A method for predicting final cooling temperature in a power station steel plate production process

Abstract

A kind of final cooling temperature prediction method of power station steel plate production process belongs to power station steel plate production and manufacturing technical field, to solve the problem of poor accuracy and large error of existing final cooling temperature prediction method of steel plate in cooling process. It includes: in the cooling process of power station steel plate, the data of steel plate parameter in cooling process is collected, and the data is preprocessed;With the data of steel plate parameter in cooling process as the input data of XGBoost prediction model, with the final cooling temperature of steel plate as the output data of XGBoost prediction model;The hyperparameters of XGBoost prediction model are optimized and adjusted using Bayes law, a BO-XGBoost-based prediction model of final cooling temperature of power station steel plate is constructed, and a BO-XGBoost prediction model is obtained;The final cooling temperature of power station steel plate is predicted in real time using BO-XGBoost prediction model.The present application is used for monitoring of steel plate in cooling process.

Description

TECHNICAL FIELD

[0001] The application relates to a final cooling temperature prediction method for a power station steel plate production process and belongs to the technical field of power station steel plate production and manufacturing. BACKGROUND

[0002] In recent years, with the economy shifting from the high-speed development stage to the high-quality development stage, the steel industry has entered a stage of structural adjustment and transformation. High quality, low cost, environmental friendliness and good stability are problems that the steel industry urgently needs to solve, and steel enterprises are increasing their efforts to technological innovation, focusing on developing green, intelligent new technologies, new equipment and new products, and enhancing core competitiveness.

[0003] Steel plates are important raw materials for power station boilers and pressure vessel manufacturing. In the current steel plate post-rolling cooling control system, an energy balance equation is generally used to describe various heat exchange phenomena of the steel plate during the cooling process. By continuously optimizing model construction and parameter prediction, the control cooling mechanism model has played an important role in the steel plate cooling process. Although the heat transfer mechanism model plays an important role in the cooling process, since the steel plate involves multiple heat exchange forms in the heat exchange process, it is a multiple input and output process, and the heat exchange process is influenced by multiple factors such as the chemical composition, thickness, speed, cooling header flow, water temperature, open cooling temperature of the steel plate, and the interference of unmeasurable intermediate variables and external environmental changes. Therefore, the mechanism modeling method is difficult to accurately describe the physical and chemical changes of the steel plate in the cooling process in the process of constructing the cooling process mathematical model, resulting in low control accuracy of the mathematical model when the production conditions and equipment working conditions change. Especially in the case of changing specifications and changing batches, the temperature control deviation of the first steel plate is often large, thereby causing poor stability of the steel plate product. SUMMARY

[0004] The application aims to solve the problems of poor accuracy and large error of the existing final cooling temperature prediction method for the steel plate in the cooling process, and provides a final cooling temperature prediction method for a power station steel plate production process.

[0005] The final cooling temperature prediction method for the power station steel plate production process comprises the following steps:

[0006] S1, in the cooling process of the power station steel plate, collecting data of the steel plate parameters in the cooling process and preprocessing the data;

[0007] S2, taking the data of the steel plate parameters in the cooling process as input data of an XGBoost prediction model and taking the final cooling temperature of the steel plate as output data of the XGBoost prediction model;

[0008] Bayes' theorem was used to optimize and adjust the hyperparameters of the XGBoost prediction model, and a prediction model for the final cooling temperature of power plant steel plates based on BO-XGBoost was constructed to obtain the BO-XGBoost prediction model.

[0009] S3. Input the data of the cooling process steel plate parameters obtained in S1 into the BO-XGBoost prediction model to predict the final cooling temperature of the power plant steel plate in real time.

[0010] Preferably, the cooling process parameters collected in S1 include: chemical composition content, starting cooling temperature, cooling rate, number of manifolds, steel plate speed, steel plate thickness, steel plate length, water temperature, total flow rate, furnace exit temperature, and final rolling temperature.

