Steel process design method and related equipment

Abstract

The invention discloses a steel process design method and related equipment, relates to the field of steel production, and mainly aims to solve the problem that the process design for steel production depends too much on manual trial and error. The method comprises the steps that target technological parameters of target steel in the production process are determined; a neural network model is established based on the target process parameters, and the neural network model is related to the room temperature tensile property of target steel; and determining the optimal process design of the target steel through the neural network model based on a product performance demand target value. The method is used for the steel process design process.

Description

Technical Field

[0001] The present invention relates to the field of steel production, and in particular to a steel process design method and related equipment. Background Art

[0002] Current production process design for specialty steels relies primarily on traditional experience and trial-and-error methods. However, these methods require repeated testing and adjustments to achieve the optimal match between process parameters, including steelmaking composition, hot rolling finish temperature, hot rolling coiling temperature, cold rolling reduction ratio, and annealing temperature. This trial-and-error approach is not only time-consuming but also difficult to quickly respond to changes in market demand. Furthermore, the large amount of experimental materials and manpower required leads to high costs. Summary of the Invention

[0003] In view of the above problems, the present invention provides a steel process design method and related equipment, the main purpose of which is to solve the problem that the process design for steel production relies too much on manual trial and error.

[0004] To solve at least one of the above technical problems, in a first aspect, the present invention provides a steel process design method, the method comprising:

[0005] Determine the target process parameters for the target steel during production;

[0006] Establishing a neural network model based on the target process parameters, wherein the neural network model is related to the room temperature tensile properties of the target steel;

[0007] The optimal process design of the target steel is determined through the neural network model based on the target value of product performance requirements.

[0008] Optionally, determining target process parameters of the target steel during production includes:

[0009] Obtain room temperature tensile test data of target steel under different hot-dip galvanizing process conditions;

[0010] Preprocessing the room temperature tensile test data to obtain optimized room temperature tensile test data;

[0011] Obtaining the degree of correlation of the impact of the optimized room temperature tensile test data on key performance indicators;

[0012] The optimized room temperature tensile test data having an impact correlation degree greater than a preset standard is determined as the target process parameter.

[0013] Optionally, the above method further includes:

[0014] The target process parameters are one-hot encoded and normalized to form numerical inputs.

[0015] Optionally, determining the optimal process design of the target steel material through the neural network model based on the product performance requirement target value includes:

[0016] Establishing a target database based on the neural network model, wherein the target database is used to provide feedback on the relationship between the target process parameters and the room temperature tensile properties;

[0017] The target value of the product performance requirement is used as a screening condition to determine the optimal process design of the target steel through the target database.

[0018] Optionally, the above method further includes:

[0019] The performance prediction result of the process design of the target steel is determined based on the following formula taking into account the production process control error and the prediction model error:

[0020] Maximum performance prediction result of process design: μ+3σ+MSE

[0021] Average value of performance prediction results of process design: μ

[0022] Minimum performance prediction result of process design: μ-3σ-MSE

[0023] Wherein, σ is the standard deviation of the steel grade code performance data, μ is the performance prediction result, and MSE is the mean square error of the model.

[0024] Optionally, the product performance requirement target value is used as a screening condition to determine the optimal process design of the target steel through a target database, including:

[0025] Determine a process design corresponding to the maximum value and the minimum value of the performance prediction result within the target value of the product performance requirement of the target steel;

[0026] Determine the PPK index corresponding to the process design;

[0027] The process design with the best PPK index is selected as the optimal process design for the target steel.

[0028] Optionally, the above method further includes:

[0029] Construct a vehicle parts database based on the dimensions of vehicle parts;

[0030] Determine the target value of product performance requirements based on the vehicle component database.

[0031] In a second aspect, an embodiment of the present invention further provides a steel process design device, comprising:

[0032] A first determination unit is used to determine target process parameters of the target steel during the production process;

[0033] an establishing unit, configured to establish a neural network model based on the target process parameters, wherein the neural network model is related to the room temperature tensile properties of the target steel;

[0034] The second determining unit is used to determine the optimal process design of the target steel material through the neural network model based on the product performance requirement target value.

