Method for determining a control parameter, computer device and machine-readable storage medium

Abstract

Embodiments of the present application provide a kind of regulation parameter determination method, computer equipment and machine readable storage medium, belong to industrial production technical field.The method includes determining initial population for genetic algorithm, and the initial population is carried out by executing the genetic algorithm based on the optimization target of pre-set and is selected iteration;In the case where iteration number is equal to pre-set maximum genetic number, the first optimal solution obtained by executing genetic algorithm is obtained as current optimal solution;When current optimal solution does not satisfy optimization target, determine rebirth chromosome based on the chromosome eliminated in selection iteration, update initial population, and re-execute the genetic algorithm based on optimization target to carry out the initial population and is selected iteration, in the case where iteration number is equal to maximum genetic number, the second optimal solution obtained by re-executing genetic algorithm is obtained, and current optimal solution is updated based on the second optimal solution;Otherwise, current optimal solution is used as target control parameter.Intelligence and accuracy of regulation parameter determination are improved.

Description

Technical Field

[0001] This invention relates to the field of industrial production technology, and more specifically to a method for determining control parameters, a computer device, and a machine-readable storage medium. Background Technology

[0002] The Industrial Internet is a new type of infrastructure, application model, and industrial ecosystem that deeply integrates next-generation information and communication technologies with the industrial economy. It builds a brand-new manufacturing and service system covering the entire industrial chain and value chain by comprehensively connecting people, machines, things, and systems.

[0003] However, in the current technological landscape, many industrial equipment manufacturers still rely on traditional methods, such as expert experience and manual judgment, to monitor and manage their internal equipment. For example, in traditional machinery production processes, equipment parameters are manually adjusted, and the machinery operates according to these pre-set parameters to ensure that the resulting parts meet specifications. However, different raw materials require different optimal control parameters. Manually setting equipment control parameters demands a high level of expertise and experience from operators, and errors in human judgment can easily affect the final product's pass rate. Therefore, finding better ways to determine the control parameters for each piece of equipment remains a challenge. Summary of the Invention

[0004] To address the aforementioned shortcomings in the prior art, the purpose of this invention is to provide a method for determining control parameters, a computer device, and a machine-readable storage medium.

[0005] To achieve the above objectives, a first aspect of the present invention provides a method for determining control parameters, comprising:

[0006] Determine the initial population for the genetic algorithm, and perform optimal iteration on the initial population by executing a genetic algorithm based on a preset optimization objective;

[0007] If the number of iterations equals the preset maximum number of genetic iterations, obtain the first optimal solution obtained by executing the genetic algorithm, and use it as the current optimal solution.

[0008] If the current optimal solution does not meet the optimization objective, a regenerated chromosome is determined based on the chromosomes eliminated in the optimization iteration. The initial population is updated based on the regenerated chromosome and the current optimal solution. The genetic algorithm based on the optimization objective is re-executed to perform optimization iteration on the initial population. If the number of iterations is equal to the maximum number of genetic iterations, the second optimal solution obtained by re-executing the genetic algorithm is obtained. The current optimal solution is updated based on the second optimal solution.

[0009] If the current optimal solution satisfies the optimization objective, the current optimal solution is used as the target adjustment parameter.

[0010] In this embodiment of the invention, determining the regenerated chromosome based on the chromosomes eliminated in the optimal iteration includes:

[0011] Based on a preset chromosome selection range, target chromosomes are selected from all eliminated chromosomes in the optimization iteration. Target chromosomes include chromosomes that have been eliminated at least partially in the preset number of iterations.

[0012] Reverse all gene loci on the target chromosome to generate a regenerated chromosome.

[0013] In this embodiment of the invention, determining the initial population for the genetic algorithm includes:

[0014] Obtain all historical control parameters;

[0015] Key indicators are determined based on all historical control parameters.

[0016] Key indicators are preprocessed to obtain standardized data;

[0017] The initial population can be determined from standardized data based on preset population election rules, or the initial chromosomes can be randomly generated from the data range of each key indicator to form and determine the initial population.