[0011] The chemical composition content includes the contents of C, Cr, Mn, Mo and Nb.

[0012] Preferably, the specific method for preprocessing the data as described in S1 includes:

[0013] S1-1. Select the input and output parameters of the model according to the production process;

[0014] S1-2. Use the Raida criterion to remove outliers from the parameters;

[0015] The elimination method is:

[0016]

[0017] Where, x i Let i represent the i-th sample. S represents the average of all samples. x This represents the standard deviation of all samples;

[0018]

[0019] Where i = 1, 2, ..., n, n represents the total number of samples;

[0020] S1-3. Standardize the parameter data using the Z-score method;

[0021]

[0022] Where, q j This represents the value of the unstandardized power plant steel plate data in the j-th column, q'. j μ represents the value after standardization. j σ represents the average value of the j-th column of data in the power plant steel plate data. j represents the standard deviation of the j-th column of data in the power plant steel plate data, and m represents the total number of columns in the power plant steel plate data;

[0023] S1-4, divide the data set, 80% of the data is divided into a training set, and 20% of the data is divided into a test set.

[0024] Preferably, the input parameters and output parameters of the selection model of S1-1 specifically include:

[0025] The chemical composition content, cooling temperature, cooling rate, header number, steel plate speed, steel plate thickness, steel plate length, water temperature, total flow, discharge temperature and finish rolling temperature are taken as the input parameters of the model;

[0026] The final cooling temperature is taken as the output parameter of the model.

[0027] Preferably, the XGBoost prediction model of S2 includes:

[0028] S2-1, according to the training data set and the loss function, the objective function is calculated;

[0029] The training set data set T={(x1,y1),(x2,y2),...,(x n ,y n )} and the loss function The objective function Obj is calculated:

[0030]

[0031] Wherein, k represents the kth CART tree, k=1,2,...,K, K represents the total CART tree, y i represents the true value of sample x i , represents the predicted value of sample x i ,

[0032] represents the predicted value of sample x i of the t-1th iteration, f k (x i ) represents the prediction output of the kth CART tree, f t (x i ) represents the prediction output of sample x i of the tth iteration;

[0033] S2-2, the objective function is introduced into regularization;

[0034]

[0035] Wherein, Ω(f) represents the introduced regularization function, T represents the number of leaf nodes, γ represents the number of leaf nodes after pruning, l2 represents the 2-norm of the leaf node weight vector, λ represents the l2 regularization term, w jdenotes the weight of the jth leaf node, j denotes the jth leaf node;

[0036] S2-3, the model objective function of the tth iteration is calculated as:

[0037]

[0038] wherein, Ω(f k ) denotes the regularization function of the kth CART tree;

[0039] S2-4, the overall objective function of the model is calculated as:

[0040]

[0041] S2-5, the overall objective function obtained in S2-4 is expanded by using the second-order Taylor:

[0042]

[0043] wherein, g i denotes the first-order partial derivative after Taylor expansion h i denotes the second-order partial derivative after Taylor expansion

[0044]

[0045] S2-6, the iteration of the tree model is converted into the iteration of the leaf nodes of the tree, and the objective function is transformed as:

[0046]

[0047] S2-7, the constant term is omitted, and the objective function is simplified as:

[0048]

[0049] S2-8, all samples x i belonging to the jth leaf node are divided into a sample set I j of a leaf node, that is, all samples are grouped according to the leaf nodes, and finally, the following is obtained:

[0050]

[0051] S2-9, let obtain the optimal solution point of the objective function and the optimal solution Obj:

[0052]

[0053] Preferably, S2 optimizes and adjusts the hyperparameters of the XGBoost prediction model by using the Bayes theorem, and the specific method for constructing the BO-XGBoost-based power station steel plate final cooling temperature prediction model comprises the following steps:

[0054] Calculate the prior function f(x) and the acquisition function f PI (x):

[0055] f(x)~gp(m(x),k(x,x'))

[0056]

[0057] Wherein, m(x) represents the mean function, k(x,x') represents the covariance function, gp(·) represents the Gaussian regression function, u(x) represents the mean of the target function value,

[0058] σ(x) represents the variance of the target function value, f(x + ) represents the optimal target function value, ξ represents the parameter, and Φ() represents the normal distribution cumulative density function.

