Steel grade design method and system, electronic equipment and storage medium

Abstract

The invention relates to the technical field of steel design, in particular to a steel grade design method and system, electronic equipment and a storage medium, and the method comprises the steps that target values of mechanical property indexes of to-be-designed steel grades and initial values of target process parameters of the to-be-designed steel grades are obtained; inputting the initial value of the target process parameter into a pre-trained mechanical property prediction model, and performing mechanical property prediction to obtain a prediction result of a mechanical property index; and performing iterative optimization on the initial value of the target process parameter according to a preset target function, the prediction result and the target value of the mechanical property index so as to obtain a final value of the target process parameter and a final mechanical property value corresponding to the final value, the mechanical property final value refers to a prediction result obtained by inputting the final value of the target process parameter into the mechanical property prediction model; according to the method, the steel grade design efficiency is high, the steel grade design period is greatly shortened, the test cost is reduced, and the precision is high.

Description

Technical Field

[0001] The present invention relates to the technical field of steel design, and in particular to a steel grade design method, system, electronic equipment and storage medium. Background Art

[0002] Steel has become the material basis of modern society. However, with the diversification of steel application scenarios, existing steel grades can no longer meet the needs of some application scenarios. More and more customers have put forward new requirements for the performance and characteristics of steel. This has also made some steel companies face the current situation of small-batch and multi-variety steel production, which has brought great challenges to steel companies. Steel companies urgently need to respond to this challenge by continuously optimizing steel grade design and process.

[0003] The traditional steel grade design method mainly relies on the experimental method of "trial and error" and "experience", that is, through preliminary design, parameter adjustment and repeated experiments, the steel grade design is completed. However, this method has a long design cycle, high experimental costs, and is inconvenient to implement. Summary of the invention

[0004] The present invention provides a steel grade design method, system, electronic equipment and storage medium to solve the problems of long design cycle, high test cost and inconvenient implementation of the steel grade design method in the prior art.

[0005] The present invention provides a steel grade design method, the method comprising: obtaining target values ​​of mechanical property indicators of the steel grade to be designed, and initial values ​​of target process parameters of the steel grade to be designed;

[0006] Inputting the initial values ​​of the target process parameters into a pre-trained mechanical property prediction model to perform mechanical property prediction to obtain prediction results of mechanical property indicators;

[0007] According to the preset objective function, the prediction result, and the target value of the mechanical property index, the initial value of the target process parameter is iteratively optimized to obtain the final value of the target process parameter and the final value of the mechanical property corresponding to the final value. The final value of the mechanical property refers to the prediction result obtained by inputting the final value of the target process parameter into the mechanical property prediction model.

[0008] In one embodiment of the present invention, the target value of the mechanical property index of the steel grade to be designed is a preset value, and the step of obtaining the initial value of the target process parameter of the steel grade to be designed includes:

[0009] Acquire historical production data of a plurality of associated steel grades, wherein the associated steel grades refer to steel grades of the same type as the steel grade to be designed, and the historical production data include: historical values ​​of target process parameters, and performance test results;

[0010] Determine the associated steel grade corresponding to the performance test result with the smallest difference from the target value as the basic steel grade;

[0011] All historical values ​​of the target process parameters of the basic steel grade are averaged to obtain initial values ​​of the target process parameters of the steel grade to be designed.

[0012] In one embodiment of the present invention, the training step of the mechanical property prediction model includes:

[0013] Acquire historical production data of a plurality of associated steel grades, wherein the associated steel grades refer to steel grades of the same type as the steel grade to be designed, and the historical production data include: historical values ​​of target process parameters, and performance test results;

[0014] Preprocessing the historical production data to obtain preprocessed data, wherein preprocessing the historical production data includes removing null values, abnormal values, and outliers;

[0015] Performing feature screening on historical values ​​of target process parameters in the preprocessed data to obtain data to be predicted, wherein the feature screening includes variance screening, collinearity screening and contribution screening;

[0016] Inputting the data to be predicted into the mechanical property prediction model to perform mechanical property prediction to obtain mechanical property prediction results;

[0017] The mechanical property prediction model is trained according to the gap between the mechanical property prediction result and the corresponding performance test result in the preprocessed data.

[0018] In one embodiment of the present invention, the target process parameters include preset common process parameters and key process parameters, and the prediction results include predicted values ​​of multiple mechanical property indicators of the steel grade to be designed;

[0019] The step of iteratively optimizing the initial value of the target process parameter according to the preset objective function, the prediction result, and the target value of the mechanical property index to obtain the final value of the target process parameter and the final value of the mechanical property corresponding to the final value includes:

[0020] According to the preset objective function, the prediction results, and the target values ​​of the mechanical performance indicators, the initial values ​​of the key process parameters are iteratively optimized, and the values ​​of the common process parameters are kept unchanged to obtain the final values ​​of all the target process parameters and the final values ​​of the mechanical properties corresponding to the final values.

[0021] In one embodiment of the present invention, the step of iteratively optimizing the initial values ​​of the key process parameters according to the preset objective function, the prediction results, and the target values ​​of the mechanical performance indicators includes:

[0022] When there is a gap between the prediction result and the target value of the mechanical property index, and the current cumulative number of iterations is less than a preset iteration number threshold, obtaining the initial value of the key process parameter, wherein the cumulative number of iterations refers to the number of iterations for obtaining the value of the target process parameter;

[0023] Determine the key process parameter as a fruit fly, and determine the initial position of each fruit fly in a preset spatial coordinate system based on the initial value of the key process parameter;

[0024] Perform multiple random flights on each fruit fly to obtain multiple sets of fruit fly position information, wherein the fruit fly position information includes the target position after each fruit fly has randomly flown, wherein each random flight controls the fruit fly to randomly fly a preset distance from an initial position;

[0025] Obtaining a target distance from the target position to the origin of the spatial coordinate system, and determining the reciprocal of the target distance as a new value of the current key process parameter;

[0026] Inputting the new values ​​of all the key process parameters and the initial values ​​of the common process parameters into the mechanical property prediction model to perform mechanical property prediction and obtain new prediction results;

[0027] According to the new prediction results, the objective function, and the target values ​​of the mechanical performance indicators, iterative optimization of the initial values ​​of the key process parameters is completed.

