the multiple matching degree values and a preset matching degree threshold, determine whether there is a historical process plan in the database that matches the target process parameters.

[0009] Optionally, based on the NSGA-II genetic algorithm, performing parameter combination on the multi-dimensional data to obtain the optimal parameter combination corresponding to the target process parameters includes: Randomly generate multiple initial parameter combinations in the multi-dimensional data according to the target process parameters; Input the multiple initial parameter combinations into a multi-objective optimization framework based on the NSGA-II algorithm and a neural network prediction model to iterate on the multiple initial parameter combinations, and calculate to obtain multiple predicted target parameter combinations; Optimize the process priority in the parameter optimization request, and extract the optimal parameter combination corresponding to the target process parameter from multiple combinations of the predicted target parameters.

[0010] Optionally, after performing parameter combination on the multi-dimensional data based on the NSGA-II genetic algorithm to obtain the optimal parameter combination corresponding to the target process parameter, it further includes: Generate a process plan for the target process parameter based on the optimal parameter combination, and record the process plan as a historical process plan in the database.

[0011] Optionally, recording the process plan as a historical process plan in the database includes: Calculate the quality score of the process plan based on a preset scoring rule; Record the process plan with a quality score greater than the preset scoring threshold as a historical process plan in the database.

[0012] In a second aspect, the present invention provides a device for optimizing process parameters of silicon carbide grinding, including: A preprocessing module for preprocessing the original parameters of the silicon carbide grinding process to obtain preprocessed grinding process parameters; A partitioning module for partitioning the preprocessed grinding process parameters according to a preset dimension classification rule to obtain multi-dimensional data; An extraction module for extracting the target process parameter from the parameter optimization request in response to a user's parameter optimization request; A judgment module for judging whether there is a historical process plan in the database that matches the target process parameter; if so, execute the determination module; if not, execute the optimization module; A determination module for extracting the target process plan from the matching historical process plans according to the priority target set by the user, and determining the process parameters in the target process plan as the optimal parameter combination corresponding to the target process parameter; An optimization module for performing parameter combination on the multi-dimensional data based on the NSGA-II genetic algorithm to obtain the optimal parameter combination corresponding to the target process parameter.

[0013] Optionally, the preprocessing module includes: A cleaning sub-module for performing data cleaning on the original parameters of the silicon carbide grinding process to remove noise, anomalies, invalid or duplicate data, and obtain data-cleaned grinding process parameters; The first conversion sub-module is used to perform unit standardization operation, timestamp standardization operation, encoding conversion operation, and numerical reduction operation on the ground-grinding process parameters after data cleaning, so as to convert the ground-grinding process parameters after data cleaning into data in the same format, and obtain the ground-grinding process parameters after format conversion; The naming sub-module is used to perform field name standardization operation, multi-source field mapping operation, data table association modeling operation, and data label consistency operation on the ground-grinding process parameters after format conversion, so as to standardize the naming of the ground-grinding process parameters after format conversion, and obtain the preprocessed ground-grinding process parameters.

[0014] Optionally, the judgment module includes: The second conversion sub-module is used to convert the target process parameters into a standardized process vector in the same format as the multi-dimensional data; The calculation sub-module is used to calculate the matching degrees between the standardized process vector and all historical process plans in the database based on a preset similarity matching mechanism, and obtain multiple matching degree values; The judgment sub-module is used to judge whether there is a historical process plan in the database that matches the target process parameters according to the magnitude relationship between multiple matching degree values and a preset matching degree threshold.

[0015] Optionally, the optimization module includes: The generation sub-module is used to randomly generate multiple initial parameter combinations in the multi-dimensional data according to the target process parameters; The iteration sub-module is used to input multiple initial parameter combinations into a multi-objective optimization framework based on the NSGA-II algorithm and a neural network prediction model to iterate multiple initial parameter combinations, and calculate and obtain multiple predicted target parameter combinations; The extraction sub-module is used to extract the optimal parameter combination corresponding to the target process parameters from multiple predicted target parameter combinations according to the process priority in the parameter optimization request.

[0016] Optionally, the optimization module further includes: The recording sub-module is used to generate a process plan for the target process parameters based on the optimal parameter combination, and record the process plan as a historical process plan into the database.

[0017] Optionally, the recording sub-module includes: The calculation unit is used to calculate the quality score of the process plan based on a preset scoring rule; The recording unit is used to record the process plan with a quality score greater than a preset scoring threshold as a historical process plan into the database.

