Part heat treatment control method and system and computer equipment

Abstract

The invention relates to a part heat treatment control method and system and computer equipment. Comprising the following steps: obtaining model characteristic parameters of a sample part, determining material performance parameters of the sample part after simulation heat treatment, and then constructing a response relationship among sample appearance characteristics, sample process characteristics and the material performance parameters based on machine learning; and when the target appearance characteristics and the target performance parameters of the target part associated with the type of the sample part are determined, the corresponding target process characteristics can be quickly determined through the response relationship, so that heat treatment control is carried out on the target part according to the target process characteristics. By adopting the method, the heat treatment process of the part can be efficiently and accurately determined.

Description

Technical Field

[0001] The present application relates to the field of industrial control technology, and in particular to a method, system and computer equipment for controlling heat treatment of parts. Background Art

[0002] Heat treatment is an important process method that changes the physical and chemical properties of metals by heating and cooling them. Heat treatment can be used to regulate the mechanical properties and organizational structure of metal materials to obtain ideal material processing properties. It is widely used in mechanical manufacturing, aerospace, automobile manufacturing, tool manufacturing and other fields.

[0003] At present, the formulation and setting of heat treatment processes lack scientific basis and are largely restricted by the workers' practical experience and operating level. The work is mainly completed through simple PID adjustment blocks in PLC. For parts of different materials or shapes, the appropriate heating speed, holding time, cooling speed and other process parameters are not selected. There is not enough process information support for process improvement and optimization, which is prone to unstable quality during heat treatment of different batches. For new parts of the same type, it is often necessary to simulate the heat treatment process for the new parts and obtain process parameters to guide on-site production. However, the heat treatment simulation process is time-consuming. For some parts that urgently need to carry out heat treatment processes, it is impossible to formulate a reasonable heat treatment process in a timely and efficient manner. Summary of the invention

[0004] Based on this, the purpose of this application is to provide a part heat treatment control method, system, computer equipment and computer-readable storage medium that integrates simulation and machine learning, which can quickly generate target process characteristics of parts of the same type based on the material performance parameters of the parts after heat treatment, thereby stabilizing the performance of the parts after controlled heat treatment.

[0005] In a first aspect, the present application provides a method for controlling heat treatment of a part, comprising:

[0006] Obtaining model characteristic parameters of the sample parts and determining material performance parameters after the sample parts are subjected to simulated heat treatment; the model characteristic parameters include sample shape characteristics and sample process characteristics;

[0007] Construct the response relationship between sample appearance characteristics, sample process characteristics and material performance parameters based on machine learning;

[0008] When the target shape characteristics and target performance parameters of the target part are determined, the corresponding target process characteristics are determined through the response relationship; the target part is associated with the type of the sample part;

[0009] Control the heat treatment of the target parts according to the target process characteristics.

[0010] In one embodiment, determining material performance parameters of a sample part after simulated heat treatment includes: constructing a simulation model for the sample part when the simulated heat treatment is performed; determining a parameter range corresponding to a model characteristic parameter, and sampling the model characteristic parameter within the parameter range to obtain multiple sets of data sets; and processing the simulation model based on the multiple sets of data sets to obtain multiple sets of material performance parameters.

[0011] In one embodiment, the response relationship is realized by a target learning model of an artificial neural network; a response relationship between sample shape characteristics, sample process characteristics and material performance parameters is constructed based on machine learning, including: determining a training set and a test set according to the sample shape characteristics, sample process characteristics and material performance parameters; training the artificial neural network model through the training set to obtain an initial learning model; verifying the initial learning model through the test set to obtain an evaluation index; the evaluation index at least includes goodness of fit and root mean square error; evaluating the initial learning model according to the evaluation index to obtain a target learning model.

[0012] In one embodiment, the artificial neural network model includes an input layer, a hidden layer and an output layer; the input layer includes sample shape characteristics and material performance parameters; the output layer includes sample process characteristics; the number of layers and the number of units in the hidden layer are determined by a neural network architecture search method and an empirical formula; the sample shape characteristics include at least part diameter and part length; the material performance parameters include at least material strength, material hardness and grain size; the sample process characteristics include at least aging temperature, aging time and cooling rate.