[0059] Calculate the Bayes theorem:

[0060]

[0061] Wherein, f represents the target function, D 1:t represents the observed set, p(f|D 1:t ) represents the likelihood distribution of y, p(f) represents the prior probability distribution of f, and p(D 1:t ) represents the marginal likelihood distribution of f.

[0062] Calculate the optimal parameter combination x * :

[0063]

[0064] Wherein, χ represents the decision space.

[0065] Preferably, S2 obtains the BO-XGBoost prediction model, which further comprises the following steps:

[0066] Input the data of the training set obtained in S1 into the BO-XGBoost prediction model, obtain the BO-XGBoost prediction model and save it;

[0067] Input the data of the test set obtained in S1 into the BO-XGBoost prediction model, and verify the accuracy of the BO-XGBoost prediction model.

[0068] Advantages of this invention: The method for predicting the final cooling temperature of power plant steel plates proposed in this invention utilizes the BO-XGBoost model. Compared with DNN and BO-GBDT models, the BO-XGBoost model exhibits the highest prediction accuracy and the smallest prediction error. The model's training set shows excellent MAE, RMSE, and R-values. 2 The values ​​were 1.47℃, 2.03℃, and 0.9947℃, respectively, with MAE, RMSE, and R values ​​on the test set. 2 The temperatures were 1.97℃, 2.91℃, and 0.9886℃, respectively. Attached Figure Description

[0069] Figure 1 This is a flowchart of the final cooling temperature prediction method for the power plant steel plate production process described in this invention.

[0070] Figure 2 This is a diagram showing the comparison between the model's predicted and actual final cooling temperatures. Detailed Implementation

[0071] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0072] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0073] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the scope of the invention.

[0074] Example 1:

[0075] The following is combined Figure 1 This embodiment describes a method for predicting the final cooling temperature in the production process of power plant steel plates, which includes:

[0076] S1. During the cooling process of the power plant steel plate, collect data on the parameters of the steel plate during the cooling process and preprocess the data;

[0077] S2. Use the data of the steel plate parameters during the cooling process as the input data of the XGBoost prediction model, and use the final cooling temperature of the steel plate as the output data of the XGBoost prediction model.

[0078] The Bayesian theorem is used to optimize and adjust the hyperparameters of the XGBoost prediction model, a BO-XGBoost-based power station steel plate final cooling temperature prediction model is constructed, and a BO-XGBoost prediction model is obtained.

[0079] S3, input the data of the cooling process steel plate parameters obtained in S1 into the BO-XGBoost prediction model to realize real-time prediction of the power station steel plate final cooling temperature.

[0080] Further, the collection of cooling process parameters in S1 includes chemical composition content, open cooling temperature, cooling rate, number of headers, steel plate speed, steel plate thickness, steel plate length, water temperature, total flow, discharge temperature and finish rolling temperature.

[0081] The chemical composition content includes the contents of C, Cr, Mn, Mo and Nb.

[0082] Further, the specific method of S1 for pre-processing the data includes:

[0083] S1-1, according to the production process, the input parameters and output parameters of the model are selected;

[0084] S1-2, the outliers in the parameters are removed by using the Laplace criterion;

[0085] The removal method is:

[0086]

[0087] Wherein, x i represents the i-th sample, represents the average value of all samples, S x represents the standard deviation of all samples;

[0088]

[0089] Wherein, i=1,2, …, n, n represents the total number of samples;

[0090] S1-3, the parameter data is standardized by using the Z-score method;

[0091]

[0092] Wherein, q j represents the value of the power station steel plate data in the j-th column before standardization, q' j represents the value after standardization, μ j represents the average value of the j-th column data in the power station steel plate data, σ j represents the standard deviation of the j-th column data in the power station steel plate data, and m represents the total number of columns of the power station steel plate data.