[0018] In this embodiment of the invention, the optimization objectives include:

[0019] The current optimal solution corresponds to a product qualification rate that reaches the first preset qualification rate;

[0020] Alternatively, the product qualification rate corresponding to the current optimal solution reaches the first preset qualification rate, and the current optimal solution is within the data range of each key indicator and / or the current optimal solution meets the preset mechanism rules;

[0021] Among them, the key indicators are determined based on all historical control parameters, and the mechanism rules include pre-set standards that meet actual production needs or actual production conditions.

[0022] In this embodiment of the invention, the key indicators are determined in the following ways:

[0023] The weights of all historical control parameters are determined based on the rough set attribute reduction algorithm and the preset mechanism rules.

[0024] Key indicators among all historical control parameters are determined based on weights.

[0025] In this embodiment of the invention, the data range of each key indicator is determined by using a clustering algorithm and a linear regression algorithm with a second preset pass rate as the target to filter the standardized data of each key indicator.

[0026] In this embodiment of the invention, the product qualification rate corresponding to the current optimal solution is obtained in the following way:

[0027] Input the current optimal solution into the parameter evaluation model to obtain the product qualification rate corresponding to the current optimal solution output by the parameter evaluation model.

[0028] In this embodiment of the invention, the parameter evaluation model is trained in the following manner:

[0029] Labels are set for each historical control parameter based on the product qualification rate corresponding to all historical control parameters.

[0030] The historical control parameters after label setting are input into the initial XGBoost algorithm model, and the initial XGBoost algorithm model is trained to obtain the preset parameter evaluation model.

[0031] A second aspect of the present invention provides a computer device comprising: a memory, a processor, and a program stored in the memory and executable on the processor, the program being configured to implement the steps of the control parameter determination method as described in the above embodiments.

[0032] A third aspect of the present invention provides a machine-readable storage medium storing instructions that cause a machine to perform the steps of the control parameter determination method as described in the above embodiments.

[0033] The above technical solution determines the initial population for the genetic algorithm. The initial population is then iteratively optimized using a genetic algorithm based on a preset optimization objective. If the number of iterations equals the preset maximum number of genetic iterations, the first optimal solution obtained from the genetic algorithm is acquired and designated as the current optimal solution. If the current optimal solution does not satisfy the optimization objective, a regenerated chromosome is determined based on the chromosomes eliminated during the optimization iteration. The initial population is updated based on the regenerated chromosome and the current optimal solution, and the genetic algorithm based on the optimization objective is re-executed to iterate on the initial population. If the number of iterations equals the maximum number of genetic iterations, the second optimal solution obtained from the re-execution of the genetic algorithm is acquired, and the current optimal solution is updated based on the second optimal solution. If the current optimal solution satisfies the optimization objective, it is used as the target control parameter. Compared to current manual control and adjustment systems, this system improves the intelligence and accuracy of determining the control parameters of production equipment. Furthermore, the automated method of determining control parameters eliminates the need for manual intervention, saving human resources while overcoming the drawbacks of long manual inspection times and large errors. This effectively improves the product qualification rate. Moreover, by determining the control parameters, it provides reliable information and guiding data for workers lacking extensive experience, accelerating production speed and ensuring production quality.

[0034] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0035] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0036] Figure 1 This is a flowchart illustrating a method for determining control parameters according to an embodiment of the present invention. Detailed Implementation

[0037] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0038] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0039] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0040] Figure 1 This is a flowchart illustrating a method for determining control parameters according to an embodiment of the present invention. Figure 1 As shown, in this embodiment of the invention, a method for determining control parameters is provided. Taking the application of this method to a processor as an example, the method may include the following steps:

[0041] Step S100: Determine the initial population for the genetic algorithm, and perform optimization iteration on the initial population by executing a genetic algorithm based on a preset optimization objective;

[0042] It should be noted that genetic algorithms are global optimization algorithms that can generate high-quality solutions to optimization and search problems without needing to know the internal logic between various indicators. Genetic algorithms borrow from the perspective of biological genetics, using mechanisms such as natural selection, heredity, and mutation. Starting from any initial population, through random selection, crossover, and mutation operations, a group of individuals better adapted to the environment is generated, allowing the population to evolve into increasingly better regions in the search space. This process continues generation after generation, eventually converging to a group of individuals best adapted to the environment, thus obtaining the optimal solution to the problem. The initial population is the set of individuals used for individual optimization in the genetic algorithm. In this embodiment, the initial population used for the genetic algorithm can be determined from historical control parameters. The method for determining the initial population can be based on a preset population election rule, selecting a certain number of control parameters from historical control parameters as the initial population.