[0013] In one embodiment, the initial learning model is verified by a test set to obtain evaluation indicators, including: determining the predicted value and the corresponding actual value after the test set is input into the initial learning model; determining the goodness of fit and the root mean square error according to the predicted value, the actual value and the number of samples in the test set output by the initial learning model; evaluating the initial learning model according to the goodness of fit and the root mean square error to obtain a target learning model; the above method also includes: performing denormalization processing on the predicted value output by the target learning model to obtain a heat treatment process feature; the heat treatment process feature is used as a new label for model training.

[0014] In one embodiment, heat treatment control of a target part is performed according to target process characteristics, including: during heat treatment control of the target part, actual performance parameters are monitored; actual performance parameters are used as new material performance parameters, target shape characteristics are used as new sample shape characteristics, and target process characteristics are used as new sample process characteristics; and feedback is provided to the process of constructing a response relationship between sample shape characteristics, sample process characteristics and material performance parameters based on machine learning.

[0015] In a second aspect, the present application also provides a parts heat treatment control system, the system includes a simulation analysis module, a machine learning module and an intelligent control module, wherein:

[0016] A simulation analysis module is used to obtain model characteristic parameters of sample parts and determine material performance parameters after simulated heat treatment of the sample parts; the model characteristic parameters include sample shape characteristics and sample process characteristics;

[0017] A machine learning module, for constructing a response relationship between sample shape features, sample process features and material performance parameters based on machine learning, and when determining target shape features and target performance parameters of a target part, determining the corresponding target process features through the response relationship;

[0018] The intelligent control module is used to control the heat treatment of the target parts according to the target process characteristics.

[0019] In one embodiment, the system also includes a monitoring feedback module for feeding back the monitored actual performance parameters to the machine learning module during heat treatment control of the target part according to the target process characteristics, so that the machine learning module can perform model optimization.

[0020] In a third aspect, the present application further provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned part heat treatment control method when executing the computer program.

[0021] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned part heat treatment control method are implemented.

[0022] The above-mentioned parts heat treatment control method, system, computer device and readable storage medium obtain the model characteristic parameters of the sample parts and determine the material performance parameters after the simulated heat treatment of the sample parts, and then build the response relationship between the sample appearance characteristics, sample process characteristics and material performance parameters based on machine learning, and when the target appearance characteristics and target performance parameters of the target parts associated with the type of the sample parts are determined, the corresponding target process characteristics can be quickly determined through the response relationship to control the heat treatment of the target parts according to the target process characteristics. Therefore, by associating the relationship between the heat treatment performance and characteristic parameters of parts of the same type, the problem of unstable heat treatment performance of batch parts during heat treatment is overcome, and parts with specified material properties can be obtained as required, which can greatly improve the yield of parts heat treatment; at the same time, the efficiency of heat treatment process formulation of similar parts is improved, and the cost of heat treatment simulation calculation is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A diagram showing an application environment of a method for controlling heat treatment of parts in one embodiment;

[0024] Figure 2 A schematic flow chart of a method for controlling heat treatment of parts in one embodiment;

[0025] Figure 3 is a schematic diagram of an artificial neural network model in one embodiment;

[0026] Figure 4 A structural block diagram of a parts heat treatment control system in one embodiment;

[0027] Figure 5 The schematic diagram of the execution process of a parts heat treatment control system in one embodiment;

[0028] Figure 6 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0030] The parts heat treatment control method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, and through the interaction between the terminal 102 and the computer device 104. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the computer device 104 needs to process. The data storage system can be integrated on the computer device 104, or it can be placed on the cloud or other network servers. The terminal 102 can be, but is not limited to, various parts processing devices, such as heat treatment furnaces. The computer device 104 is used to obtain the model feature parameters of the sample parts, and determine the material performance parameters after the sample parts are simulated heat treated; based on machine learning, the response relationship between the sample shape characteristics, the sample process characteristics and the material performance parameters is constructed. The computer device 104 is also used to determine the corresponding target process characteristics through the response relationship when determining the target shape characteristics and target performance parameters of the target part, so as to control the terminal 102 to heat treat the target part according to the target process characteristics. Among them, the computer device 104 is realized by an independent server or a server cluster composed of multiple servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as big data and artificial intelligence platforms.