[0018] In a third aspect, the present invention provides an electronic device, including a processor and a memory. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the steps in the method provided in the first aspect above are run.

[0019] In a fourth aspect, the present invention provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the method provided in the first aspect above are run.

[0020] In a fifth aspect, the present invention provides a computer program product, including a computer program. When the computer program is executed by a processor, the steps in the method provided in the first aspect above are run.

[0021] As can be seen from the above technical solutions, the present invention has the following advantages: The present invention provides a method and device for optimizing silicon carbide grinding process parameters. The method includes: preprocessing the obtained original silicon carbide grinding process parameters to obtain preprocessed grinding process parameters, dividing the preprocessed grinding process parameters according to a preset dimension classification rule to obtain multi-dimensional data, in response to a user's parameter optimization request, extracting target process parameters from the parameter optimization request, determining whether there is a historical process plan in the database that matches the target process parameters. If so, extracting a target process plan from the matching historical process plans according to the priority target set by the user, and determining the process parameters in the target process plan as the optimal parameter combination corresponding to the target process parameters. If not, based on the NSGA-II genetic algorithm, performing parameter combination on the multi-dimensional data to obtain the optimal parameter combination corresponding to the target process parameters. According to the target process parameters input by the user, through a preset matching mechanism or the NSGA-II genetic algorithm, performing parameter combination on the multi-dimensional data divided according to the preset dimension classification rule to obtain the optimal parameter combination, so as to realize the optimization of the silicon carbide grinding process parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. The drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0023] Figure 1 It is a flowchart of the first embodiment of the method for optimizing silicon carbide grinding process parameters of the present invention; Figure 2It is a flow chart of the second embodiment of the method for optimizing the process parameters of silicon carbide grinding according to the present invention; Figure 3 It is a structural block diagram of the embodiment of the device for optimizing the process parameters of silicon carbide grinding according to the present invention. Detailed implementation manners

[0024] The embodiments of the present invention provide a method and a device for optimizing the process parameters of silicon carbide grinding, which are used to optimize the process parameters of silicon carbide grinding.

[0025] In order to make the objectives, features, and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the embodiments described below are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0026] Embodiment 1, please refer to Figure 1 , Figure 1 It is a flow chart of the first embodiment of the method for optimizing the process parameters of silicon carbide grinding according to the present invention. The method includes: S1. Preprocess the obtained original process parameters of silicon carbide grinding to obtain the preprocessed grinding process parameters; In the embodiments of the present application, the obtained original process parameters of silicon carbide grinding include structured process data and unstructured process data collected from grinding equipment, sensors, process platforms, etc. The original process parameters of grinding are subjected to data cleaning, format conversion, and standardized naming to eliminate data noise and format differences. Among them, the data cleaning step ensures the integrity and reliability of the input data by filtering invalid or incorrect data, and avoids the failure of subsequent steps due to the interference of noise data. The format conversion step eliminates the heterogeneity problems caused by differences in data sources or storage methods by unifying the data format, and enhances the comparability and processability between different parameters. The standardized naming step provides a structured data basis for the application of subsequent dimension classification rules by standardizing the parameter naming rules, thereby improving the accuracy and scalability of multi-dimensional data partitioning.

[0027] S2. Divide the preprocessed grinding process parameters according to the preset dimension classification rules to obtain multi-dimensional data; In the embodiments of the present application, the preprocessed data is divided according to the preset dimension classification rules to realize the multi-dimensional structured recombination of parameters and establish an operable classification framework for parameter combination analysis.

[0028] S3. In response to the user's parameter optimization request, extract the target process parameters from the parameter optimization request.

[0029] In the embodiment of the present application, the target process parameters are extracted from the parameter optimization request provided by the user. The target process parameters include parameters such as the target surface roughness and processing efficiency. For example, the grinding surface roughness is less than 0.1 micrometer, and the processing time does not exceed 30 minutes.

[0030] S4. Determine whether there is a historical process plan in the database that matches the target process parameters. In the embodiment of the present application, the target process parameters are converted into a standardized process vector in the same format as the multi-dimensional data. Based on a preset similarity matching mechanism, the matching degree between the standardized process vector and all historical process plans in the database is calculated to obtain multiple matching degree values. According to the magnitude relationship between the multiple matching degree values and the preset matching degree threshold, it is determined whether there is a historical process plan in the database that matches the target process parameters.