[0043] It should be noted that the optimization objective is determined based on the requirements of the problem itself and serves as the target for individual optimization in the genetic algorithm. The optimal solution obtained by the genetic algorithm must satisfy this optimization objective. The optimization objective is pre-set based on actual needs, with a pass rate as the objective. Here, the pass rate refers to the pass rate of the final products produced by the machinery when the control parameters corresponding to the individuals in the optimal solution of the genetic algorithm are applied in the actual production process.

[0044] Step S200: If the number of iterations equals the preset maximum number of genetic iterations, obtain the first optimal solution obtained by executing the genetic algorithm, and use it as the current optimal solution;

[0045] It should be noted that the genetic algorithm iterates through selection, crossover, and mutation operations on the initial population based on the optimization objective until an optimal solution that satisfies the objective is obtained. However, in practical applications, the final optimal solution may still not satisfy the optimization objective. In this case, further selection, crossover, and mutation operations will not significantly change the optimal solution. In this embodiment, a preset maximum number of genetic iterations limits the number of iterations for the genetic algorithm. If the number of iterations is less than the maximum number of genetic iterations, the optimization iteration continues. The first optimal solution refers to the optimal solution when the fitness of individuals and the fitness of the population no longer increase after the genetic algorithm's optimization iterations based on the initial population, or the optimal solution when the maximum number of genetic iterations is reached. When the number of iterations equals the preset maximum number of genetic iterations, the first optimal solution obtained by executing the genetic algorithm is taken as the current optimal solution.

[0046] Step S300: If the current optimal solution does not meet the optimization objective, determine the regenerated chromosome based on the chromosomes eliminated in the optimization iteration, update the initial population based on the regenerated chromosome and the current optimal solution, and re-execute the genetic algorithm based on the optimization objective to perform optimization iteration on the initial population. If the number of iterations is equal to the maximum number of genetic iterations, obtain the second optimal solution obtained by re-executing the genetic algorithm, and update the current optimal solution based on the second optimal solution.

[0047] It should be noted that since genetic algorithms cannot directly process parameters in the problem space, they must encode the problem to be solved as chromosomes in the genetic space. A chromosome consists of multiple binary-coded gene bits, for example, 1010111. When the number of iterations equals this maximum number of genetic iterations, and the current optimal solution of the genetic algorithm still does not satisfy the optimization objective, a regenerated chromosome will be determined based on the chromosomes eliminated in the optimization iteration.

[0048] It should be noted that since the current optimal solution of the genetic algorithm cannot satisfy the optimization objective, i.e., the fitness of individuals in the current iteration population is not high enough, it is necessary to update the individuals to improve the fitness of individuals in the population. In this embodiment, individuals are updated by updating the initial population. The regenerated chromosome and the current optimal solution are used together as the new initial population, and the genetic algorithm is re-performed for optimal selection iteration. The second optimal solution refers to the optimal solution when the fitness of individuals and the fitness of the population no longer increase after re-performing the optimal selection iteration based on the updated initial population, or the optimal solution when the maximum number of inheritances is reached. Specifically, the initial population is updated based on the regenerated chromosome and the current optimal solution, and the genetic algorithm based on the optimization objective is re-executed to perform optimal selection iteration on the initial population. When the number of iterations equals the maximum number of inheritances, the second optimal solution obtained by re-executing the genetic algorithm is obtained, and the current optimal solution is updated based on the second optimal solution.

[0049] Step S400: If the current optimal solution satisfies the optimization objective, the current optimal solution is used as the target adjustment parameter.

[0050] If the current optimal solution obtained during the re-optimization iteration of the genetic algorithm based on the updated initial population still does not meet the optimization objective, a regenerated chromosome will be determined based on the chromosomes eliminated in the re-optimization iteration, and the initial population will be updated again. This process will continue until a current optimal solution that meets the optimization objective is obtained. The current optimal solution that meets the optimization objective will be used as the target control parameter.