[0031] In one embodiment, Figure 2 As shown, a method for controlling heat treatment of parts is provided. The method is implemented by a heat treatment control system for parts. The system is integrated in Figure 1 In the computer device in the embodiment, the steps include:

[0032] Step 202, obtaining model characteristic parameters of the sample part, and determining material performance parameters after the sample part is subjected to simulated heat treatment.

[0033] The model characteristic parameters include sample shape characteristics and sample process characteristics. Sample shape characteristics include at least part diameter and part length; sample process characteristics include at least aging temperature, aging time, and cooling rate; material performance parameters include at least material strength, material hardness, and grain size.

[0034] Specifically, the computer device constructs a finite element simulation model of the sample part during a specific heat treatment process according to user requirements in advance, and determines the model characteristic parameters of the sample part. For example, when the sample part is an aluminum alloy bar, the model characteristic parameters also include the part material. The computer device modifies the simulation model parameters based on the model characteristic parameters, and obtains the material performance parameters of the sample part after heat treatment simulation of the modified simulation model.

[0035] In one embodiment, determining material performance parameters of a sample part after simulated heat treatment includes: constructing a simulation model for the sample part when the simulated heat treatment is performed; determining a parameter range corresponding to a model characteristic parameter, and sampling the model characteristic parameter within the parameter range to obtain multiple sets of data sets; and processing the simulation model based on the multiple sets of data sets to obtain multiple sets of material performance parameters.

[0036] Specifically, based on actual production experience, the user specifies the parameter ranges corresponding to the sample appearance features and the sample process features, so that the computer device samples the model feature parameters within the parameter range by, for example, the Latin hypercube sampling method to form, for example, 1,000 sets of data sets. The computer device batch modifies the feature parameters in the simulation model in response to the user's program writing operation to form modified simulation models corresponding to different feature parameters. The computer device batch performs heat treatment simulation calculations on multiple sets of modified simulation models to obtain multiple sets of material performance parameters.

[0037] In this embodiment, by sampling based on different parameter ranges, a more comprehensive understanding of the behavior of the model under various conditions can be obtained, which helps to improve the prediction accuracy of the simulation model for the actual heat treatment process, and can also ensure that more diverse material performance parameters are obtained, ensuring the richness of the data source related to the subsequent target learning model.

[0038] Step 204, when constructing a response relationship between sample appearance characteristics, sample process characteristics and material performance parameters based on machine learning.

[0039] Among them, the response relationship is realized through the target learning model of artificial neural network.

[0040] Specifically, based on machine learning, the computer device uses the sample appearance characteristics and material performance parameters as the input data of the artificial neural network model, and uses the corresponding sample process characteristics after matching as the output data of the artificial neural network model, that is, the input data and output data are used as sample sets, so that the artificial neural network model is preliminarily trained through part of the input data and output data in advance to obtain an initial learning model. The computer device then verifies and evaluates the initial learning model through the remaining input data and output data to obtain the final trained target learning model. Therefore, the process of using the relevant data and machine learning models of existing sample parts to quickly predict the heat treatment process and part performance control of subsequent new parts is realized.

[0041] Step 206, when the target shape characteristics and target performance parameters of the target part are determined, the corresponding target process characteristics are determined through the response relationship.

[0042] Step 208, performing heat treatment control on the target part according to the target process characteristics.

[0043] The target part is associated with the type of the sample part.