[0035] In order to achieve the above-mentioned purpose, according to a third aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed by a processor, the steps of the above-mentioned steel process design method are implemented.

[0036] In order to achieve the above-mentioned purpose, according to the fourth aspect of the present invention, there is provided an electronic device, comprising at least one processor and at least one memory connected to the processor; wherein the above-mentioned processor is used to call the program instructions in the above-mentioned memory to execute the steps of the above-mentioned steel process design method.

[0037] Through the above-mentioned technical solution, the present invention provides a steel process design method and related equipment. This method addresses the problem of excessive reliance on manual trial and error in steel production process design. The present invention determines the target process parameters for the target steel during production; establishes a neural network model based on the target process parameters, wherein the neural network model is correlated with the room temperature tensile properties of the target steel; and uses the neural network model to determine the optimal process design for the target steel based on the target product performance requirements. In this solution, by determining the process parameters of the target steel and establishing a neural network model to predict its room temperature tensile properties, intelligent and data-driven process design is achieved. Traditional methods rely on manual trial and error, making it difficult to quantify the complex relationship between process parameters and performance. However, the neural network model can automatically learn nonlinear mapping laws from massive amounts of historical data, capturing the synergistic effects of multiple factors, such as steelmaking composition, hot rolling temperature, and cold rolling reduction, on material properties. The model directly predicts key indicators such as tensile strength and yield strength by inputting process parameters, forming a closed-loop "process-performance" mapping. When reverse-screening process parameters based on product performance requirements, the system matches candidate process solutions from the database that meet the target performance range, avoiding the problem of solution space redundancy caused by insufficient input dimensions in forward modeling. This approach shifts process design from an experience-driven to a data-driven approach, significantly improving the scientific and systematic nature of parameter optimization and reducing the potential for blind human intervention. Furthermore, through the model's generalization capabilities, it adapts to the needs of diverse production scenarios, providing a scalable technical framework for complex process design.

[0038] Correspondingly, the steel process design device, equipment and computer-readable storage medium provided by the embodiments of the present invention also have the above-mentioned technical effects.

[0039] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0041] Figure 1 A schematic flow chart of a steel material process design method provided by an embodiment of the present invention is shown;

[0042] Figure 2 A box line schematic diagram provided by an embodiment of the present invention is shown;

[0043] Figure 3 Another box line schematic diagram provided by an embodiment of the present invention is shown;

[0044] Figure 4 A schematic diagram of a correlation matrix provided by an embodiment of the present invention is shown;

[0045] Figure 5 A schematic diagram showing the logical relationship between process design recommendation and performance prediction provided by an embodiment of the present invention is shown;

[0046] Figure 6 A schematic diagram showing the logical relationship between a process design recommendation and product requirements provided by an embodiment of the present invention is shown;

[0047] Figure 7 A schematic diagram of a product performance requirement-oriented interactive interface provided by an embodiment of the present invention is shown;

[0048] Figure 8 A schematic diagram of a historical supply performance interaction interface provided by an embodiment of the present invention is shown;

[0049] Figure 9 A schematic diagram of a database import interaction interface provided by an embodiment of the present invention is shown;

[0050] Figure 10 A schematic diagram of an interactive interface for inputting a performance requirement range provided by an embodiment of the present invention is shown;

[0051] Figure 11 A schematic diagram of a background computing interaction interface provided by an embodiment of the present invention is shown;

[0052] Figure 12 A schematic diagram of an interactive interface for displaying recommendation results provided by an embodiment of the present invention is shown;

[0053] Figure 13 A schematic diagram of a performance verification interaction interface provided by an embodiment of the present invention is shown;

[0054] Figure 14 A schematic diagram of another performance verification interaction interface provided by an embodiment of the present invention is shown;

[0055] Figure 15 A schematic block diagram of the composition of a steel process design device provided by an embodiment of the present invention is shown;

[0056] Figure 16 A schematic block diagram of the composition of an electronic device for steel process design provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0057] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0058] In order to solve the problem that the process design for steel production relies too much on manual trial and error, an embodiment of the present invention provides a steel process design method, such as Figure 1 As shown, the method includes:

[0059] S101. Determine target process parameters for target steel during production;

[0060] For example, the target steel material may be IF steel for automobile plates or other steel materials, which are not specifically limited here. The following scheme takes hot-dip galvanized IF steel as an example.