[0031] S5. If so, extract the target process plan from the matching historical process plans according to the priority target set by the user, and determine the process parameters in the target process plan as the optimal parameter combination corresponding to the target process parameters. In the embodiment of the present application, if there is a historical process plan in the database that matches the target process parameters, the target process plan is extracted from the matching historical process plans according to the priority target set by the user.

[0032] S6. If not, based on the NSGA-II genetic algorithm, perform parameter combination on the multi-dimensional data to obtain the optimal parameter combination corresponding to the target process parameters.

[0033] In the embodiment of the present application, if there is no historical process plan in the database that matches the target process parameters, multiple initial parameter combinations are randomly generated in the multi-dimensional data according to the target process parameters. The multiple initial parameter combinations are input into a multi-objective optimization framework based on the NSGA-II algorithm and a neural network prediction model to iterate on the multiple initial parameter combinations, and multiple predicted target parameter combinations are calculated according to the process priority in the parameter optimization request.

[0034] The embodiment of the present invention provides a method for optimizing the process parameters of silicon carbide grinding. According to the target process parameters input by the user, through a preset matching mechanism or the NSGA-II genetic algorithm, parameter combination is performed on the multi-dimensional data divided according to the preset dimension classification rules to obtain the optimal parameter combination, so as to optimize the process parameters of silicon carbide grinding.

[0035] Embodiment 2. Please refer to Figure 2 ,Figure 2 This is a flowchart of the second embodiment of the method for optimizing the process parameters of a silicon carbide grinding process according to the present invention. The steps include: Step S201: Clean the original parameters of the silicon carbide grinding process to remove noise, anomalies, invalid or duplicate data, and obtain the grinding process parameters after data cleaning. In the embodiment of the present application, the original parameters of the silicon carbide grinding process are cleaned to remove noise, anomalies, invalid or duplicate data, and obtain the grinding process parameters after data cleaning. Among them, the original parameters of the silicon carbide grinding process include structured process data and unstructured process data collected from grinding equipment, sensors, process platforms, etc.

[0036] The cleaning content is shown in Table 1 below: Table 1 Cleaning Content Table

[0037] Specifically, this method can be applied to a data integration database system for a silicon carbide grinding process. The data integration database system includes: (1) a data collection and access module for collecting structured and unstructured process data from grinding equipment, sensors, process platforms, etc.; (2) a hierarchical data warehouse storage module for processing and storing the process data in sequence according to the operational data store (ODS), data warehouse detail (DWD), dimension layer (DIM), data warehouse summary (DWS), and data application layer (DAL); (3) an instance call and optimization module for completing rapid retrieval, intelligent reasoning, and multi-objective parameter optimization of process plans based on user-defined process parameter inputs; (4) a feedback learning module for scoring and evaluating the processing results, writing effective process plans back to the instance library, and updating rule knowledge.

[0038] Among them, the hierarchical data warehouse storage module specifically includes: (1) the ODS layer for receiving original collected data; (2) the DWD layer for performing cleaning, aggregation, and field unification; (3) the DIM layer for building dimension models such as grinding machine types, material types, processing tasks, etc. for classification; (4) the DWS layer for summarizing and statistically analyzing analysis indicators; (5) the DAL layer for providing data service interfaces for process instance calls and inference decisions.

[0039] The specific functions of each level are shown in Table 2 below: Table 2 Specific Functions of Each Level of the Hierarchical Data Warehouse Storage Module

[0040] Step S202: Perform unit standardization operation, timestamp standardization operation, encoding conversion operation, and numerical reduction operation on the ground grinding process parameters after data cleaning, so as to convert the ground grinding process parameters after data cleaning into data in the same format, and obtain the ground grinding process parameters after format conversion; In the embodiment of the present application, the ground grinding process parameters after data cleaning are converted into data in the same format, the heterogeneity problem caused by differences in data sources or storage methods is eliminated, the comparability and processability between different parameters are enhanced, and the ground grinding process parameters after format conversion are obtained.

[0041] The conversion content is shown in Table 3 below: Table 3 Conversion Content Table

[0042] Step S203: Perform field name standardization operation, multi-source field mapping operation, data table association modeling operation, and data label consistency operation on the ground grinding process parameters after format conversion, so as to standardize the naming of the ground grinding process parameters after format conversion, and obtain the ground grinding process parameters after preprocessing; In the embodiment of the present application, the ground grinding process parameters after format conversion are standardized in naming. By standardizing the parameter naming rules, a unified semantic identifier is established, and the ground grinding process parameters after preprocessing are obtained, providing a structured data basis for the application of subsequent dimension classification rules.