[0044] Specifically, for a target part of the same type as the sample part, the computer device pre-acquires its known part material and target shape features, such as part diameter, part length and other information, and responds to the target performance parameters input by the user, such as material strength, material hardness, grain size and other information, and can output the corresponding target process characteristics through the response relationship embodied by the target learning model, such as heating temperature, holding time, cooling rate and other information. The computer device feeds back the target process characteristics to, for example, the control program of the heat treatment furnace to control the heat treatment furnace to heat the target part according to the target process characteristics and obtain the corresponding actual performance parameters.

[0045] In the above-mentioned part heat treatment control method, by obtaining the model characteristic parameters of the sample part and determining the material performance parameters after the simulated heat treatment of the sample part, the response relationship between the sample appearance characteristics, the sample process characteristics and the material performance parameters is then constructed based on machine learning. When the target appearance characteristics and target performance parameters of the target part associated with the type of the sample part are determined, the corresponding target process characteristics can be quickly determined through the response relationship to control the heat treatment of the target part according to the target process characteristics. Therefore, by associating the relationship between the heat treatment performance and characteristic parameters of parts of the same type, the problem of unstable heat treatment performance of batch parts during heat treatment is overcome, and parts with specified material properties can be obtained as required, which can greatly improve the yield of heat treatment of parts; at the same time, the efficiency of heat treatment process formulation of similar parts is improved, and the cost of heat treatment simulation calculation is reduced.

[0046] In one embodiment, a response relationship between sample shape features, sample process features and material performance parameters is constructed based on machine learning, including: determining a training set and a test set according to the sample shape features, sample process features and material performance parameters; training an artificial neural network model through the training set to obtain an initial learning model; verifying the initial learning model through the test set to obtain an evaluation index; and evaluating the initial learning model according to the evaluation index to obtain a target learning model.

[0047] Among them, the evaluation indicators include at least goodness of fit and root mean square error.

[0048] Specifically, the computer equipment constructs a sample set according to the sample appearance characteristics, sample process characteristics and material performance parameters, and normalizes the sample set to obtain a normalized sample set. The computer equipment divides the normalized sample set into a training set and a test set according to a preset ratio. For example, when the ratio is 8:2, the training set is 800 groups of data and the test set is 200 groups of data. The computer equipment first trains the artificial neural network model through the training set to obtain the initial learning model after training; then verifies the initial learning model through the test set to obtain the evaluation index; finally, the initial learning model is evaluated according to the goodness of fit and the root mean square error to obtain the target learning model of the part heat treatment process.

[0049] In this embodiment, by determining the training set and the test set according to the sample appearance characteristics, process characteristics and material performance parameters, it can be ensured that the artificial neural network model has high relevance and accuracy when processing data. In addition, by first verifying the initial learning model through the test set, the performance of the model can be evaluated and potential improvement points can be found, thereby optimizing the model; and then the initial learning model is evaluated through the evaluation index, so that the advantages and disadvantages of the model, such as prediction accuracy, generalization ability, etc., can be accurately understood. Therefore, the target learning model finally obtained has higher prediction ability and better performance, and can more accurately process data in practical applications.

[0050] In one embodiment, a linear function is used to normalize the sample set, and the formula is as follows:

[0051]

[0052] In the formula, X inorm represents the normalized sample data, x i is the initial sample data, x max is the maximum value in the initial sample data, x min is the minimum value in the initial sample data.

[0053] In one embodiment, the artificial neural network model includes an input layer, a hidden layer and an output layer; the input layer includes sample shape characteristics and material performance parameters; the output layer includes sample process characteristics; the number of layers and units in the hidden layer are determined by a neural network architecture search method and an empirical formula.

[0054] Among them, Figure 3 As shown, Figure 3 It is a schematic diagram of the artificial neural network model. When the input layer has 6 input layer units, it can be the part material, as well as the part diameter and part length included in the sample shape characteristics, and the material strength, material hardness, and grain size included in the material performance parameters; when the output layer has 3 input layer units, it can be the aging temperature, aging time, and cooling rate in the sample process characteristics; the empirical formula is used to estimate the maximum number of hidden layer units, and the empirical formula is:

[0055]

[0056] Among them, h is the number of hidden layer units, h is an integer, i is the number of input layer units, o is the number of output layer units, and a is a constant less than or equal to 10, that is, the number of hidden layer units h≤13.