[0051] In the above scheme, an initial population for the genetic algorithm is determined, and the initial population is iteratively selected by executing a genetic algorithm based on a preset optimization objective. When the number of iterations equals the preset maximum number of genetic iterations, the first optimal solution obtained by executing the genetic algorithm is obtained as the current optimal solution. When the current optimal solution does not meet the optimization objective, a regenerated chromosome is determined based on the chromosomes eliminated in the selection iteration. The initial population is updated based on the regenerated chromosome and the current optimal solution, and the genetic algorithm based on the optimization objective is re-executed to iterate the initial population. When the number of iterations equals the maximum number of genetic iterations, the second optimal solution obtained by re-executing the genetic algorithm is obtained, and the current optimal solution is updated based on the second optimal solution. When the current optimal solution meets the optimization objective, the current optimal solution is used as the target control parameter. Compared to current manual control and adjustment systems, this system improves the intelligence and accuracy of determining the control parameters of production equipment. Furthermore, the automated method of determining control parameters eliminates the need for manual intervention, saving human resources while overcoming the drawbacks of long manual inspection times and large errors. This effectively improves the product qualification rate. Moreover, by determining the control parameters, it provides reliable information and guiding data for workers lacking extensive experience, accelerating production speed and ensuring production quality.

[0052] In one embodiment, determining the regenerated chromosome based on the chromosomes eliminated in the selection iteration includes:

[0053] Based on a preset chromosome selection range, target chromosomes are selected from all eliminated chromosomes in the optimization iteration. Target chromosomes include chromosomes that have been eliminated at least partially in the preset number of iterations.

[0054] Reverse all gene loci on the target chromosome to generate a regenerated chromosome.

[0055] It should be noted that the regenerated chromosome is determined based on chromosomes eliminated during the optimization iteration. Specifically, in this embodiment, the target chromosome is selected from all chromosomes eliminated during the optimization iteration based on a preset chromosome selection range. All gene positions of the target chromosome are reversed to generate the regenerated chromosome. The preset chromosome selection range can be determined according to the actual application scenario and experience. For example, it can be that after obtaining the chromosome eliminated in the first optimization iteration, all gene positions of the eliminated chromosome are reversed to obtain the regenerated chromosome. For example, if the chromosome includes 000111, the reversed chromosome is 111000, and chromosome 111000 is used as the regenerated chromosome.

[0056] In this embodiment, the genetic algorithm is improved by reversing all gene positions of some eliminated chromosomes to obtain regenerated chromosomes. This increases the individual diversity of the initial population of the genetic algorithm, increases the probability of improving the fitness of individuals in the genetic algorithm, and thus accelerates the speed of the best-fit iteration to meet the optimization goal.

[0057] In one embodiment, determining the initial population for the genetic algorithm includes:

[0058] Obtain all historical control parameters;

[0059] Key indicators are determined based on all historical control parameters.

[0060] Key indicators are preprocessed to obtain standardized data;

[0061] The initial population can be determined from standardized data based on preset population election rules, or the initial chromosomes can be randomly generated from the data range of each key indicator to form and determine the initial population.

[0062] In this embodiment, it should be noted that historical control parameters include reference values ​​such as temperature, air pressure, magnetic powder quantity, carbon content, current, and pressure, used for equipment control during the production process of mechanical equipment. These historical control parameters are stored in a pre-set parameter database, including parameter settings for controlling the mechanical equipment over a relatively long period. The control parameters are recorded along with production activities on an industrial internet platform associated with the production equipment. Data needs to be extracted from the industrial internet platform beforehand and stored in this parameter database to obtain the historical control parameters. Specifically, the process of storing historical control parameters in the parameter database includes collecting equipment operating condition data (which may be stored on platforms such as MySQL, Oracle, or SAP) into topic data message groups through NiFi customized components and Kafka. Then, the collected data files are uploaded to the Minio file system, and a scheduled execution script is set to download and upload the Minio files from the Minio file system to HDFS (Hadoop Distributed File System). The HDFS path is the file path of the Hive (data warehouse tool) tables, and the Hive metadata is refreshed and stored in the parameter database. In one embodiment, in order to quickly find the data source when performing control parameter analysis, after extracting the equipment data, the equipment condition data of various industrial equipment will be screened and classified to determine the data type corresponding to all equipment condition data. The classified data will be transmitted to the parameter database to provide a reliable data source for data analysis.