[0093] S1-4, divide the data set, 80% of the data is divided into a training set, and 20% of the data is divided into a test set.

[0094] Further, the input parameters and output parameters of the model selected in S1-1 specifically include:

[0095] The chemical composition content, cooling temperature, cooling rate, header number, steel plate speed, steel plate thickness, steel plate length, water temperature, total flow, discharge temperature and finish rolling temperature are taken as input parameters of the model;

[0096] The final cooling temperature is taken as the output parameter of the model.

[0097] Further, the XGBoost prediction model of S2 includes:

[0098] S2-1, according to the training data set and the loss function, the objective function is calculated;

[0099] The training set data set T={(x1, y1), (x2, y2),..., (x n , y n )} and the loss function The objective function Obj is calculated:

[0100]

[0101] Wherein, k represents the kth CART tree, k=1, 2,..., K, K represents the total CART tree, y i represents the true value of sample x i , represents the predicted value of sample x i , represents the predicted value of sample x i of the t-1th iteration, f k (x i ) represents the prediction output of the kth CART tree, f t (x i ) represents the prediction output of sample x i of the tth iteration;

[0102] S2-2, the objective function is introduced into regularization;

[0103]

[0104] Wherein, Ω(f) represents the introduced regularization function, T represents the number of leaf nodes, γ represents the number of leaf nodes after pruning, l2 represents the 2-norm of the leaf node weight vector, λ represents the l2 regularization term, w j represents the weight of the jth leaf node, and j represents the jth leaf node.

[0105] S2-3, the model objective function of the tthiteration is calculated as:

[0106]

[0107] wherein, Ω(f k ) represents the regularization function of the kthCART tree;

[0108] S2-4, the overall objective function of the model is calculated as:

[0109]

[0110] S2-5, the overall objective function obtained in S2-4 is expanded using the second-order Taylor:

[0111]

[0112] wherein, g i represents the first-order partial derivative after Taylor expansion h i represents the second-order partial derivative after Taylor expansion

[0113]

[0114] S2-6, the iteration of the tree model is converted into the iteration about the leaves of the tree, and the objective function is transformed as:

[0115]

[0116] S2-7, the constant term is omitted, and the objective function is simplified as:

[0117]

[0118] S2-8, all samples x i belonging to the jthleaf node are divided into a sample set I j , that is, all samples are grouped according to the leaf node, and finally:

[0119]

[0120] S2-9, let obtain the optimal solution point of the objective function and the optimal solution Obj:

[0121]

[0122] Further, the method for optimizing and adjusting the hyperparameters of the XGBoost prediction model by using the Bayes theorem in S2 comprises:

[0123] Calculate the prior function f(x) and the acquisition function f PI (x):

[0124] f(x)~gp(m(x),k(x,x'))

[0125]

[0126] Wherein, m(x) represents the mean function, k(x,x') represents the covariance function, gp(·) represents the Gaussian regression function, u(x) represents the mean of the target function value, sigma(x) represents the variance of the target function value, f(x + ) represents the optimal target function value, xi represents the parameter, and Phi() represents the normal distribution cumulative density function.

[0127] Calculate the Bayes theorem:

[0128]

[0129] Wherein, f represents the target function, D 1:t represents the observed set, p(f|D 1:t ) represents the likelihood distribution of y, p(f) represents the prior probability distribution of f, and p(D 1:t ) represents the marginal likelihood distribution of f.

[0130] Calculate the optimal parameter combination x * :

[0131]

[0132] Wherein, chi represents the decision space.