[0057] In one embodiment, the initial learning model is verified by a test set to obtain evaluation indicators, including: determining the predicted value and the corresponding actual value after the test set is input into the initial learning model; determining the goodness of fit and the root mean square error based on the predicted value, the actual value and the number of samples in the test set output by the initial learning model; evaluating the initial learning model based on the goodness of fit and the root mean square error to obtain a target learning model.

[0058] Specifically, the computer device verifies the trained initial learning model through the test set to obtain the evaluation index, and evaluates the trained initial learning model according to the goodness of fit and the root mean square error to obtain the target learning model.

[0059] The goodness of fit calculation formula is as follows:

[0060]

[0061] In the formula, n is the number of samples in the test set, y i is the actual value corresponding to the test sample, is the predicted value output by the test sample model, is the mean of the actual value of the test sample, R 2 It is the goodness of fit, and its value range is [0,1]. The closer its value is to 1, the higher the prediction accuracy of the model.

[0062] The root mean square error RMSE calculation formula is as follows:

[0063]

[0064] In the formula, n is the number of samples in the test set, y i is the actual value of the test sample; The predicted value output by the test sample model. If the goodness of fit is low or the root mean square error is high, this may indicate that there are problems with the model, such as underfitting, overfitting, or improper feature selection. These indicators need to be used to identify the direction in which the model needs to be improved.

[0065] In this embodiment, the prediction performance of the model can be accurately evaluated by calculating the goodness of fit and the root mean square error. The goodness of fit reflects the strength of the linear relationship between the model prediction value and the actual value, while the root mean square error quantifies the average error between the model prediction value and the actual value. These indicators can reveal the accuracy and reliability of the model.

[0066] In one embodiment, the above method further includes: performing denormalization processing on the predicted value output by the target learning model to obtain the heat treatment process characteristics.

[0067] Specifically, the computer device performs a denormalization process on the predicted value output by the target learning model according to the denormalization process function, and the method is: ipred =(y max -y min )*y ipred_n +y min

[0068] In the formula, y ipred is the heat treatment process characteristics predicted by the denormalized model, y ipred_n is the predicted value output by the target learning model, y max is the maximum value of the sample, y min is the minimum value of the sample. The heat treatment process features are used as new labels for model training; the heat treatment process features include the target process features corresponding to the target parts. Therefore, the denormalization process enables the prediction results to be restored to the range of actual physical quantities or engineering parameters, so that the results are more practical.

[0069] In one embodiment, heat treatment control of a target part is performed according to target process characteristics, including: during heat treatment control of the target part, actual performance parameters are monitored; actual performance parameters are used as new material performance parameters, target shape characteristics are used as new sample shape characteristics, and target process characteristics are used as new sample process characteristics; and feedback is provided to the process of constructing a response relationship between sample shape characteristics, sample process characteristics and material performance parameters based on machine learning.

[0070] Specifically, in the process of controlling the heat treatment furnace to heat the target part according to the target process characteristics, the temperature history data of the heat treatment furnace and the actual performance parameters of the part after heat treatment can be collected in real time. Therefore, the collected data, the corresponding target appearance characteristics and the target process characteristics can be used as feedback data, and the sample set can be updated accordingly, so that the updated sample set can be used to return to step 204 to continue training the target learning model to obtain a more accurate learning model.

[0071] In this embodiment, by feeding back data such as actual performance parameters into the model, the model can gradually learn the relationship between the actual process and material performance, thereby improving the accuracy and reliability of the prediction. In addition, real-time monitoring and feedback of performance parameters can help to timely discover and correct abnormalities in the production process, improve process control capabilities and part quality management levels.