[0043] The specific standardized naming is shown in Table 4 below: Table 4 Specific Standardized Naming Table

[0044] Step S204: Divide the preprocessed ground grinding process parameters according to the preset dimension classification rules to obtain multi-dimensional data; In the embodiment of the present invention, according to the preset dimension classification rules, the preprocessed ground grinding process parameters are divided, and a large number of index data existing longitudinally are classified and structured by dimension to obtain multi-dimensional data, providing a statistical caliber and analysis perspective for subsequent summary analysis.

[0045] Specifically, according to the characteristics of the system focusing on the silicon carbide grinding process, the preprocessed ground grinding process parameters are divided into five categories according to the preset dimension classification rules. The multi-dimensional data specifically includes equipment operation type indicators (i.e., equipment dimension), process process type indicators (i.e., process parameter dimension), material removal / quality type indicators (i.e., material dimension), time series trend type indicators (i.e., time dimension), and multi-dimensional combination type indicators (i.e., cross-dimensional association).

[0046] Each multi-dimensional data is specifically shown in Tables 5 to 9 below: (1) Equipment operation - related metrics (equipment dimension): Table 5 Equipment Dimension Division Table

[0047] (2) Process - related metrics (process parameter dimension): Table 6 Process Parameter Dimension Division Table

[0048] (3) Material removal / quality - related metrics (material dimension): Table 7 Material Dimension Division Table

[0049] (4) Time - series trend - related metrics (time dimension): Table 8 Time Dimension Division Table

[0050] (5) Multi - dimensional combination - related metrics (cross - dimension association): Table 9 Cross - Dimension Association Division Table

[0051] Step S205: In response to the user's parameter optimization request, extract the target process parameters from the parameter optimization request. In the embodiment of the present application, the target process parameters are obtained from the parameter optimization request input by the user. The target process parameters include parameters such as the target surface roughness and machining efficiency. For example, the grinding surface roughness is less than 0.1 micron and the machining time does not exceed 30 minutes.

[0052] Step S206: Convert the target process parameters into a standardized process vector in the same format as the multi - dimensional data; based on a preset similarity matching mechanism, calculate the matching degrees between the standardized process vector and all historical process plans in the database to obtain multiple matching degree values; according to the magnitude relationship between the multiple matching degree values and a preset matching degree threshold, determine whether there is a historical process plan in the database that matches the target process parameters. In the embodiments of the present application, the dimensional difference of the parameter quantity is eliminated through normalization processing, and the target process parameter is converted into a standardized process vector. Based on a preset similarity matching mechanism (the cosine similarity algorithm can be selected for the similarity matching mechanism), the matching degree between the standardized process vector and all historical process plans in the database is calculated to obtain multiple matching degree values. According to the magnitude relationship between the multiple matching degree values and a preset matching degree threshold (the matching degree threshold can be set according to actual needs, generally set to 0.6), it is determined whether there is a historical process plan in the database that matches the target process parameter. When the matching degree value is greater than or equal to the matching degree threshold, it is determined that the corresponding historical process plan matches the target process parameter; when the matching degree value is less than the matching degree threshold, it is determined that the corresponding historical process plan does not match the target process parameter. The quantitative evaluation mechanism established thereby can eliminate the subjective interference of manual experience. For example, it can convert the "high-precision mode" that originally relied on text description into a process parameter combination with clear numerical boundaries. Through automatic threshold determination, the matching result can avoid misjudgment caused by inconsistent units. For example, after automatically converting the spindle speed unit marked as "rpm" in the historical plan to the "rad / s" unit of the target parameter, it participates in the calculation. This technical solution converts the semantic difference of process parameters into a computable spatial distance, solving the problem of insufficient accuracy caused by data heterogeneity in traditional matching methods.

[0053] Step S207, if so, extract the target process plan from the matching historical process plans according to the priority target set by the user, and determine the process parameters in the target process plan as the optimal parameter combination corresponding to the target process parameter; In the embodiments of the present application, if there is a matching historical process plan, the target process plan is extracted from the matching historical process plans according to the priority target set by the user. For example, if the user gives priority to surface quality, the historical plan with the lowest surface roughness is selected. The process parameters in the target process plan are used as the optimal parameter combination corresponding to the target process parameter.