[0133] Further, the method for obtaining the BO-XGBoost prediction model in S2 further comprises:

[0134] Input the data of the training set obtained in S1 into the BO-XGBoost prediction model, obtain the BO-XGBoost prediction model and save it;

[0135] Input the data of the test set obtained in S1 into the BO-XGBoost prediction model, and verify the precision of the BO-XGBoost prediction model.

[0136] In the present application, the terminal cooling temperature prediction method of the power station steel plate production process comprises the following steps:

[0137] Step 1: Hot rolled medium plate cooling process data acquisition and preprocessing;

[0138] Step 1.1: The chemical composition content (C, Cr, Mn, Mo, Nb), open cooling temperature, cooling rate, header number, steel plate speed, steel plate thickness, steel plate length, water temperature, total flow, discharge temperature, final rolling temperature, a total of 15 parameters, as the input of the model, and the final cooling temperature as the output of the model;

[0139] Step 1.2: Adopt Pauta criterion to eliminate outliers, and after elimination, there are a total of 31971 data samples left.

[0140]

[0141]

[0142] In the formula: and S y are the mean and standard deviation of all samples; L is the sample number; y i is the i-th sample;

[0143] Step 1.3: Standardize the power station steel plate data by Z-score method;

[0144]

[0145] In the formula: x i is the value of the unstandardized power station steel plate data on the i-th column; x i ' is the standardized value; μ i is the average value of the i-th column data in the power station steel plate data; σ i is the standard deviation of the i-th column data in the power station steel plate data;

[0146] Step 1.4: Randomly divide the data set, and divide 80% of the data into the training set and 20% of the data into the test set.

[0147] Step 2: Construct the XGBoost-based power station steel plate final cooling temperature prediction model;

[0148] Step 2.1: According to the training data set T={(x1,y1),(x2,y2),...,(x n ,y n )}, loss function Calculate the objective function:

[0149]

[0150] In the formula: i is the sample number; k is the k-th tree; y i and are the sample xi the true value and the predicted value; f k (x i ) is the predicted output of the kth CART tree.

[0151] Step 2.2: The objective function introduces regularization:

[0152]

[0153] where γT is the number of leaf nodes; l2 is the 2-norm of the leaf node weight vector.

[0154] Step 2.3: Calculate the model objective function at the tth iteration:

[0155]

[0156] Step 2.4: Calculate the overall objective function of the model:

[0157]

[0158] Step 2.5: The objective function adopts a second-order Taylor expansion:

[0159]

[0160] where g i and h i are the first-order partial derivatives and the second-order partial derivatives

[0161] Step 2.6: Convert the iteration of the tree model into iteration about the leaf nodes of the tree, and deform the objective function:

[0162]

[0163] Step 2.7: The regularization term involving t-1 trees is omitted as a constant, and the objective function is simplified:

[0164]

[0165] Step 2.8: All samples x i belonging to the jth leaf node are included in a leaf node sample set I j , that is, all samples are grouped according to leaf nodes, and finally we get:

[0166]

[0167] where w j is the weight of the leaf node j; and respectively represent the first-order partial derivative accumulation sum and the second-order partial derivative accumulation sum of the samples contained in the leaf node j.

[0168] Step 2.9: Finally, let The optimal solution point and the optimal solution of the objective function are as follows:

[0169]

[0170] Step 3: Adjust the XGBoost model hyperparameters using Bayesian optimization to build a power station steel plate final cooling temperature prediction model based on BO-XGBoost;

[0171] Step 3.1: Calculate the prior function and acquisition function:

[0172] f(x)~gp(m(x),k(x,x'))

[0173]

[0174] In the formula: m(x) is the mean function, m(x)=E[f(x)], and for simplicity, the mean function m(x)=0 is usually set; k(x,x') is the covariance function, k(x,x')=E[(f(x)-m(x))(f(x')-m(x'))]; u(x),σ(x) are the mean and variance of the objective function value respectively; f(x + ) is the best objective function value; ξ is a parameter; Φ(*) is the normal distribution cumulative density function.