[0061] Furthermore, Python, as a concise, easy-to-read and write programming language, has a wide range of applications in data processing and machine learning. It can help developers quickly build algorithm models, process large-scale data, and implement personalized recommendation functions. In the design and development of intelligent recommendation systems, Python libraries such as Pandas and NumPy can be used for data preprocessing; libraries such as Surprise and scikit-learn provide implementations of various recommendation algorithms, such as collaborative filtering and content recommendation; libraries such as LightFM support hybrid recommendation algorithms that can simultaneously consider user behavior data and item attribute information. This application can be implemented based on Python.

[0062] In one embodiment, room temperature tensile test data of a target steel material under different hot-dip galvanizing process conditions are obtained;

[0063] Preprocessing the room temperature tensile test data to obtain optimized room temperature tensile test data;

[0064] Obtaining the degree of correlation of the impact of the optimized room temperature tensile test data on key performance indicators;

[0065] The optimized room temperature tensile test data having an impact correlation degree greater than a preset standard is determined as the target process parameter.

[0066] For example, this application collects room temperature tensile test data of IF steel under different hot-dip galvanizing process conditions, organizes the data, filters missing values, zero values, outliers, removes duplicates, and other preprocessing to obtain optimized room temperature tensile test data.

[0067] This application analyzes the data through pivot tables, box plots (such as Figure 2 and Figure 3 As shown in Figure 2), determine the basic characteristics of the data and provide a data basis for subsequent model selection.

[0068] Furthermore, this application uses correlation analysis and heat maps to determine the degree of correlation between the production process parameters and key performance indicators such as tensile strength, yield strength, A80, R90, and N90. When building the prediction model, these target process parameters with high correlation are prioritized as input features.

[0069] Based on the above scheme, the present invention ensures the representativeness and reliability of the model training data by acquiring and preprocessing the tensile test data under the hot-dip galvanizing process. Data preprocessing includes filtering missing values, outliers and duplicate data to eliminate the interference of noise on model learning; analyzing the correlation between key process parameters and performance indicators through box plots and heat maps, screening out characteristic variables that have a significant impact on performance, and avoiding the risk of overfitting caused by the introduction of irrelevant variables into the model. Determining highly correlated parameters as target process parameters is essentially to extract core influencing factors based on the principle of statistical significance, so that the model focuses on the optimization of key process nodes (such as annealing temperature, coiling temperature, etc.). This step constructs a scientific feature engineering system from the source of the data, which not only reduces the complexity of the model, but also improves the prediction accuracy, provides high-quality training samples for subsequent neural network modeling, and ensures the pertinence and effectiveness of process parameter optimization.

[0070] In one embodiment, the method further includes:

[0071] The target process parameters are one-hot encoded and normalized to form numerical inputs.

[0072] Exemplarily, the present application performs one-hot encoding on the target process parameters. Due to the different equipment structures of the annealing furnace and the skin-finishing machine of the galvanizing production line, the process code needs to be used as a key input item for model training. Since the target process parameters are character-type data, there are two problems when they are used directly for model training: first, the model cannot directly understand non-numeric inputs; second, the similarity or difference between characters is opaque to the model, which may lead to the inability to correctly learn the intrinsic connection between the processes. Therefore, this application adopts the one-hot encoding method. The basic idea of ​​one-hot encoding is to convert each process code into a vector containing only 0 and 1, where the length of the vector is equal to the total number of samples.