[0028] In one embodiment of the present invention, the step of completing iterative optimization of the initial value of the key process parameter according to the new prediction result, the objective function, and the target value of the mechanical performance index includes:

[0029] According to the new prediction result, the objective function, and the target value of the mechanical performance index, the flavor concentration value corresponding to each group of fruit fly position information is obtained;

[0030] Determine the maximum value among multiple flavor concentration values, determine the fruit fly position information corresponding to the maximum value as the initial position of the next generation fruit fly population, and repeat the steps of random flight, key process parameter value update, mechanical property prediction, flavor concentration value acquisition and screening based on the initial position of the next generation fruit fly population until the cumulative number of iterations reaches the iteration number threshold;

[0031] If the cumulative number of iterations reaches the iteration number threshold, the current value of the target process parameter is determined as the final value of the target process parameter, and the prediction result of the mechanical property index corresponding to the final value of the target process parameter is determined as the final value of the mechanical property.

[0032] In one embodiment of the present invention, according to the new prediction result, the objective function, and the target value of the mechanical performance index, the mathematical expression of the flavor concentration value corresponding to each group of fruit fly position information is obtained as follows:

[0033]

[0034] Among them, F(α, β, γ…) represents the flavor concentration value, (α, β, γ…) represents the values ​​of multiple key process parameters, min represents minimization, K represents the number of mechanical performance indicators, and w j represents the important coefficient of the preset j-th mechanical performance index, x represents the common process parameters in the target process parameters, y represents the key process parameters in the target process parameters, and f j (x, y) represents the prediction result predicted by the mechanical property prediction model according to the input values ​​of x and y, and represents the target value of the jth mechanical property index.

[0035] The present invention also provides a steel grade design system, the system comprising: an initial data acquisition module, used to acquire target values ​​of mechanical property indicators of the steel grade to be designed, and initial values ​​of target process parameters of the steel grade to be designed;

[0036] A mechanical property prediction module is used to input the initial value of the target process parameter into a pre-trained mechanical property prediction model to perform mechanical property prediction and obtain a prediction result of the mechanical property index;

[0037] The steel grade design module is used to iteratively optimize the initial values ​​of the target process parameters according to a preset objective function, the prediction results, and the target values ​​of the mechanical property indicators to obtain the final values ​​of the target process parameters and the final values ​​of the mechanical properties corresponding to the final values, wherein the final values ​​of the mechanical properties refer to the prediction results obtained by inputting the final values ​​of the target process parameters into the mechanical property prediction model.

[0038] The present invention also provides an electronic device, comprising a processor, a memory and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute a computer program stored in the memory to implement a method provided in any one of the above embodiments.

[0039] The present invention also provides a computer-readable storage medium on which a computer program is stored, wherein the computer program is used to enable a computer to execute the method provided in any one of the above embodiments.

[0040] Beneficial effects of the present invention: The steel grade design method, system, electronic device and storage medium proposed in the present invention obtain target values ​​of mechanical property indicators of the steel grade to be designed and initial values ​​of target process parameters of the steel grade to be designed; input the initial values ​​of the target process parameters into a pre-trained mechanical property prediction model to perform mechanical property prediction and obtain prediction results of the mechanical property indicators; iteratively optimize the initial values ​​of the target process parameters according to the preset objective function, the prediction results, and the target values ​​of the mechanical property indicators to obtain final values ​​of the target process parameters and final values ​​of mechanical properties corresponding to the final values, wherein the final values ​​of mechanical properties refer to prediction results obtained by inputting the final values ​​of the target process parameters into the mechanical property prediction model; the method can quickly obtain the steel grade design results, i.e., the final values ​​of the target process parameters of the steel grade to be designed and the final values ​​of mechanical properties corresponding to the final values, has high design efficiency, greatly shortens the steel grade design cycle, reduces the test cost, has a high degree of automation, and iteratively optimizes the initial values ​​of the target process parameters so that the mechanical property indicators corresponding to the optimized values ​​of the target process parameters gradually approach the corresponding target values, thereby effectively improving the design accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic flow chart of a steel grade design method provided in one embodiment of the present invention;

[0042] Figure 2 It is a schematic diagram of the optimization of the tensile strength obtained by adopting the steel grade design method in one embodiment of the present invention;

[0043] Figure 3 It is a schematic diagram of the optimization of the yield strength designed by the steel grade design method in one embodiment of the present invention;

[0044] Figure 4 It is a schematic diagram of the optimization of the elongation obtained by adopting the steel grade design method in one embodiment of the present invention;

[0045] Figure 5 A schematic diagram of the structure of a steel grade design system provided by an embodiment of the present invention;

[0046] Figure 6 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0048] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0049] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it is obvious to those skilled in the art that the embodiments of the present invention can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present invention difficult to understand.

[0050] Combine the following Figures 1 to 6 , the steel grade design method, system, electronic device and storage medium provided by the present invention are explained.

[0051] See also Figure 1 , Figure 1 A schematic diagram of a process flow of a steel grade design method provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the method includes:

[0052] S110: Obtain target values ​​of mechanical property indicators of the steel type to be designed, and initial values ​​of target process parameters of the steel type to be designed.