[0175] Step 3.2: Calculate the Bayesian law:

[0176]

[0177] In the formula: f is the objective function; D 1:t ={(x1,y1),(x2,y2),...,(x t ,y t )} is the observed set; x t is the decision vector; y t is the observed value; p(D 1:t |f) is the likelihood distribution of y; p(f) is the prior probability distribution of f; p(D 1:t ) is the marginal likelihood distribution of f.

[0178] Step 3.3: Calculate the optimal parameter combination, and the final BO-XGBoost model parameters are shown in Table 1:

[0179]

[0180] In the formula: x *where x is the optimal parameter combination; chi is the decision space; and f(x) is the objective function.

[0181] Table 1

[0182] XGBoost parameters Parameter meaning Search space Optimal parameter value learning_rate Learning rate (0.1,1) 0.3583 n_estimators Number of decision trees (0,1000) 1000 max_depth Maximum tree depth (1,10) 3 min_child_weight Minimum leaf node sample weight and (1,10) 3

[0183] Step 4: training and saving the BO-XGBoost-based power station steel plate final cooling temperature prediction model;

[0184] Step 4.1: inputting the training set data obtained in step 1 into the BO-XGBoost model to obtain a BO-XGBoost prediction model and save the model;

[0185] Step 4.2: inputting the test set data obtained in step 1 into the BO-XGBoost model obtained in step 4.1, using three model evaluation indexes of mean absolute error (MAE), root mean square absolute error (RMSE) and determination coefficient (R2) to verify the prediction accuracy, and repeating steps 3.1 to 4.1 until the requirement is met. By comparing the DNN model, the BO-GBDT model and the XGBoost, the prediction result is as shown in Table 2, which shows that the prediction result of the BO-XGBoost model in the application is the highest. Figure 2

[0186] Although the application has been described herein with reference to particular embodiments, it is to be understood that these examples are merely illustrative of the principles and applications of the present application. It is therefore to be understood that numerous modifications can be made to the illustrative embodiments and that other arrangements can be devised without departing from the spirit and scope of the present application as defined by the appended claims. It will be understood that different embodiments described herein can be combined with each other in different ways. It will also be understood that features described with respect to one embodiment can be used in other embodiments described herein.​

Claims

1. A method for predicting the final cooling temperature in the production process of power plant steel plates, characterized in that, It includes: S1. During the cooling process of the power plant steel plate, collect data on the parameters of the steel plate during the cooling process and preprocess the data; S2. Use the data of the steel plate parameters during the cooling process as the input data of the XGBoost prediction model, and use the final cooling temperature of the steel plate as the output data of the XGBoost prediction model. Bayes' theorem was used to optimize and adjust the hyperparameters of the XGBoost prediction model, and a prediction model for the final cooling temperature of power plant steel plates based on BO-XGBoost was constructed to obtain the BO-XGBoost prediction model. S3. Input the data of the cooling process steel plate parameters obtained in S1 into the BO-XGBoost prediction model to predict the final cooling temperature of the power plant steel plate in real time. The XGBoost prediction model described in S2 includes: S2-1. Calculate the objective function based on the training dataset and the loss function; Training set dataset loss function Calculate the objective function : ； ； in, Indicates the first A CART tree, , This represents the total number of CART trees. Indicates sample The true value, Indicates sample The predicted value, Indicates the first Round of iteration samples The predicted value, Indicates the first The predicted output of a CART tree, Indicates the first Round of iteration samples The predicted output; S2-2, Introduce regularization into the objective function; ； in, This indicates the introduced regularization function. Indicates the number of leaf nodes. The number of pruned branches indicates the number of leaf nodes. The 2-norm represents the weight vector of the leaf nodes. express Regularization term, Indicates the first The weight of each leaf node, Indicates the first leaf nodes; S2-3, Calculate the... The objective function of the round-iteration model is: ； in, Indicates the first The regularization function for a CART tree; S2-4. Calculate the overall objective function of the model: ； S2-5. Expand the overall objective function obtained in S2-4 using second-order Taylor: in, This represents the first-order partial derivative after Taylor expansion. , Indicates the second-order partial derivative after Taylor expansion ; S2-6. Transform the iteration of the tree model into iteration over the leaf nodes of the tree. The objective function is then transformed into: ； S2-7. By omitting the constant term, the objective function is simplified to: ； S2-8, belonging to the first All samples of leaf nodes The sample set assigned to a leaf node In this process, all samples are grouped according to the leaf nodes, resulting in: ； S2-9, Order , To obtain the optimal solution point of the objective function and optimal solution : 。 2. The method for predicting the final cooling temperature in the production process of power plant steel plates according to claim 1, characterized in that, The parameters collected in S1 for the cooling process include: chemical composition content, starting cooling temperature, cooling rate, number of manifolds, steel plate speed, steel plate thickness, steel plate length, water temperature, total flow rate, furnace exit temperature, and final rolling temperature. The chemical composition content includes the contents of C, Cr, Mn, Mo and Nb.