[0073] Furthermore, the goal of normalization is mainly to adjust the numerical range of the data to a specific interval, usually [0, 1] or [-1, 1], but it can also be any other interval that is convenient for model processing. The goal of this process is to eliminate the dimensional differences and scale differences between different features so that different features can be treated equally during the model training process. After normalization, all features are on the same order of magnitude, which can accelerate model convergence, improve the prediction accuracy of the model, prevent gradient explosion or disappearance, and enhance the generalization ability of the model. The present invention uses the max_abs normalization method to process data. The max_abs normalization calculation formula is as follows:

[0074] x′=x / max

[0075] Wherein, x is the original data, x' is the standardized data, and max is the maximum value in the target process parameter data set.

[0076] This application uses the set_Sequential_model() function to construct a simple three-layer neural network: one input layer, one hidden layer, and one output layer, for a regression problem. The model is configured to learn and predict continuous values ​​from input data using the ReLU activation function, the mean squared error loss function, and the Adam optimization algorithm. The dataset is split into training and test sets by importing the train_test_split function from the sklearn.model_selection module. The model is trained using the yield_model.fit method. The trained model is saved using yield_model.save('yield_model.keras') .

[0077] It's important to note that training deep learning models typically requires significant computing resources and time. Especially for large models or datasets, the training process can take days or even weeks. Saving a trained model avoids having to repeat the training process when the same model is needed in the future, saving computing resources and time.

[0078] Based on the above scheme, the negative impact of non-numerical data and dimensional differences on model training is resolved by performing one-hot encoding and normalization on the process parameters. Discrete variables such as process codes are converted into binary vectors through one-hot encoding, enabling the model to recognize the independence and differences between different processes. For example, the process differences between the annealing furnace and the skin-finishing machine are encoded as features of different dimensions, avoiding the false order relationship that may be introduced by traditional numerical encoding. Normalization unifies the numerical range of each process parameter to the same scale (such as \[0,1\]), eliminating the feature weight deviation caused by different dimensions. For example, the cold rolling reduction rate (percentage) and the annealing temperature (degrees Celsius) have a large numerical difference in the original scale. After normalization, the model can treat each feature equally. These two preprocessing technologies optimize the model input structure from the data expression level, enhance the compatibility of neural networks with multi-source heterogeneous data, improve the model convergence speed and generalization performance, and lay the foundation for the collaborative optimization of complex process parameters.

[0079] S102, establishing a neural network model based on the target process parameters, wherein the neural network model is related to the room temperature tensile properties of the target steel;

[0080] For example, considering that ensuring data quality is the basis for establishing an effective prediction model, this application uses data cleaning and exploratory data analysis. Based on the big data foundation, a suitable neural network model is selected to predict the room temperature tensile properties of hot-dip galvanized IF steel.

[0081] Specifically,

[0082] It is important to note that the selection of a model must take into account factors such as data characteristics and prediction accuracy requirements. The model training process also includes key steps such as parameter tuning and overfitting control.

[0083] S103. Determine the optimal process design of the target steel material through the neural network model based on the target value of product performance requirements.

[0084] For example, the above-mentioned operation of establishing a performance prediction model involves finding correlation patterns between data through data cleaning and exploratory data analysis. Based on big data, a neural network model for predicting the room temperature tensile properties of hot-dip galvanized IF steel is established.

[0085] Based on this application, the system model can be used to simulate and predict the process design results, provide early warning of the risk of unqualified performance, improve the efficiency of data analysis, reduce the error rate in the process design process, thereby reducing the production cost of IF steel and improving the economic benefits of the enterprise.

[0086] In one embodiment, a target database is established based on the neural network model, wherein the target database is used to provide feedback on the relationship between the target process parameters and the room temperature tensile properties;

[0087] The target value of the product performance requirement is used as a screening condition to determine the optimal process design of the target steel through the target database.

[0088] Specifically, such as Figure 5 As shown in the figure, the idea of ​​process design recommendation is equivalent to the reverse of the performance prediction model. Given the target value and upper and lower limits of the performance requirements, the target value and upper and lower limits of the process parameters of the entire process are obtained.