[0063] Key performance indicators (KPIs) refer to the parameters that are more important in practical applications among all historical control parameters, and the control parameters corresponding to these KPIs have a greater impact on the entire production activity. Specifically, KPIs can be determined based on the actual application scenario. After determining the KPIs, data preprocessing is performed. Parameter data of different data types are placed in a single table, and null values, unnecessary data, and obviously problematic data are deleted. Preprocessing methods such as data cleaning, data integration, data transformation, and data reduction can be used to obtain standardized data. Population election rules can include random generation. For example, based on inherent knowledge of the problem, efforts are made to grasp the distribution range of the space occupied by the optimal solution in the entire problem space, and then an initial population is set within this distribution range. Another example is to first randomly generate a certain number of individuals, and then select the best individuals from them to add to the initial population. This process is iterated until the number of individuals in the initial population reaches a predetermined size. Population election rules can also be manually defined, selecting a predetermined size of individuals from all standardized data based on experience as the initial population.

[0064] In this embodiment, by identifying key indicators and preprocessing the data for these key indicators, the data cleanliness of the initial population is improved, thereby increasing the accuracy of the genetic algorithm calculation.

[0065] Specifically, key indicators are determined in the following ways:

[0066] The weights of all historical control parameters are determined based on the rough set attribute reduction algorithm and the preset mechanism rules.

[0067] Key indicators among all historical control parameters are determined based on weights.

[0068] In this embodiment, it should be noted that the mechanism rules include pre-set standards that conform to actual production needs or actual production conditions. For example, indoor temperature is generally between 24℃ and 26℃; human body temperature should be between 36℃ and 37℃, etc. These mechanism rules can be adaptively adjusted according to actual production needs. The rough set attribute reduction algorithm is a method for handling uncertain information. It can derive existing patterns from incomplete information and extract some rules from them. In this embodiment, the rough set attribute reduction algorithm and mechanism rules are used to determine the weights of historical control parameters. For example, the rough set attribute reduction algorithm is used to determine some more important data in the historical control parameters, and the mechanism rules are used to filter some data that conform to the rules. When a historical control parameter is neither more important data nor conforming to the rules, the weight is set to 1; when a historical control parameter is the more important data or conforming to the rules, the weight is set to 2; when a historical control parameter is both the more important data and conforming to the rules, the weight is set to 3. The key indicators in historical control parameters are determined based on weights. This can be done by determining the weight values ​​and setting weight thresholds in advance based on actual production needs. Historical control parameters that meet these weight thresholds are considered key indicators. For example, if the weight threshold is 2, then historical control parameters with a weight of 2 or higher are all key indicators.

[0069] In this embodiment, attribute redundancy is reduced by rough set attribute reduction algorithm, and key indicators are constrained by mechanism rules to improve the effectiveness of the initial population of the genetic algorithm.

[0070] In this embodiment, initial chromosomes can also be randomly generated within the data ranges of each key indicator to form and determine the initial population. Specifically, the data ranges of each key indicator are determined by interval filtering of the standardized data of each key indicator using a clustering algorithm and a linear regression algorithm with a second preset pass rate as the target.

[0071] It is understandable that the range of data corresponding to the same data type in standardized data may be very large, and not all data can be applied to actual production. Therefore, it is necessary to filter the standardized data to obtain a more effective data range, i.e., a data interval. The second preset pass rate refers to the expected production pass rate corresponding to the filtered standardized data when determining the data interval. In this embodiment, clustering algorithms and linear regression algorithms are used to filter the standardized data intervals with the second preset pass rate as the target, so as to obtain the data intervals for each key indicator.

[0072] In this embodiment, clustering and linear regression algorithms are used to determine data intervals, thereby improving the effectiveness and accuracy of the data intervals.