[0072] In one embodiment, a parts heat treatment control system is provided, such as Figure 4As shown, the system includes a simulation analysis module, a machine learning module and an intelligent control module, wherein: the simulation analysis module is used to obtain the model characteristic parameters of the sample parts and determine the material performance parameters after the sample parts are subjected to simulated heat treatment; the model characteristic parameters include sample shape characteristics and sample process characteristics; the machine learning module is used to construct a response relationship between the sample shape characteristics, the sample process characteristics and the material performance parameters based on machine learning, and when determining the target shape characteristics and target performance parameters of the target parts, determine the corresponding target process characteristics through the response relationship; the intelligent control module is used to control the heat treatment of the target parts according to the target process characteristics.

[0073] Specifically, the actual execution process of the parts heat treatment control system is as follows: Figure 5 As shown. The simulation model of the sample parts undergoing simulated heat treatment is constructed through the simulation analysis module, and the model characteristic parameters are extracted; the parameter range corresponding to the model characteristic parameters is determined, and the model characteristic parameters are sampled within the parameter range to obtain multiple sets of data sets; the characteristic parameters in the simulation model are modified in batches to form modified simulation models corresponding to different characteristic parameters; heat treatment simulation calculations are performed on multiple sets of modified simulation models in batches to obtain multiple sets of material performance parameters.

[0074] The machine learning module constructs a sample set by combining the sample appearance features, sample process features and corresponding material performance parameters; the sample set is normalized and divided into a training set and a test set; the training set and the test set are trained, verified and evaluated by an artificial neural network model to obtain a target learning model. The predicted value output by the target learning model is denormalized by the intelligent control module to obtain the target process features corresponding to the target part; the target process features are fed back to the control program to heat the target part and obtain the corresponding actual performance parameters.

[0075] In this embodiment, the performance parameters of the material can be optimized by heat treating the sample parts through the simulation analysis module to meet specific working conditions and requirements. The machine learning module can build the response relationship between the appearance characteristics, process characteristics and material properties based on historical data, thereby optimizing the process flow and reducing the cost of trial and error. The intelligent control module can accurately control the heat treatment of the target parts according to the target process characteristics to ensure the quality and performance of the parts.

[0076] In one embodiment, reference Figure 5 As shown, the system also includes a monitoring feedback module, which is used to feed back the monitored actual performance parameters to the machine learning module during the heat treatment control of the target part according to the target process characteristics, so that the machine learning module can optimize the model.

[0077] Among them, the collected data, the corresponding target appearance features and the target process features are used as feedback data through the monitoring feedback module, and fed back to the simulation analysis module and the machine learning module, such as the CAE simulation module. The machine learning module updates the sample set based on the feedback data, realizes iterative optimization training of the target learning model, obtains a more accurate learning model, and continues to control the heating of the heat treatment furnace through the intelligent control module to obtain more accurate part performance control, thus realizing a high-precision intelligent closed-loop control system.

[0078] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0079] Based on the same inventive concept, the embodiment of the present application also provides a part heat treatment control device for implementing the above-mentioned part heat treatment control method. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more embodiments of the part heat treatment control device provided below can refer to the limitations of the part heat treatment control method above, and will not be repeated here.

[0080] In one embodiment, a part heat treatment control device is provided, comprising: an initial data acquisition module, a target machine code determination module and a target machine code loading module, wherein:

[0081] The sample part processing module is used to obtain the model characteristic parameters of the sample part and determine the material performance parameters after the sample part is subjected to simulated heat treatment; the model characteristic parameters include the sample shape characteristics and the sample process characteristics.

[0082] The response relationship determination module is used to construct the response relationship between sample appearance characteristics, sample process characteristics and material performance parameters based on machine learning.

[0083] The target part processing module is used to determine the corresponding target process characteristics through the response relationship when the target shape characteristics and target performance parameters of the target part are determined; and to perform heat treatment control on the target part according to the target process characteristics.

[0084] Each module in the above-mentioned parts heat treatment control can be implemented in whole or in part by software, hardware and their combination. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each of the above modules.

[0085] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store a target learning model. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a part heat treatment control method is implemented.

[0086] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0087] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0088] In one embodiment, a computer-readable storage medium is provided, storing a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0089] In one embodiment, a computer program product or computer program is provided, the computer program product or computer program includes computer instructions, the computer instructions are stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the steps in the above-mentioned method embodiments.