[0089] This application takes into account that if the performance requirement target value is used as the input variable X and the process parameter target value of the entire process is used as the output variable Y, the dimension of X will be much smaller than the dimension of Y. This means that for the same X value (i.e., the performance requirement target value), there may be multiple Y values ​​(i.e., the process parameter target value of the entire process) that meet the conditions, making it difficult to ensure that the final process parameter target value is both accurate and reliable.

[0090] like Figure 6 As shown, this application establishes a process design-material performance database through a predictive model, and then uses the product performance requirements as a screening condition to obtain a process design that meets the requirements. This avoids the problem of the X dimension being smaller than the Y dimension.

[0091] By leveraging the aforementioned technical solution, the present application establishes a database of process design and material performance results, and utilizes the aforementioned neural network model to determine the performance results corresponding to the existing quality designs in the manufacturing system, thereby forming a comprehensive database of quality design and material performance results.

[0092] The above solution constructs a target database based on a neural network model and establishes a dynamic mapping relationship library between process parameters and material properties. The database associates and stores historical process data with predicted performance results, forming a queryable process-performance knowledge graph. When product performance requirements are input, the system reversely searches the database for process solutions that meet the requirements, replacing traditional trial-and-error experiments. For example, when the user specifies A80 ≥ 30%, the database automatically filters out all process parameter combinations that predict A80 to meet the standard and recommends them based on process stability. This mechanism essentially achieves process reverse design through data-driven pattern matching, solving the problem of low search efficiency caused by the explosion of process parameter combinations in traditional methods. Continuous updates to the database can also accumulate production performance data, forming a self-reinforcing learning loop, so that the recommendation system is continuously optimized as the amount of data grows, gradually approaching the global optimal solution.

[0093] In one embodiment, the method further includes:

[0094] The performance prediction result of the process design of the target steel is determined based on the following formula taking into account the production process control error and the prediction model error:

[0095] Maximum performance prediction result of process design: μ+3σ+MSE

[0096] Average value of performance prediction results of process design: μ

[0097] Minimum performance prediction result of process design: μ-3σ-MSE

[0098] Wherein, σ is the standard deviation of the steel grade code performance data, μ is the performance prediction result, and MSE is the mean square error of the model.

[0099] Based on this approach, an error compensation algorithm is introduced to expand the performance prediction range, significantly improving the industrial applicability of the recommended results. The maximum and minimum performance predictions for the process design are corrected by superimposing the model's mean square error (MSE) and production fluctuations (3σ), expanding the theoretical predictions to a practically possible range. For example, when the model predicts a yield strength of 300 MPa, considering MSE = 5 and σ = 2, the actual value is likely to be distributed within the range of 300 ± (3 × 2 + √5) ≈ 300 ± 7.2 MPa. This algorithm probabilistically quantifies the impact of production uncertainty, ensuring that the parameter tolerance band of the recommended process covers the actual fluctuation range. During secondary screening, the system only retains process solutions whose expanded range fully falls within the target performance requirements, avoiding the risk of performance overshoot due to model errors or equipment fluctuations. This mechanism integrates statistical principles into process design and enhances the adaptability of the recommendation system to complex production environments.

[0100] In one embodiment, the method of using the product performance requirement target value as a screening condition to determine the optimal process design of the target steel material through a target database includes:

[0101] Determine a process design corresponding to the maximum value and the minimum value of the performance prediction result within the target value of the product performance requirement of the target steel;

[0102] Determine the PPK index corresponding to the process design;

[0103] The process design with the best PPK index is selected as the optimal process design for the target steel.

[0104] Illustratively, this application recommends the optimal process design based on the performance requirements of the new product, and the performance results are within the performance requirements of the new product; and the indicator characterizing the performance control capability - PPK is arranged in descending order from highest to lowest, and the top five are taken as the final recommendation results, thereby determining the optimal process design for the target steel.

[0105] Specifically:

[0106]

[0107] σ is the standard deviation of the performance data of the steel grade code, μ is the performance prediction result, USL is the upper limit of the performance requirement, and LSL is the lower limit of the performance requirement.