[0053] It should be noted that the target values ​​of the mechanical properties of the steel to be designed are preset values. By presetting the target values ​​of the mechanical properties of the steel to be designed, it is possible to facilitate the subsequent steel design. The mechanical properties include yield strength, tensile strength, elongation, and impact energy. The target process parameters of the steel to be designed include copper content, phosphorus content, final rolling temperature, coiling temperature, niobium content, and manganese content. The initial values ​​of the parameters of the target process can be obtained by presetting, or by obtaining them based on historical process parameters of the same type of steel in the following embodiment.

[0054] In some embodiments, the initial value of the target process parameter is determined based on the historical process parameters of a base steel grade, and the base steel grade and the steel grade to be designed are of the same type.

[0055] S120: Inputting the initial values ​​of the target process parameters into a pre-trained mechanical property prediction model to perform mechanical property prediction to obtain prediction results of mechanical property indicators.

[0056] It should be noted that the prediction results of mechanical performance indicators include the predicted values ​​of various mechanical performance indicators. By using the mechanical performance prediction model to predict mechanical performance, it is possible to adjust the initial value or value of the target process parameter based on the prediction results output by the mechanical performance prediction model, so that the prediction results of the mechanical performance indicators corresponding to the adjusted values ​​of the target process parameters gradually approach the target value.

[0057] S130: According to a preset objective function, the prediction result, and the target value of the mechanical property index, the initial value of the target process parameter is iteratively optimized to obtain the final value of the target process parameter and the final value of the mechanical property corresponding to the final value, wherein the final value of the mechanical property refers to the prediction result obtained by inputting the final value of the target process parameter into the mechanical property prediction model.

[0058] It should be noted that, by iteratively optimizing the initial values ​​of the target process parameters according to the objective function, the prediction results, and the target values ​​of the mechanical properties, the prediction results of the mechanical properties corresponding to the values ​​of the target process parameters can gradually approach the corresponding target values, thereby completing the design of the steel type to be designed, that is, obtaining the final values ​​of the target process parameters of the steel type to be designed, and the final values ​​of the mechanical properties corresponding to the final values, with high design efficiency, low cost, high accuracy, and strong real-time performance.

[0059] In some embodiments, the target value of the mechanical property index of the steel type to be designed is a preset value, and the step of obtaining the initial value of the target process parameter of the steel type to be designed includes:

[0060] 1. Obtain historical production data of multiple related steel grades, where the related steel grades refer to steel grades of the same type as the steel grade to be designed, and the historical production data include: historical values ​​of target process parameters, and performance test results.

[0061] It should be noted that the performance test results refer to the values ​​of the mechanical performance indicators of the relevant steel types during the historical production process.

[0062] 2. Determine the associated steel grade corresponding to the performance test result with the smallest difference from the target value as the basic steel grade.

[0063] 3. All historical values ​​of the target process parameters of the basic steel grade are averaged to obtain initial values ​​of the target process parameters of the steel grade to be designed.

[0064] It should be noted that by adopting the above method to determine the initial values ​​of the target process parameters of the steel grade to be designed, the subsequent design difficulty can be effectively reduced and the steel grade design efficiency and response speed can be improved.

[0065] In some embodiments, the training step of the mechanical property prediction model includes:

[0066] 1. Obtain historical production data of multiple related steel grades, where the related steel grades refer to steel grades of the same type as the steel grade to be designed, and the historical production data include: historical values ​​of target process parameters, and performance test results.

[0067] 2. Preprocessing the historical production data to obtain preprocessed data, wherein preprocessing the historical production data includes removing null values, abnormal values, and outliers.

[0068] It should be noted that the outlier refers to the extreme value in the data, that is, the value that is significantly different from other data. For example, if the difference between a certain data value and the target mean is greater than or equal to the preset difference threshold, the data value can be determined to be an outlier, etc. The target mean refers to the mean of all data. Outliers refer to values ​​that significantly deviate from the preset normal data range, and the normal data range can be set according to actual conditions. The above preprocessing can help improve the accuracy of the mechanical properties prediction model.

[0069] 3. Perform feature screening on the historical values ​​of the target process parameters in the preprocessed data to obtain the data to be predicted, wherein the feature screening includes variance screening, collinearity screening and contribution screening.

[0070] It should be noted that variance screening refers to reducing the features that contribute less to model prediction by excluding features with too small variance; collinearity screening refers to reducing the multicollinearity problem between features by identifying and removing highly correlated features; contribution screening refers to retaining the most important features for model effect by evaluating the contribution of each feature to model prediction performance. The above steps can extract as much useful feature information as possible, i.e., the data to be predicted, by performing variance screening, collinearity screening, and contribution screening in sequence.

[0071] 4. Inputting the data to be predicted into the mechanical property prediction model, performing mechanical property prediction, and obtaining mechanical property prediction results.

[0072] 5. The mechanical property prediction model is trained according to the gap between the mechanical property prediction result and the corresponding performance test result in the preprocessed data.

[0073] It should be noted that by adopting the above-mentioned method to train the mechanical property prediction model, the accuracy of the mechanical property prediction model can be effectively improved, and the implementation is relatively convenient. In addition, by using the historical production data of the same type of steel as the steel to be designed for model training, the trained mechanical property prediction model can have a relatively accurate mechanical property prediction capability for this type of steel. Compared with the method of casting a wide net (such as obtaining historical production data of multiple types of steel for model training), this method can not only improve the mechanical property prediction accuracy of the mechanical property prediction model for the type of steel to be designed, but also effectively reduce the difficulty of model training and improve the efficiency of model training.