[0054] Step S208, if not, randomly generate multiple initial parameter combinations in the multi-dimensional data according to the target process parameter; input the multiple initial parameter combinations into a multi-objective optimization framework based on the NSGA-II algorithm and a neural network prediction model to iterate the multiple initial parameter combinations, and calculate to obtain multiple predicted target parameter combinations; extract the optimal parameter combination corresponding to the target process parameter from the multiple predicted target parameter combinations according to the process priority in the parameter optimization request; In the embodiment of the present application, if there is no matching historical process plan, parameter combinations are performed on multi-dimensional data based on the NSGA-II genetic algorithm. The multi-dimensional data such as grinding pressure and rotation speed are used as decision variables, and the target process parameters such as surface roughness and processing efficiency (processing time) are used as optimization objectives (the decision variables and optimization objectives are set according to the target process parameters input by the user). A multi-objective optimization model is constructed and trained. Based on the historical instance mean or the set range, N initial individuals (parameter combinations) are randomly generated. Candidate solutions are generated through crossover and mutation operations, and non-dominated sorting is performed based on the two dimensions of surface roughness and processing efficiency. The multi-objective optimization model is iterated through the NSGA-II algorithm to predict the objective function values (the objective function values are sorted according to "the minimum surface roughness" and "the fastest processing efficiency (the minimum processing time)"), and one or more sets of Pareto optimal solutions are obtained. The solution that best meets the user's needs is selected as the optimal parameter combination corresponding to the target process parameters. This method effectively balances accuracy, efficiency, and cost, and is applicable to the automatic reasoning and optimization of silicon carbide grinding parameters under complex conditions.

[0055] Step S209, based on the optimal parameter combination, generate a process plan for the target process parameter, and calculate the quality score of the process plan based on a preset scoring rule, and record the process plan with a quality score greater than the preset scoring threshold as a historical process plan in the database; In the embodiment of the present application, when the NSGA-II genetic algorithm completes parameter combination optimization, a corresponding process plan is generated. The process plan includes specific operation steps, equipment configuration, and parameter settings. For example, for the silicon carbide grinding process, the process plan may include parameters such as the rotation speed of the grinding disk, the rotation speed of the workpiece, the grinding pressure, and the grinding time. The generated process plan is stored in the database as a historical process plan. The database adopts a relational database structure, and each process plan record includes fields such as a unique identifier, process parameters, generation time, and execution result. During the storage process, the system automatically assigns labels to the process plans for subsequent retrieval and classification. In addition, the system also records the execution situation and actual effect of the process plan to provide a basis for subsequent optimization.

[0056] However, in the above process of generating the process plan, the database may be redundant and the retrieval efficiency may decrease due to the indiscriminate recording of low-quality process plans, affecting the generation effect of the optimal parameter combination. In response to this, the present application further proposes to calculate the quality score of the process plan based on a preset scoring rule, and record the process plan with a quality score greater than the preset scoring threshold as a historical process plan in the database. Among them, the preset scoring rule includes the weight assignment of the physical indicators of the process parameters (that is, a multi-index weighted scoring mechanism).

[0057] Read the index data such as the grinding surface roughness Ra value, processing time, and yield rate recorded in the process plan, combine the weight coefficients corresponding to the priority goals set by the user, and calculate the quality score of the corresponding process plan. The quality score is verified through the database write interface, and only when the quality score exceeds the score threshold set in the current database, a storage operation is triggered, and the process plan with a quality score greater than the preset score threshold is recorded as a historical process plan in the database. The process candidate plans generated based on the NSGA-II genetic algorithm need to pass the quality score screening before entering the database to avoid contaminating the data source with low-quality parameter combinations.

[0058] A method for optimizing the process parameters of silicon carbide grinding disclosed in an embodiment of the present invention realizes the efficient optimization of the process parameters of silicon carbide grinding through a standardized data processing process and an intelligent optimization decision-making. It solves the problem of integrating multi-source heterogeneous data through a preprocessing step, establishes an association framework between parameters through multi-dimensional division, balances the optimization efficiency and accuracy through a hybrid decision-making mechanism of a matching mechanism and the NSGA-II genetic algorithm, and forms a closed loop in the entire parameter optimization process to continuously optimize the process parameters.