[0108] This application will use the process design whose performance prediction results (maximum value, minimum value) are within the product performance requirement target value range of the new product, and arrange the screening results in descending order from highest to lowest according to the indicator characterizing the control capability - PPK, as the optimal process design for the target steel.

[0109] Based on the above scheme, this application uses the process capability index (PPK) as the basis for sorting process schemes, and gives priority to recommending process parameters with the best stability and controllability. PPK quantifies the probability that the process meets performance requirements by calculating the relative position of the mean (μ) of the process output and the upper and lower specification limits (USL / LSL). For example, PPK = 1.33 means that the process fluctuation is within 75% of the specification range, while PPK = 2.0 corresponds to a pass rate of 99.7%. The present invention selects the top five processes by PPK sorting, which is essentially to select the solution with the largest tolerance margin and the strongest anti-interference ability from the perspective of statistical process control (SPC). This strategy reduces the risk of performance deviation caused by equipment fluctuations or differences in raw material batches during the production process, ensures the robustness of the recommended process in actual production, and provides a quantitative evaluation benchmark for the continuous optimization of process parameters.

[0110] In one embodiment, the method further includes:

[0111] Construct a vehicle parts database based on the dimensions of vehicle parts;

[0112] Determine the target value of product performance requirements based on the vehicle component database.

[0113] Exemplarily, this application constructs a forward-looking guidance system for product process design driven by a vehicle model parts knowledge base. Specifically, this application implants the dimensions of vehicle model parts into the product process design ideas. To achieve system-guided products, with vehicle model parts as the entry point, the three knowledge bases of product characteristics, three major issues and one major issue, objections and complaints, and historical supply performance of vehicle model parts are connected in series as advance guidance for product process design. Incorporating the dimensions of vehicle model parts into the product process design process provides forward-looking guidance for product process design, realizing the integration of vehicle model part dimensions with product process design.

[0114] Specifically, this application first establishes a database: organizing the names, images, and user needs of typical automotive parts. The images are stored in a local computer folder in PNG format, and text and other information are stored in an Excel spreadsheet. An information query window is created. Using the Python graphical user interface (GUI) module PySimpleGUI, the database is organized into tabs to display the product features, major issues, objections, and complaints, as well as historical supply performance of the vehicle model parts, using the part name as a clue.

[0115] It should be noted that the above three major issues are used to represent the key problems that occur in key products at key users, thus forming a major comparative advantage.

[0116] Furthermore, this application combines the protocol or standard with the vehicle component knowledge base to ultimately determine the performance range of the product. Entering this range into the "Process Design Recommendation" interface, the system automatically recommends process design results that meet this performance range.

[0117] Finally, this application selects process design results that meet the product performance requirements. At the same time, a performance verification mechanism for the recommended results is established, ensuring the accuracy and reliability of the recommended results by comparing historical performance results with product performance requirements. This application takes into account the errors in the prediction model and fluctuations in the production control process. The recommended process design will be formally adopted after verifying that historical performance results are consistent with product performance requirements.

[0118] Based on the above solution, by building a vehicle parts database and binding it to product performance requirements, demand-oriented optimization of process design is achieved. The database integrates the application scenarios of parts (such as the need for high formability of door panels), user complaints (such as surface defects) and historical supply performance data to form multi-dimensional constraints. For example, for a certain model's roof outer panel, the system automatically associates its "high surface quality" and "anti-sag" requirements, and prioritizes the solution with the best surface finish prediction value in the annealing process parameter combination. This design reversely injects downstream application requirements into the process design process, breaking through the limitations of traditional process optimization that only focuses on material performance indicators, so that process parameters can meet both performance requirements and the special needs of actual application scenarios, thereby improving the systematicness and market responsiveness of product development.