[0074] In order to further reduce the difficulty and complexity of steel grade design, in some embodiments, the target process parameters include: preset common process parameters and key process parameters. Common process parameters refer to parameters that have little impact on the mechanical properties of the steel to be designed, and key process parameters refer to parameters that have a greater impact on the mechanical properties of the steel to be designed. The common process parameters and key process parameters can be set and adjusted according to actual conditions. For example: copper content and phosphorus content are set as common process parameters, and finishing temperature, coiling temperature, niobium content, and manganese content are set as key process parameters. Through the above settings, it is convenient to iteratively optimize only the key process parameters during the subsequent iterative optimization of process parameters, so as to gradually approach the target value of the mechanical properties. The common process parameters remain unchanged to reduce the difficulty and complexity of iterative optimization, improve design efficiency, and shorten response time.

[0075] In some embodiments, the prediction result includes predicted values ​​of multiple mechanical property indicators of the steel type to be designed.

[0076] In some embodiments, the step of iteratively optimizing the initial value of the target process parameter according to the preset objective function, the prediction result, and the target value of the mechanical property index to obtain the final value of the target process parameter and the final value of the mechanical property corresponding to the final value includes:

[0077] According to the preset objective function, the prediction results, and the target values ​​of the mechanical performance indicators, the initial values ​​of the key process parameters are iteratively optimized, and the values ​​of the common process parameters are kept unchanged to obtain the final values ​​of all the target process parameters and the final values ​​of the mechanical properties corresponding to the final values.

[0078] It should be noted that the objective function is to minimize the gap between the predicted result and the corresponding target value. By adopting the above method, i.e. iteratively optimizing the initial values ​​of the key process parameters and keeping the values ​​of the common process parameters unchanged, the complexity of the optimization can be effectively reduced.

[0079] It should be mentioned that in the process of iterative optimization of the initial values ​​or values ​​of key process parameters, it is necessary to ensure that the common process parameters are within a preset reasonable range, which can be set according to actual conditions.

[0080] The above steps of iteratively optimizing the initial values ​​of the key process parameters according to the preset objective function, the prediction results, and the target values ​​of the mechanical performance indicators are equivalent to a multi-objective optimization model, that is, optimizing multiple key process parameters to minimize the gap between the predicted values ​​and the target values ​​of the mechanical performance indicators. The following is an explanation of the process of solving the multi-objective optimization model.

[0081] In some embodiments, the step of iteratively optimizing the initial values ​​of the key process parameters according to the preset objective function, the prediction results, and the target values ​​of the mechanical performance indicators includes:

[0082] 1. When there is a gap between the prediction result and the target value of the mechanical performance index, and the current cumulative number of iterations is less than a preset iteration number threshold, the initial value of the key process parameter is obtained, and the cumulative number of iterations refers to the number of iterations for taking the value of the target process parameter.

[0083] It should be noted that the above iterative optimization process is completed in this embodiment by using the fruit fly optimization method. Therefore, before the iterative optimization is performed, the parameters are initialized, including setting the iteration number threshold (maximum iteration number) and the fruit fly population size.

[0084] 2. Determine the key process parameter as fruit fly, and determine the initial position of each fruit fly in a preset spatial coordinate system based on the initial value of the key process parameter.

[0085] It should be noted that the distance from the initial position to the origin in the spatial coordinate system is the corresponding initial value. Each key process parameter corresponds to one or more fruit flies, and the initial positions of different fruit flies corresponding to each key process parameter are different. The number of fruit flies corresponding to each key process parameter depends on the preset fruit fly population size. For example: assuming that a key process parameter corresponds to three fruit flies, then each of the three fruit flies has the same distance from the origin in the spatial coordinate system, but is located in a different position, etc. The dimension of the spatial coordinate system can be set according to actual conditions.

[0086] 3. Perform multiple random flights on each fruit fly to obtain multiple sets of fruit fly position information, wherein the fruit fly position information includes the target position after each fruit fly's random flight, wherein each random flight controls the fruit fly to randomly fly a preset distance from an initial position.

[0087] It should be noted that, assuming that there are currently three key process parameters (the first key process parameter, the second key process parameter, and the third key process parameter), then a set of fruit fly position information includes the target positions of each of the three key process parameters, that is, the target position of the fruit fly after random flight corresponding to the first key process parameter, the target position of the fruit fly after random flight corresponding to the second key process parameter, and the target position of the fruit fly after random flight corresponding to the third key process parameter.

[0088] It should be mentioned that the number of random flights can be set according to actual conditions, such as 10 times. By performing multiple random flights on each fruit fly, multiple target positions after the random flight corresponding to each fruit fly can be obtained. It can be understood that the target positions of the three key process parameters are combined into a set of fruit fly position information.

[0089] It should also be mentioned that the preset distance can be set according to actual conditions.

[0090] For example, suppose there are three key process parameters, namely α, β, and γ. Each key process parameter corresponds to a fruit fly, and the initial position of each fruit fly is The distance from the initial position to the origin of the spatial coordinate system is the initial value of the corresponding key process parameter. Each fruit fly randomly flies a preset distance in the six-dimensional spatial coordinate system, and the new position after the flight, that is, the target position, is (X α , Y α )(X β , Y β )(X γ , Y γ ),in,

[0091] represents the random flight distance of the fruit fly corresponding to α, represents the random flight distance of the fruit fly corresponding to β, represents the random flight distance of the fruit fly corresponding to γ.

[0092] Fourth, obtaining a target distance from the target position to the origin of the spatial coordinate system, and determining the reciprocal of the target distance as a new value of the current key process parameter.