[0059] Embodiment 3, please refer to Figure 3 , Figure 3 which is a structural block diagram of an embodiment of a device for optimizing the process parameters of silicon carbide grinding according to the present invention. The device includes: A preprocessing module 301 for preprocessing the original parameters of the silicon carbide grinding process to obtain the preprocessed grinding process parameters; A division module 302 for dividing the preprocessed grinding process parameters according to a preset dimension classification rule to obtain multi-dimensional data; An extraction module 303 for extracting target process parameters from the parameter optimization request in response to the user's parameter optimization request; A judgment module 304 for judging whether there is a historical process plan matching the target process parameters in the database; if so, execute a determination module 305; if not, execute an optimization module 306; A determination module 305 for extracting a target process plan from the matching historical process plans according to the priority goals set by the user, and determining the process parameters in the target process plan as the optimal parameter combination corresponding to the target process parameters; An optimization module 306 for performing parameter combination on the multi-dimensional data based on the NSGA-II genetic algorithm to obtain the optimal parameter combination corresponding to the target process parameters.

[0060] In an optional embodiment, the preprocessing module 301 includes: A cleaning sub-module, configured to perform data cleaning on the original parameters of the silicon carbide grinding process obtained, so as to remove noise, anomalies, invalid or duplicate data, and obtain the grinding process parameters after data cleaning; A first conversion sub-module, configured to perform unit standardization operation, timestamp standardization operation, coding conversion operation and numerical reduction operation on the grinding process parameters after data cleaning, so as to convert the grinding process parameters after data cleaning into data in the same format, and obtain the grinding process parameters after format conversion; A naming sub-module, configured to perform field name standardization operation, multi-source field mapping operation, data table association modeling operation and data label consistency operation on the grinding process parameters after format conversion, so as to standardize the naming of the grinding process parameters after format conversion, and obtain the pre-processed grinding process parameters.

[0061] In an optional embodiment, the judgment module 304 includes: A second conversion sub-module, configured to convert the target process parameters into a standardized process vector in the same format as the multi-dimensional data; A calculation sub-module, configured to calculate the matching degrees between the standardized process vector and all historical process plans in the database based on a preset similarity matching mechanism, and obtain a plurality of matching degree values; A judgment sub-module, configured to judge whether there is a historical process plan matching the target process parameters in the database according to the magnitude relationship between the plurality of matching degree values and a preset matching degree threshold.

[0062] In an optional embodiment, the optimization module 306 includes: A generation sub-module, configured to randomly generate a plurality of initial parameter combinations in the multi-dimensional data according to the target process parameters; An iteration sub-module, configured to input the plurality of initial parameter combinations into a multi-objective optimization framework based on the NSGA-II algorithm and a neural network prediction model to iterate on the plurality of initial parameter combinations, and calculate and obtain a plurality of predicted target parameter combinations; An extraction sub-module, configured to extract the optimal parameter combination corresponding to the target process parameters from the plurality of predicted target parameter combinations according to the process priority in the parameter optimization request.

[0063] In an optional embodiment, the optimization module 306 further includes: A recording sub-module, configured to generate a process plan for the target process parameters based on the optimal parameter combination, and record the process plan as a historical process plan in the database.

[0064] Optionally, the recording sub-module includes: A calculation unit that calculates the quality score of the process plan based on a preset scoring rule; A recording unit that records the process plan with a quality score greater than a preset scoring threshold into the database as a historical process plan.

[0065] Example 4: The embodiments of the present invention also provide an electronic device, including a memory and a processor. A computer program is stored in the memory. When the computer program is executed by the processor, the processor executes the steps of an optimization method for silicon carbide grinding process parameters according to any one of the embodiments.

[0066] Example 5: The embodiments of the present invention also provide a computer storage medium, on which a computer program is stored. When the computer program is executed by the processor, the steps of an optimization method for silicon carbide grinding process parameters according to any one of the embodiments are implemented.

[0067] Example 6: The embodiments of the present invention also provide a computer program product, on which a computer program is stored. When the computer program is executed by the processor, the steps of an optimization method for silicon carbide grinding process parameters according to any one of the embodiments are implemented.

[0068] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments and will not be repeated here.

[0069] In several embodiments provided by the present application, it should be understood that the methods, devices, electronic devices, and storage media disclosed by the present invention can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0070] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0071] In addition, in each embodiment of the present invention, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0072] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned readable storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store program codes.