[0119] The steel process design method provided by the present invention addresses the problem of excessive reliance on manual trial and error in steel production process design. The method determines target process parameters for a target steel during production; establishes a neural network model based on these target process parameters, wherein the neural network model is correlated with the room-temperature tensile properties of the target steel; and uses the neural network model to determine the optimal process design for the target steel based on the target product performance requirements. In this approach, by determining the process parameters for the target steel and establishing a neural network model to predict its room-temperature tensile properties, intelligent and data-driven process design is achieved. Traditional methods rely on manual trial and error, making it difficult to quantify the complex relationship between process parameters and performance. However, the neural network model can automatically learn nonlinear mapping patterns from massive amounts of historical data, capturing the synergistic effects of multiple factors, such as steelmaking composition, hot rolling temperature, and cold rolling reduction, on material properties. The model directly predicts key indicators, such as tensile strength and yield strength, by inputting process parameters, forming a closed-loop "process-performance" mapping. When reversely screening process parameters based on product performance requirements, the system matches candidate process solutions from a database that meet the target performance range, avoiding the problem of solution space redundancy caused by insufficient input dimensions in forward modeling. This method shifts process design from experience-driven to data-driven, significantly improving the scientific nature and systematic nature of parameter optimization and reducing the blindness of manual intervention. At the same time, it adapts to the needs of different production scenarios through model generalization capabilities, providing a scalable technical framework for complex process design.

[0120] The following is a specific embodiment of the present application:

[0121] The information of new product parts to be designed is known. Through the parts information database, the product characteristics, three major issues and one major issue, as well as objection complaints of vehicle parts can be obtained. Figure 7 ), historical supply performance ( Figure 8 ). Given the performance requirements of a new product to be designed, the process design recommendation system can be used to obtain the optimal process design recommendation. The steps are as follows:

[0122] Step S1: Import database: Click the button "Import database", wait for about 30 seconds, the status description bar will display: "Database import completed" ( Figure 9 ).

[0123] Step S2: Enter the performance requirement range: In the performance requirement input module, enter the upper and lower limits of the performance requirements and click the "Confirm" button ( Figure 10 ).

[0124] Step S3: Background Calculation: Click the "Start Calculation" button. The background program will begin selecting the optimal process. The calculation process ends when the status bar displays "Calculation Completed."

[0125] ( Figure 11 )

[0126] Step S4: Recommendation result display: Click the "Display Recommendation Result" button. The window will show the recommended five optimal process design recommendations, including performance results, steelmaking, hot rolling, cold rolling and annealing processes. ( Figure 12 )

[0127] Step S5: Performance Verification: Click the "Performance Verification" button. A new window will pop up to display the box plot of historical performance results under the recommended process parameters. ( Figure 13 and Figure 14 )

[0128] Furthermore, as a response to the above Figure 1 In order to realize the method shown in the figure, the embodiment of the present invention also provides a steel process design device for Figure 1 This device embodiment corresponds to the aforementioned method embodiment. For ease of reading, this device embodiment will not describe the details of the aforementioned method embodiment one by one, but it should be clear that the device in this embodiment can implement all the contents of the aforementioned method embodiment. Figure 15 As shown, the device includes: a first determining unit 21, an establishing unit 22, and a second determining unit 23, wherein

[0129] The first determining unit 21 is used to determine target process parameters of the target steel material during the production process;

[0130] an establishing unit 22 for establishing a neural network model based on the target process parameters, wherein the neural network model is related to the room temperature tensile properties of the target steel;

[0131] The second determining unit 23 is configured to determine the optimal process design of the target steel material through the neural network model based on the target value of product performance requirements.

[0132] The processor includes a kernel, which retrieves the corresponding program unit from memory. One or more kernels can be configured, and by adjusting kernel parameters, a steel process design method can be implemented, addressing the problem of steel production process design being overly reliant on manual trial and error.

[0133] An embodiment of the present invention provides a computer-readable storage medium, which includes a stored program. When the program is executed by a processor, the steel process design method is implemented.

[0134] An embodiment of the present invention provides a processor, which is used to run a program, wherein the steel process design method is executed when the program is run.