[0093] It should be noted that the data of the target distance is expressed as: Among them, D α represents the target distance of the fruit fly corresponding to α, D β represents the target distance of the fruit fly corresponding to β, D γ represents the target distance of the fruit fly corresponding to γ.

[0094] On this basis, the mathematical expression of the new value of the corresponding key process parameters is: in, is the new value of the key process parameter α, is the new value of the key process parameter β, is the new value of the key process parameter γ.

[0095] 5. Input the new values ​​of all the key process parameters and the initial values ​​of the common process parameters into the mechanical property prediction model, perform mechanical property prediction, and obtain new prediction results.

[0096] 6. Based on the new prediction results, the objective function, and the target values ​​of the mechanical performance indicators, iterative optimization of the initial values ​​of the key process parameters is completed.

[0097] It should be noted that, through the above steps, iterative optimization of the initial values ​​of key process parameters can be well achieved.

[0098] It should also be noted that the mathematical expression of the objective function is:

[0099]

[0100] Among them, min means minimization, K represents the number of mechanical performance indicators, and w j represents the important coefficient of the preset j-th mechanical performance index, x represents the common process parameters in the target process parameters, y represents the key process parameters in the target process parameters, and f j (x, y) represents the prediction result of the mechanical property prediction model based on the input values ​​of x and y, target j Represents the target value of the jth mechanical performance index.

[0101] In addition, the following constraints are included:

[0102]

[0103] target j_min <target j <target j_max

[0104] Among them, y irepresents the i-th key process parameter, represents the minimum value preset for the i-th key process parameter, Indicates the maximum value of the i-th key process parameter preset, target j represents the target value of the jth mechanical performance index, target j_min Indicates the preset minimum value of the jth mechanical performance index, target j_max Indicates the preset maximum value of the jth mechanical performance index. Wherein, j = 1, 2, ..., K, K represents the number of mechanical performance indicators.

[0105] In some embodiments, the step of completing iterative optimization of the initial value of the key process parameter according to the new prediction result, the objective function, and the target value of the mechanical performance index includes:

[0106] 1. According to the new prediction result, the objective function, and the target value of the mechanical performance index, the flavor concentration value corresponding to each group of fruit fly position information is obtained.

[0107] 2. Determine the maximum value among multiple flavor concentration values, and determine the fruit fly position information corresponding to the maximum value as the initial position of the next generation of fruit fly population. Based on the initial position of the next generation of fruit fly population, repeat the steps of random flight, updating key process parameter values, predicting mechanical properties, and acquiring and screening flavor concentration values ​​until the cumulative number of iterations reaches the iteration number threshold.

[0108] 3. If the cumulative number of iterations reaches the iteration number threshold, the current value of the target process parameter is determined as the final value of the target process parameter, and the predicted result of the mechanical property index corresponding to the final value of the target process parameter is determined as the final value of the mechanical property.

[0109] It should be noted that, through the above steps, multi-objective optimization can be well achieved, and the final values ​​of the target process parameters and the final values ​​of the mechanical properties with high accuracy can be obtained. The final value of the mechanical properties is the final design value of each mechanical performance index.

[0110] It should be mentioned that the final values ​​of the target process parameters include the final values ​​of the key process parameters after iterative optimization and the initial values ​​of the common process parameters. It can be understood that during the entire iterative optimization process, the values ​​of the key process parameters are continuously optimized to approach the target values ​​of the mechanical performance indicators, and the values ​​of the common process parameters remain unchanged. Finally, the values ​​of the key process parameters after iterative optimization are determined as their final values, and the initial values ​​of the common process parameters are determined as their final values.

[0111] In some embodiments, according to the new prediction result, the objective function, and the target value of the mechanical performance index, the mathematical expression of the flavor concentration value corresponding to each group of fruit fly position information is obtained as follows:

[0112]

[0113] Among them, F(α, β, γ…) represents the flavor concentration value, and (α, β, γ…) represents the values ​​of multiple key process parameters.

[0114] It should be noted that the above mathematical expression can facilitate subsequent iterative optimization based on the flavor concentration values ​​corresponding to each group of fruit fly position information.

[0115] The steel grade design method in the above embodiment is explained below with a specific embodiment.

[0116] Assume that it is necessary to design a new type of beam steel (steel type to be designed), the target value of the tensile strength of this steel type is 600 MPa, the target value of the yield strength is 540 MPa, and the target value of the elongation is 24%.

[0117] N (nitrogen) content, Ti (titanium) content, Nb (niobium) content, Mn (manganese) content, furnace temperature and finishing rolling outlet temperature are selected as key process parameters, and the remaining components and rolling parameters are used as common process parameters.

[0118] First, the historical production data of the related steel grades of the same type as the new steel grade is obtained. Here, it is assumed that the related steel grades include 510L, 590L, 610L and 700L. The numbers represent the lower limit of the tensile strength of the corresponding related steel grades. The mechanical property prediction model is trained using the historical production data of the related steel grades to obtain a trained mechanical property prediction model.

[0119] Secondly, since steel type 510L is closest to the target value of tensile strength of the above-mentioned new steel type, steel type 510L is determined as the basic steel type, and then the initial values ​​of the target process parameters (key process parameters and common process parameters) of the new steel type are obtained.

[0120] Then, the value ranges of key process parameters are set as constraints, such as the value range of N content is [0, 0.01%], the value range of Ti content is [0, 0.011%], the value range of Nb content is [0, 0.1%], the value range of Mn content is [0.7%, 2%], the value range of furnace temperature is [1150℃, 1300℃], the value range of finishing rolling outlet temperature is [800℃, 950℃], etc.