[0090] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited thereto.

[0091] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0092] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for controlling heat treatment of parts, characterized in that: The method comprises: Obtaining model characteristic parameters of a sample part, and determining material performance parameters of the sample part after simulated heat treatment; the model characteristic parameters include sample shape characteristics and sample process characteristics; Constructing a response relationship between the sample appearance characteristics, the sample process characteristics and the material performance parameters based on machine learning; When determining the target shape feature and the target performance parameter of the target part, determining the corresponding target process feature through the response relationship; the target part is associated with the type of the sample part; The target part is heat treated and controlled according to the target process characteristics.

2. The method according to claim 1, characterized in that The determining of the material performance parameters of the sample part after the simulated heat treatment comprises: Constructing a simulation model for the sample part when undergoing simulated heat treatment; Determining a parameter range corresponding to the model characteristic parameter, and sampling the model characteristic parameter within the parameter range to obtain multiple sets of data sets; The simulation model is processed based on multiple sets of data sets to obtain multiple sets of material performance parameters.

3. The method according to claim 1, characterized in that: The response relationship is realized by a target learning model of an artificial neural network; the response relationship between the sample appearance feature, the sample process feature and the material performance parameter is constructed based on machine learning, including: Determine the training set and the test set according to the sample appearance characteristics, sample process characteristics and material performance parameters; The artificial neural network model is trained by using the training set to obtain an initial learning model; The initial learning model is verified by the test set to obtain evaluation indicators; the evaluation indicators at least include goodness of fit and root mean square error; The initial learning model is evaluated according to the evaluation index to obtain a target learning model.

4. The method according to claim 3, characterized in that The artificial neural network model includes an input layer, a hidden layer and an output layer; the input layer includes sample shape characteristics and material performance parameters; the output layer includes sample process characteristics; the number of layers and the number of units in the hidden layer are determined by a neural network architecture search method and an empirical formula; the sample shape characteristics include at least part diameter and part length; the material performance parameters include at least material strength, material hardness and grain size; the sample process characteristics include at least aging temperature, aging time and cooling rate.

5. The method according to claim 3, characterized in that: The initial learning model is verified by the test set to obtain evaluation indicators, including: Determine a predicted value and a corresponding actual value after the test set is input into the initial learning model; Determine the goodness of fit and the root mean square error according to the predicted value output by the initial learning model, the actual value and the number of samples in the test set; Evaluate the initial learning model according to the goodness of fit and the root mean square error to obtain a target learning model; The method also includes: performing a denormalization process on the predicted value output by the target learning model to obtain a heat treatment process feature; the heat treatment process feature is used as a new label for model training.

6. The method according to claim 1, characterized in that The step of controlling the heat treatment of the target part according to the target process characteristics comprises: In controlling the heat treatment of the target part, monitoring actual performance parameters; Using the actual performance parameters as new material performance parameters, using the target shape characteristics as new sample shape characteristics, and using the target process characteristics as new sample process characteristics; The process of feeding back to constructing the response relationship between the sample appearance characteristics, the sample process characteristics and the material performance parameters based on machine learning continues.

7. A parts heat treatment control system, characterized in that: The system includes a simulation analysis module, a machine learning module and an intelligent control module, wherein: A simulation analysis module, used to obtain model characteristic parameters of a sample part and determine material performance parameters after the sample part is subjected to simulated heat treatment; the model characteristic parameters include sample shape characteristics and sample process characteristics; A machine learning module, for constructing a response relationship between the sample shape feature, the sample process feature and the material performance parameter based on machine learning, and when determining a target shape feature and a target performance parameter of a target part, determining a corresponding target process feature through the response relationship; The intelligent control module is used to control the heat treatment of the target part according to the target process characteristics.

8. The system according to claim 1, characterized in that The system also includes a monitoring feedback module, which is used to feed back the monitored actual performance parameters to the machine learning module during the heat treatment control of the target part according to the target process characteristics, so that the machine learning module can perform model optimization.

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

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

Images