[0073] In one embodiment, the optimization objectives include:

[0074] The current optimal solution corresponds to a product qualification rate that reaches the first preset qualification rate;

[0075] Alternatively, the product qualification rate corresponding to the current optimal solution reaches the first preset qualification rate, and the current optimal solution is within the data range of each key indicator and / or the current optimal solution meets the preset mechanism rules;

[0076] Among them, the key indicators are determined based on all historical control parameters, and the mechanism rules include pre-set standards that meet actual production needs or actual production conditions.

[0077] In this embodiment, it should be noted that the first preset pass rate refers to the computational target of the genetic algorithm, that is, the product pass rate that the control parameters in the optimal solution obtained by the genetic algorithm after selecting the best from the initial population are expected to achieve in actual production. The data interval is used to impose rule-based restrictions on the optimal solution of the genetic algorithm. The data interval can represent the data standard of the control parameters; that is, the control parameters in the optimal solution must be within this data interval to conform to actual production activities. Mechanism rules include pre-set standards that conform to actual production needs or actual production conditions.

[0078] Specifically, the product qualification rate corresponding to the current optimal solution of the genetic algorithm must at least meet a first preset qualification rate. Furthermore, to improve the accuracy of the genetic algorithm, while ensuring the product qualification rate of the current optimal solution meets the first preset qualification rate, it is also necessary to, based on actual application conditions, ensure that the current optimal solution falls within the data range of each key indicator or meets at least one of the preset mechanism rules. Among these, the most accurate determination is that the current optimal solution meets the optimization objective if the product qualification rate of the current optimal solution meets the first preset qualification rate, the control parameters in the current optimal solution fall within their corresponding data ranges, and the control parameters in the current optimal solution meet the mechanism rules.

[0079] If the product qualification rate corresponding to the current optimal solution does not meet the first preset qualification rate, and / or the control parameter in the current optimal solution is not within the corresponding data range, and / or the control parameter in the current optimal solution does not meet the mechanism rule, then it is determined that the current optimal solution does not meet the optimization objective.

[0080] In this embodiment, the optimal solution of the genetic algorithm is judged by the first preset pass rate, data range and preset mechanism rules, which effectively improves the accuracy and practicality of the control parameters in the optimal solution.

[0081] In this embodiment of the invention, the product qualification rate corresponding to the current optimal solution is obtained in the following way:

[0082] Input the current optimal solution into the parameter evaluation model to obtain the product qualification rate corresponding to the current optimal solution output by the parameter evaluation model.

[0083] It should be noted that the current optimal solution is input into a preset parameter evaluation model to obtain the evaluation result. This parameter evaluation model is used to assess the pass rate corresponding to the control parameter; inputting the control parameter into the model outputs the pass rate evaluation value corresponding to that parameter. It should also be noted that during the execution of the genetic algorithm, the fitness (product pass rate) of the currently surviving chromosome solution generated in each round of optimization iteration is also calculated using this method.

[0084] In one embodiment, the parameter evaluation model is trained in the following manner:

[0085] Labels are set for each historical control parameter based on the product qualification rate corresponding to all historical control parameters.

[0086] The historical control parameters after label setting are input into the initial XGBoost algorithm model, and the initial XGBoost algorithm model is trained to obtain the preset parameter evaluation model.

[0087] In this embodiment, it should be noted that the parameter evaluation model is determined based on the XGBoost algorithm model. The parameter evaluation model is obtained after training the XGBoost algorithm model. The training data for training the XGBoost algorithm model consists of historical control parameters. The product qualification rate corresponding to the historical control parameters is used as the label for the historical control parameters. The labeled historical control parameters are input into the initial XGBoost algorithm model to complete the training of the XGBoost algorithm model. It can be understood that when training the XGBoost algorithm model, a portion of the historical control parameters can be extracted as training data, and another portion can be extracted as test data. The test data and training data are used as input data to the XGBoost algorithm model respectively, and the output results are obtained separately. Based on the difference between the output results of the training data and the output results of the test data, the XGBoost algorithm model training is optimized and improved.

[0088] In this embodiment, a parameter evaluation model is determined based on the XGBoost algorithm model, and the optimal solution obtained by the genetic algorithm is further verified to ensure the accuracy of the target regulation parameters.