[0073] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A method for optimizing process parameters of silicon carbide grinding, characterized in that, Including: Preprocess the original parameters of the silicon carbide grinding process to obtain the preprocessed grinding process parameters; According to the preset dimension classification rules, divide the preprocessed grinding process parameters to obtain multi-dimensional data; In response to the user's parameter optimization request, extract the target process parameters from the parameter optimization request; Determine whether there is a historical process plan in the database that matches the target process parameters; If so, extract the target process plan from the matching historical process plans according to the priority target set by the user, and determine the process parameters in the target process plan as the optimal parameter combination corresponding to the target process parameters; If not, based on the NSGA-II genetic algorithm, perform parameter combination on the multi-dimensional data to obtain the optimal parameter combination corresponding to the target process parameters.

2. The method for optimizing the process parameters of silicon carbide grinding according to claim 1, wherein Preprocess the original parameters of the silicon carbide grinding process to obtain the preprocessed grinding process parameters, including: Perform data cleaning on the obtained original parameters of the silicon carbide grinding process to remove noise, anomalies, invalid or duplicate data, and obtain the ground process parameters after data cleaning; Perform unit standardization operation, timestamp standardization operation, encoding conversion operation and numerical reduction operation on the ground process parameters after data cleaning to convert the ground process parameters after data cleaning into data in the same format, and obtain the ground process parameters after format conversion; Perform field name standardization operation, multi-source field mapping operation, data table association modeling operation and data label consistency operation on the ground process parameters after format conversion to standardize the naming of the ground process parameters after format conversion, and obtain the preprocessed grinding process parameters.

3. The method for optimizing the process parameters of silicon carbide grinding according to claim 1, wherein, Determine whether there is a historical process plan in the database that matches the target process parameters, including: Convert the target process parameters into a standardized process vector in the same format as the multi-dimensional data; Based on the preset similarity matching mechanism, calculate the matching degrees between the standardized process vector and all historical process plans in the database to obtain multiple matching degree values; According to the magnitude relationship between the multiple matching degree values and the preset matching degree threshold, determine whether there is a historical process plan in the database that matches the target process parameters.

4. The method for optimizing the process parameters of silicon carbide grinding according to claim 1, characterized in that, Based on the NSGA-II genetic algorithm, perform parameter combination on the multi-dimensional data to obtain the optimal parameter combination corresponding to the target process parameters, including: Randomly generate multiple initial parameter combinations in the multi-dimensional data according to the target process parameters; Input the multiple initial parameter combinations into a multi-objective optimization framework based on the NSGA-II algorithm and a neural network prediction model to iterate on the multiple initial parameter combinations, and calculate to obtain multiple predicted target parameter combinations; According to the process priority in the parameter optimization request, extract the optimal parameter combination corresponding to the target process parameters from the multiple predicted target parameter combinations.

5. The method for optimizing the process parameters of silicon carbide grinding according to claim 1, characterized in that, After obtaining the optimal parameter combination corresponding to the target process parameters by performing parameter combination on the multi-dimensional data based on the NSGA-II genetic algorithm, it further includes: Generate a process plan for the target process parameters based on the optimal parameter combination, and record the process plan as a historical process plan in the database.

6. The method for optimizing the process parameters of silicon carbide grinding according to claim 5, wherein Recording the process plan as a historical process plan in the database includes: Calculate the quality score of the process plan based on a preset scoring rule; Record the process plan with a quality score greater than the preset score threshold as a historical process plan in the database.

7. An apparatus for optimizing process parameters of silicon carbide grinding, characterized in that, It includes: A preprocessing module for preprocessing the original parameters of the silicon carbide grinding process obtained to obtain preprocessed grinding process parameters; A partitioning module for partitioning the preprocessed grinding process parameters according to a preset dimension classification rule to obtain multi-dimensional data; An extraction module for extracting target process parameters from the parameter optimization request in response to a user's parameter optimization request; A judgment module for judging whether there is a historical process plan in the database that matches the target process parameters; if so, execute the determination module; If not, execute the optimization module; A determination module for extracting a target process plan from the matching historical process plans according to the priority target set by the user, and determining the process parameters in the target process plan as the optimal parameter combination corresponding to the target process parameters; An optimization module for performing parameter combination on the multi-dimensional data based on the NSGA-II genetic algorithm to obtain the optimal parameter combination corresponding to the target process parameters.

8. An electronic device, characterized in that, It includes a processor and a memory, and the memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the method according to any one of claims 1-6 is run.

9. A storage medium, on which a computer program is stored, characterized in that, When the computer program is executed by the processor, the method according to any one of claims 1-6 is run.

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

Images