[0121] Finally, the initial values ​​of the target process parameters are input into the trained mechanical property prediction model to perform mechanical property prediction and obtain the prediction results of the mechanical property indicators. According to the preset objective function, the prediction results, and the target values ​​of the mechanical property indicators, the initial values ​​of the target process parameters are iteratively optimized to obtain the final values ​​of the target process parameters and the final values ​​of the mechanical properties corresponding to the final values. Specifically, the above-mentioned iterative optimization is completed by fruit fly optimization. It should be mentioned that the horizontal and vertical coordinates of the initial position of the fruit fly population of each key process parameter conform to the uniform distribution of [0,1]. The fruit fly population size can be 30, etc. The threshold of the number of iterations can be 100, etc. The important coefficients of each mechanical property indicator can be set according to actual conditions, such as 1, 2, etc.

[0122] The comparison between the final value and the initial value of the target process parameter is shown in Table 1 below:

[0123] Table 1 Comparison between the final and initial values ​​of target process parameters

[0124]

[0125] It can be seen from Table 1 that the final values ​​of each target process parameter are obtained after iterative optimization.

[0126] Table 2 below is a comparison example of mechanical performance indicators corresponding to Table 1:

[0127] Table 2 Comparison example of mechanical performance indicators before and after optimization

[0128]

[0129] Among them, the initial values ​​in Table 2 refer to the predicted values ​​of various mechanical performance indicators predicted according to the initial values ​​of the above target process parameters. It can be seen that the final values ​​of the above mechanical performance indicators are close to their corresponding target values, with high accuracy.

[0130] Correspondingly, please refer to Figures 2 to 4 ,Through the above iterative optimization, the optimized values ​​(values ​​after iterative optimization) of various mechanical performance indicators (tensile strength, yield strength, and elongation) gradually approach their target values. Figures 2 to 4 The horizontal axis in is the cumulative number of iterations.

[0131] This embodiment selects the production data of automobile beam steel of a hot-rolled plate and strip production line for modeling and analysis. The specific steel grades include 510L, 590L, 610L and 700L, and the numbers represent the lower limit of the tensile strength of the corresponding steel grades. The data includes the molten steel composition, rolling temperature, thickness information and mechanical properties test results of the steel coil dimension. Consider designing a new type of beam steel, the average tensile strength target value of this steel is 600, the average yield strength target value is 540, and the average elongation target value is 24. According to the analysis of the data distribution, N, Ti, Nb, Mn, furnace temperature and finishing outlet temperature are selected as key process parameters, and the remaining components and rolling parameters are used as common process parameters. In addition, since the performance of the steel grade to be designed is closest to 510L, 510L is selected as the basic steel grade, the value of the common process parameter is the average value of the corresponding parameter of 510L, and the initial value of the key process parameter is the average value of the corresponding parameter of 510L. In addition, the value ranges of key process parameters are as follows: N content range is [0, 0.01], Ti content range is [0, 0.011], Nb content range is [0, 0.1], Mn content range is [0.7, 2], furnace temperature range is [1150, 1300], and finishing rolling outlet temperature range is [800, 950].

[0132] The steel grade design system provided by the present invention is described below. The steel grade design system described below and the steel grade design method described above can be referenced to each other.

[0133] Please refer to Figure 5 The steel grade design system provided in this embodiment includes:

[0134] An initial data acquisition module 510 is used to acquire target values ​​of mechanical property indicators of the steel type to be designed, and initial values ​​of target process parameters of the steel type to be designed;

[0135] The mechanical property prediction module 520 is used to input the initial value of the target process parameter into a pre-trained mechanical property prediction model to perform mechanical property prediction and obtain the prediction result of the mechanical property index;

[0136] The steel grade design module 530 is used to iteratively optimize the initial value of the target process parameter according to the preset objective function, the prediction result, and the target value of the mechanical property index, so as to obtain the final value of the target process parameter and the final value of the mechanical property corresponding to the final value, wherein the final value of the mechanical property refers to the prediction result obtained by inputting the final value of the target process parameter into the mechanical property prediction model. The initial data acquisition module 510, the mechanical property prediction module 520, and the steel grade design module 530 are connected. The steel grade design system in this embodiment can quickly obtain the steel grade design result, that is, the final value of the target process parameter of the steel grade to be designed, and the final value of the mechanical property corresponding to the final value, with high design efficiency, greatly shortening the steel grade design cycle, reducing the test cost, and having a high degree of automation. Moreover, by iteratively optimizing the initial value of the target process parameter, the mechanical property index corresponding to the optimized value of the target process parameter can gradually approach the corresponding target value, thereby effectively improving the design accuracy and having strong feasibility.

[0137] In some embodiments, the initial data acquisition module 510 is specifically used to acquire historical production data of multiple related steel grades, where the related steel grades refer to steel grades of the same type as the steel grade to be designed, and the historical production data include: historical values ​​of target process parameters, and performance test results;

[0138] Determine the associated steel grade corresponding to the performance test result with the smallest difference from the target value as the basic steel grade;

[0139] All historical values ​​of the target process parameters of the basic steel grade are averaged to obtain initial values ​​of the target process parameters of the steel grade to be designed.

[0140] In some embodiments, a model training module is further included, which is used to obtain historical production data of multiple related steel grades, wherein the related steel grades refer to steel grades of the same type as the steel grade to be designed, and the historical production data include: historical values ​​of target process parameters, and performance test results;

[0141] Preprocessing the historical production data to obtain preprocessed data, wherein preprocessing the historical production data includes removing null values, abnormal values, and outliers;

[0142] Performing feature screening on historical values ​​of target process parameters in the preprocessed data to obtain data to be predicted, wherein the feature screening includes variance screening, collinearity screening and contribution screening;

[0143] Inputting the data to be predicted into the mechanical property prediction model to perform mechanical property prediction to obtain mechanical property prediction results;

[0144] The mechanical property prediction model is trained according to the gap between the mechanical property prediction result and the corresponding performance test result in the preprocessed data.