[0089] This invention provides a computer device, which includes a memory, a processor, and a program stored in the memory and executable on the processor. The program is configured to implement the steps of the control parameter determination method as described in the above embodiments.

[0090] This invention provides a machine-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the control parameter determination method as described in the above embodiments.

[0091] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0092] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0093] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0094] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0095] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0096] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0097] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0098] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0099] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for determining control parameters, characterized in that, include: An initial population for the genetic algorithm is determined, and the initial population is iteratively selected by executing the genetic algorithm based on a preset optimization objective; If the number of iterations equals the preset maximum number of genetic iterations, the first optimal solution obtained by executing the genetic algorithm is taken as the current optimal solution; If the current optimal solution does not satisfy the optimization objective, a regenerated chromosome is determined based on the chromosomes eliminated in the optimization iteration. The initial population is updated based on the regenerated chromosome and the current optimal solution. The genetic algorithm based on the optimization objective is re-executed to perform optimization iteration on the initial population. If the number of iterations is equal to the maximum number of genetic iterations, a second optimal solution obtained by re-executing the genetic algorithm is obtained. The current optimal solution is updated based on the second optimal solution. If the current optimal solution satisfies the optimization objective, the current optimal solution shall be used as the target adjustment parameter; The determination of regenerated chromosomes based on chromosomes eliminated in the optimal iteration includes: Based on a preset chromosome selection range, a target chromosome is selected from all eliminated chromosomes in the optimization iteration, wherein the target chromosome includes chromosomes that have been eliminated at least partially in a preset number of iterations; All gene loci of the target chromosome are reversed to generate a regenerated chromosome.

2. The method for determining control parameters according to claim 1, characterized in that, The determination of the initial population for the genetic algorithm includes: Obtain all historical control parameters; Key indicators are determined based on all the historical control parameters mentioned above. The key indicators are preprocessed to obtain standardized data. The initial population is determined from the standardized data based on a preset population election rule, or the initial chromosomes are randomly generated in the data range of each key indicator to form and determine the initial population.

3. The method for determining control parameters according to claim 1, characterized in that, The optimization objectives include: The product qualification rate corresponding to the current optimal solution reaches the first preset qualification rate; Alternatively, the product qualification rate corresponding to the current optimal solution reaches the first preset qualification rate, and the current optimal solution is within the data range of each key indicator and / or the current optimal solution satisfies the preset mechanism rules; The key indicators are determined based on all historical control parameters, and the mechanism rules include pre-set standards that meet actual production needs or actual production conditions.

4. The method for determining the control parameters according to claim 2 or 3, characterized in that, The key indicators were determined in the following ways: The weights of all historical control parameters are determined based on the rough set attribute reduction algorithm and the preset mechanism rules. Based on the weights, the key indicators among all historical control parameters are determined.

5. The method for determining the control parameters according to claim 2 or 3, characterized in that, The data ranges for each key indicator are determined by using a clustering algorithm and a linear regression algorithm with a second preset pass rate as targets to filter the standardized data of each key indicator within the specified ranges.

6. The method for determining control parameters according to claim 3, characterized in that, The product qualification rate corresponding to the current optimal solution is obtained in the following way: The current optimal solution is input into the parameter evaluation model to obtain the product qualification rate corresponding to the current optimal solution as output by the parameter evaluation model.

7. The method for determining control parameters according to claim 6, characterized in that, The parameter evaluation model is trained in the following way: Labels are set for each of the historical control parameters based on the product qualification rate corresponding to all the historical control parameters. The historical control parameters after label setting are input into the initial XGBoost algorithm model, and the initial XGBoost algorithm model is trained to obtain the preset parameter evaluation model.

8. A computer device, characterized in that, The computer device includes: a memory, a processor, and a program stored in the memory and executable on the processor, the program being configured to implement the steps of the method for determining control parameters as described in any one of claims 1 to 7.

9. A machine-readable storage medium, characterized in that, The machine-readable storage medium stores instructions for causing the machine to perform the steps of the method for determining control parameters as described in any one of claims 1 to 7.

Images