[0135] An embodiment of the present invention provides an electronic device, comprising at least one processor and at least one memory connected to the processor; wherein the processor is configured to call program instructions in the memory to execute the steel process design method as described above.

[0136] An embodiment of the present invention provides an electronic device 30, such as Figure 16 As shown, the electronic device includes at least one processor 301, and at least one memory 302 and a bus 303 connected to the processor; wherein the processor 301 and the memory 302 communicate with each other through the bus 303; the processor 301 is used to call the program instructions in the memory to execute the above-mentioned steel process design method.

[0137] The intelligent electronic devices in this article can be PCs, PADs, mobile phones, etc.

[0138] The present application also provides a computer program product, which, when executed on a process management electronic device, is suitable for executing a program that initializes the above-mentioned steel process design method steps.

[0139] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0140] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0141] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0142] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0143] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0144] The present application also provides a computer program product, which includes computer software instructions. When the computer software instructions are executed on a processing device, the processing device is caused to execute the following Figure 1 This corresponds to the flow of memory control in the embodiment.

[0145] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium can be any available medium that a computer can store or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state drive (SSD)).

[0146] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0147] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0148] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0149] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0150] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0151] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A steel process design method, characterized in that: include: Determine the target process parameters for the target steel during production; Establishing a neural network model based on the target process parameters, wherein the neural network model is related to the room temperature tensile properties of the target steel; The optimal process design of the target steel is determined through the neural network model based on the target value of product performance requirements.

2. The method according to claim 1, characterized in that Determining target process parameters of the target steel during production includes: Obtain room temperature tensile test data of target steel under different hot-dip galvanizing process conditions; Preprocessing the room temperature tensile test data to obtain optimized room temperature tensile test data; Obtaining the degree of correlation of the impact of the optimized room temperature tensile test data on key performance indicators; The optimized room temperature tensile test data having an impact correlation degree greater than a preset standard is determined as the target process parameter.

3. The method according to claim 1, characterized in that Also includes: The target process parameters are one-hot encoded and normalized to form numerical inputs.

4. The method according to claim 1, wherein The determining the optimal process design of the target steel material by the neural network model based on the product performance requirement target value includes: Establishing a target database based on the neural network model, wherein the target database is used to provide feedback on the relationship between the target process parameters and the room temperature tensile properties; The target value of the product performance requirement is used as a screening condition to determine the optimal process design of the target steel through the target database.

5. The method according to claim 4, characterized in that Also includes: The performance prediction result of the process design of the target steel is determined based on the following formula taking into account the production process control error and the prediction model error: Maximum performance prediction result of process design: μ+3σ+MSE Average value of performance prediction results of process design: μ Minimum performance prediction result of process design: μ-3σ-MSE Wherein, σ is the standard deviation of the steel grade code performance data, μ is the performance prediction result, and MSE is the mean square error of the model.

6. The method according to claim 5, characterized in that The method of using the target value of the product performance requirement as a screening condition to determine the optimal process design of the target steel through a target database includes: Determine a process design corresponding to the maximum value and the minimum value of the performance prediction result within the target value of the product performance requirement of the target steel; Determine the PPK index corresponding to the process design; The process design with the best PPK index is selected as the optimal process design for the target steel.

7. The method according to claim 1, characterized in that Also includes: Construct a vehicle parts database based on the dimensions of vehicle parts; Determine the target value of product performance requirements based on the vehicle component database.

8. A steel process design device, characterized in that: Also includes: A first determination unit is used to determine target process parameters of the target steel during the production process; an establishing unit, configured to establish a neural network model based on the target process parameters, wherein the neural network model is related to the room temperature tensile properties of the target steel; The second determining unit is used to determine the optimal process design of the target steel material through the neural network model based on the product performance requirement target value.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored program, wherein when the program is executed by a processor, the steps of the steel process design method according to any one of claims 1 to 7 are implemented.

10. An electronic device, characterized in that: The electronic device includes at least one processor and at least one memory connected to the processor; wherein the processor is used to call program instructions in the memory to execute the steps of the steel process design method according to any one of claims 1 to 7.

Images