[0145] In some embodiments, the steel grade design module 530 is specifically used to iteratively optimize the initial values ​​of the key process parameters according to a preset objective function, the prediction results, and the target values ​​of the mechanical properties indicators, and keep the values ​​of the common process parameters unchanged, so as to obtain the final values ​​of all the target process parameters and the final values ​​of the mechanical properties corresponding to the final values.

[0146] In some embodiments, the steel grade design module 530 is further specifically used to obtain the initial value of the key process parameter when there is a gap between the prediction result and the target value of the mechanical property index and the current cumulative number of iterations is less than a preset iteration number threshold, wherein the cumulative number of iterations refers to the number of iterations for obtaining the value of the target process parameter;

[0147] Determine the key process parameter as a fruit fly, and determine the initial position of each fruit fly in a preset spatial coordinate system based on the initial value of the key process parameter;

[0148] Perform multiple random flights on each fruit fly to obtain multiple sets of fruit fly position information, wherein the fruit fly position information includes the target position after each fruit fly has randomly flown, wherein each random flight controls the fruit fly to randomly fly a preset distance from an initial position;

[0149] Obtaining a target distance from the target position to the origin of the spatial coordinate system, and determining the reciprocal of the target distance as a new value of the current key process parameter;

[0150] Inputting the new values ​​of all the key process parameters and the initial values ​​of the common process parameters into the mechanical property prediction model to perform mechanical property prediction and obtain new prediction results;

[0151] According to the new prediction results, the objective function, and the target values ​​of the mechanical performance indicators, iterative optimization of the initial values ​​of the key process parameters is completed.

[0152] In some embodiments, the steel grade design module 530 is further specifically used to obtain the flavor concentration value corresponding to each group of fruit fly position information according to the new prediction result, the objective function, and the target value of the mechanical performance index;

[0153] Determine the maximum value among multiple flavor concentration values, determine the fruit fly position information corresponding to the maximum value as the initial position of the next generation fruit fly population, and repeat the steps of random flight, key process parameter value update, mechanical property prediction, flavor concentration value acquisition and screening based on the initial position of the next generation fruit fly population until the cumulative number of iterations reaches the iteration number threshold;

[0154] If the cumulative number of iterations reaches the iteration number threshold, the current value of the target process parameter is determined as the final value of the target process parameter, and the prediction result of the mechanical property index corresponding to the final value of the target process parameter is determined as the final value of the mechanical property.

[0155] In some embodiments, an electronic device is also provided, which may be a server, and its internal structure is shown in FIG. Figure 6 As shown. The electronic device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile and / or 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 network interface of the electronic device is used to communicate with an external client via a network connection. When the computer program is executed by the processor, the functions or steps on the server side of the above method are implemented.

[0156] In some embodiments, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented: obtaining target values ​​of mechanical properties of the steel type to be designed and initial values ​​of target process parameters of the steel type to be designed; inputting the initial values ​​of the target process parameters into a pre-trained mechanical property prediction model to perform mechanical property prediction and obtain prediction results of the mechanical property indicators; iteratively optimizing the initial values ​​of the target process parameters according to preset objective functions, prediction results, and target values ​​of the mechanical property indicators to obtain final values ​​of the target process parameters and final values ​​of the mechanical properties corresponding to the final values, wherein the final values ​​of the mechanical properties refer to the prediction results obtained by inputting the final values ​​of the target process parameters into the mechanical property prediction model.

[0157] In some embodiments, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: obtaining target values ​​of mechanical properties of the steel type to be designed and initial values ​​of target process parameters of the steel type to be designed; inputting the initial values ​​of the target process parameters into a pre-trained mechanical property prediction model to perform mechanical property prediction and obtain prediction results of the mechanical property indicators; iteratively optimizing the initial values ​​of the target process parameters according to preset objective functions, prediction results, and target values ​​of the mechanical property indicators to obtain final values ​​of the target process parameters and final values ​​of the mechanical properties corresponding to the final values, wherein the final values ​​of the mechanical properties refer to the prediction results obtained by inputting the final values ​​of the target process parameters into the mechanical property prediction model.

[0158] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or electronic device can refer to the relevant descriptions on the server side and the client side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0159] The flow chart and block diagram in the accompanying drawings illustrate the possible implementation architecture, function and operation of the method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0160] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A steel grade design method, characterized in that: include: Obtaining target values ​​of mechanical property indicators of the steel grade to be designed and initial values ​​of target process parameters of the steel grade to be designed; Inputting the initial values ​​of the target process parameters into a pre-trained mechanical property prediction model to perform mechanical property prediction to obtain prediction results of mechanical property indicators; According to the preset objective function, the prediction result, and the target value of the mechanical property index, the initial value of the target process parameter is iteratively optimized to obtain the final value of the target process parameter and the final value of the mechanical property corresponding to the final value. The final value of the mechanical property refers to the prediction result obtained by inputting the final value of the target process parameter into the mechanical property prediction model.

2. The steel grade design method according to claim 1, characterized in that: The target value of the mechanical property index of the steel grade to be designed is a preset value, and the step of obtaining the initial value of the target process parameter of the steel grade to be designed includes: Acquire historical production data of a plurality of associated steel grades, wherein the associated steel grades refer to steel grades of the same type as the steel grade to be designed, and the historical production data include: historical values ​​of target process parameters, and performance test results; Determine the associated steel grade corresponding to the performance test result with the smallest difference from the target value as the basic steel grade; All historical values ​​of the target process parameters of the basic steel grade are averaged to obtain initial values ​​of the target process parameters of the steel grade to be designed.