3. The method for predicting the final cooling temperature in the production process of power plant steel plates according to claim 2, characterized in that, The specific methods for data preprocessing described in S1 include: S1-1. Select the input and output parameters of the model according to the production process; S1-2. Use the Raida criterion to remove outliers from the parameters; The elimination method is: ； in, Indicates the first One sample, This represents the average of all samples. This represents the standard deviation of all samples; ， ； in, , Indicates the total number of samples; S1-3. Standardize the parameter data using the Z-score method; ； ； in, This indicates that the power plant steel plate data that has not undergone standard processing is in the first... The values ​​in the column, This represents the value after standardization. This indicates the first [item] in the power plant steel plate data. The average of the column data, This indicates the first [item] in the power plant steel plate data. Standard deviation of the data This indicates the total number of columns for the power plant steel plate data; S1-4. Divide the dataset into two parts: 80% of the data is divided into the training set and 20% of the data is divided into the test set.

4. The method for predicting the final cooling temperature in the production process of power plant steel plates according to claim 3, characterized in that, The input and output parameters of the selection model described in S1-1 specifically include: The chemical composition content, cooling temperature, cooling rate, number of manifolds, steel plate speed, steel plate thickness, steel plate length, water temperature, total flow rate, furnace exit temperature, and final rolling temperature are used as input parameters for the model. The final cooling temperature is used as the output parameter of the model.

5. The method for predicting the final cooling temperature in the production process of power plant steel plates according to claim 1, characterized in that, The specific method described in S2 for optimizing and adjusting the hyperparameters of the XGBoost prediction model using Bayes' theorem to construct a prediction model for the final cooling temperature of power plant steel plates based on BO-XGBoost includes: Calculate the prior function and acquisition function : in, Represents the mean function, Represents the covariance function. This represents the Gaussian regression function. This represents the mean of the objective function values. This represents the variance of the objective function value. This represents the optimal objective function value. Indicates parameters, This represents the accumulation density function of a normal distribution. Calculate Bayes' theorem: in, Describe the objective function. Indicates the observed set, express The likelihood distribution, express The prior probability distribution, Indicates marginalization The marginal likelihood distribution; Calculate the optimal parameter combination : in, This represents the decision-making space.

6. The method for predicting the final cooling temperature in the production process of power plant steel plates according to claim 1, characterized in that, S2 also includes obtaining the BO-XGBoost prediction model: Input the training set data obtained in S1 into the BO-XGBoost prediction model to obtain and save the BO-XGBoost prediction model; The test set data obtained from S1 is input into the BO-XGBoost prediction model to verify the accuracy of the BO-XGBoost prediction model.

Images