3. The steel grade design method according to claim 1, characterized in that: The training steps of the mechanical property prediction model include: Acquire historical production data of a plurality of associated steel grades, wherein the associated steel grades refer to steel grades of the same type as the steel grade to be designed, and the historical production data include: historical values ​​of target process parameters, and performance test results; Preprocessing the historical production data to obtain preprocessed data, wherein preprocessing the historical production data includes removing null values, abnormal values, and outliers; Performing feature screening on historical values ​​of target process parameters in the preprocessed data to obtain data to be predicted, wherein the feature screening includes variance screening, collinearity screening and contribution screening; Inputting the data to be predicted into the mechanical property prediction model to perform mechanical property prediction to obtain mechanical property prediction results; The mechanical property prediction model is trained according to the gap between the mechanical property prediction result and the corresponding performance test result in the preprocessed data.

4. The steel grade design method according to claim 1, characterized in that: The target process parameters include preset common process parameters and key process parameters, and the prediction results include predicted values ​​of multiple mechanical property indicators of the steel grade to be designed; The step of iteratively optimizing the initial value of the target process parameter according to the preset objective function, the prediction result, and the target value of the mechanical property index to obtain the final value of the target process parameter and the final value of the mechanical property corresponding to the final value includes: According to the preset objective function, the prediction results, and the target values ​​of the mechanical performance indicators, the initial values ​​of the key process parameters are iteratively optimized, and the values ​​of the common process parameters are kept unchanged to obtain the final values ​​of all the target process parameters and the final values ​​of the mechanical properties corresponding to the final values.

5. The steel grade design method according to claim 4, characterized in that: The step of iteratively optimizing the initial values ​​of the key process parameters according to the preset objective function, the prediction result, and the target value of the mechanical performance index comprises: When there is a gap between the prediction result and the target value of the mechanical property index, and the current cumulative number of iterations is less than a preset iteration number threshold, obtaining the initial value of the key process parameter, wherein the cumulative number of iterations refers to the number of iterations for obtaining the value of the target process parameter; Determine the key process parameter as a fruit fly, and determine the initial position of each fruit fly in a preset spatial coordinate system based on the initial value of the key process parameter; Perform multiple random flights on each fruit fly to obtain multiple sets of fruit fly position information, wherein the fruit fly position information includes the target position after each fruit fly has randomly flown, wherein each random flight controls the fruit fly to randomly fly a preset distance from an initial position; Obtaining a target distance from the target position to the origin of the spatial coordinate system, and determining the reciprocal of the target distance as a new value of the current key process parameter; Inputting the new values ​​of all the key process parameters and the initial values ​​of the common process parameters into the mechanical property prediction model to perform mechanical property prediction and obtain new prediction results; According to the new prediction results, the objective function, and the target values ​​of the mechanical performance indicators, iterative optimization of the initial values ​​of the key process parameters is completed.

6. The steel grade design method according to claim 5, characterized in that: The step of completing iterative optimization of the initial values ​​of the key process parameters according to the new prediction results, the objective function, and the target values ​​of the mechanical performance indicators includes: According to the new prediction result, the objective function, and the target value of the mechanical performance index, the flavor concentration value corresponding to each group of fruit fly position information is obtained; Determine the maximum value among multiple flavor concentration values, determine the fruit fly position information corresponding to the maximum value as the initial position of the next generation fruit fly population, and repeat the steps of random flight, key process parameter value update, mechanical property prediction, flavor concentration value acquisition and screening based on the initial position of the next generation fruit fly population until the cumulative number of iterations reaches the iteration number threshold; If the cumulative number of iterations reaches the iteration number threshold, the current value of the target process parameter is determined as the final value of the target process parameter, and the prediction result of the mechanical property index corresponding to the final value of the target process parameter is determined as the final value of the mechanical property.

7. The steel grade design method according to claim 6, characterized in that: According to the new prediction results, the objective function, and the target value of the mechanical performance index, the mathematical expression of the flavor concentration value corresponding to each group of fruit fly position information is obtained as follows: Among them, F(α, β, γ…) represents the flavor concentration value, (α, β, γ…) represents the values ​​of multiple key process parameters, min represents minimization, K represents the number of mechanical performance indicators, and w j represents the important coefficient of the preset j-th mechanical performance index, x represents the common process parameters in the target process parameters, y represents the key process parameters in the target process parameters, and f j (x, y) represents the prediction result of the mechanical property prediction model based on the input values ​​of x and y, target j Represents the target value of the jth mechanical performance index.

8. A steel grade design system, characterized in that: include: An initial data acquisition module, used to acquire target values ​​of mechanical property indicators of the steel type to be designed, and initial values ​​of target process parameters of the steel type to be designed; A mechanical property prediction module is used to input the initial value of the target process parameter into a pre-trained mechanical property prediction model to perform mechanical property prediction and obtain a prediction result of the mechanical property index; The steel grade design module is used to iteratively optimize the initial values ​​of the target process parameters according to a preset objective function, the prediction results, and the target values ​​of the mechanical property indicators to obtain the final values ​​of the target process parameters and the final values ​​of the mechanical properties corresponding to the final values, wherein the final values ​​of the mechanical properties refer to the prediction results obtained by inputting the final values ​​of the target process parameters into the mechanical property prediction model.

9. An electronic device, characterized in that: It comprises a processor, a memory and a communication bus; the communication bus is used to connect the processor and the memory; the processor is used to execute the computer program stored in the memory to implement the steel grade design method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program is used to enable a computer to execute the steel grade design method according to any one of claims 1 to 7.

Images