Heat flow density adaptive coefficient calculation method based on data driving

Abstract

The invention relates to a data driving-based heat flow density adaptive coefficient calculation method, which comprises the following steps of: acquiring cooling process parameters, strip steel information parameters and cooling water temperature as input characteristic parameters, taking a heat flow density adaptive coefficient as an output characteristic parameter, and forming an original data set by the input characteristic parameters and the output characteristic parameter; performing data cleaning on the original data set, and dividing the original data set into a training set, a test set and a verification set; constructing a heat flux density adaptive coefficient stacking model, and obtaining the optimal number and type of base learners by adopting a chaos Harris eagle optimization algorithm; performing optimization on hyper-parameters of the optimal basis learner by adopting a chaos Harris eagle optimization algorithm; inputting the training set into a stacking model for training; the test set is adopted to verify the generalization performance of the stacking model; the information parameters of the current strip steel are collected, and the heat flux density adaptive coefficient of the current strip steel is predicted through the tested stacking model.

Description

Technical Field

[0001] The invention belongs to the technical field of hot rolling and relates to a method for calculating a heat flux adaptive coefficient based on data driving. Background Art

[0002] Post-rolling cooling is an important process for hot-rolled steel production. By controlling the temperature and cooling path of the strip during the cooling process, hot-rolled products with expected microstructure and mechanical properties can be obtained. As the market's requirements for the quality of hot-rolled products continue to increase, the temperature control accuracy of the strip during the cooling process has also increased. However, post-rolling cooling is a process involving multiple variables and factors, and these variables have the characteristics of strong coupling, nonlinearity, and time-varying. It is difficult to quantitatively describe the process with high precision using a simplified heat transfer model alone. Therefore, the temperature control of the post-rolling cooling process has always been a key and difficult issue of concern in this industry, especially for the control of the temperature of the strip head. Due to the inability to perform feedback control, the temperature control accuracy of the head is lower. In actual production, introducing a model correction coefficient has become a simple and effective way. Since this coefficient is generally directly involved in the calculation of the heat transfer model, the initial value of this coefficient directly determines the calculation accuracy of the heat transfer model for strips of different steel grades and specifications, thereby determining the temperature control accuracy of the strip head.

[0003] In order to improve the calculation accuracy of the heat transfer model of the post-rolling cooling process, domestic scholars have successively proposed different calculation methods for model correction coefficients for different control systems. The Chinese patent "CN105032951B A control method for improving the accuracy and self-learning efficiency of ultra-fast cooling temperature model" comprehensively considers the influence of factors such as steel physical parameters, steel plate specifications, medium water temperature, steel plate temperature, and aging on the heat transfer coefficient. Construct a multidimensional spatial relationship to effectively describe the relationship between the heat transfer coefficients under the action of different influencing factors. Quickly and accurately predict the value of the heat transfer coefficient under the target influence conditions, and finally achieve precise control of the temperature model. The Chinese patent "CN111814861B An online cooling control method based on dual self-learning models" combines the characteristics of unsupervised spatial clustering models and supervised deep neural network prediction models to establish a system structure with parallel and weight-sharing dual self-learning models. It realizes rapid combined self-learning under low-cost data conditions, improving the robustness and learning efficiency of the overall cooling control system. The Chinese patent "CN108446454B A method for improving the calculation accuracy of layer cooling model setting" considers the influence of threading speed on heat transfer, and establishes a threading speed influence coefficient function according to the strip steel grade, cooling strategy, target thickness layer and target coiling temperature layer. The threading speed influence coefficient is used to correct the water cooling self-learning coefficient. It effectively solves the problem of low calculation accuracy of layer cooling model setting. The Chinese patent "CN104942025B A method for compensating the genetic coefficient of the hot rolling coiling temperature model after stopping rolling" is a method for compensating the genetic coefficient of the hot rolling coiling temperature model after stopping rolling. It can appropriately compensate for the genetic coefficient when the rolling line stops rolling. When the rolling line resumes rolling, the coiling temperature can be basically controlled within the tolerance range, thereby reducing quality loss.

[0004] The above research has improved the calculation accuracy of the heat transfer model to a certain extent, but basically only considers the influence of a single factor on heat transfer. The temperature change of the strip in the cooling zone is affected by many factors such as rolling speed, steel type, specification, cooling water temperature, pressure, cooling mode, etc. Especially as market orders tend to be small batches and diversified, the steel type and specification change frequently in production, which has a more significant impact on the calculation accuracy of the heat transfer model. However, the above invention does not take into account the steel type, specification, cooling water state and cooling mode. Summary of the invention

[0005] In order to solve the above technical problems, the purpose of the present invention is to provide a method for calculating the heat flux adaptive coefficient based on data driving.

[0006] The present invention provides a method for calculating heat flux adaptive coefficient based on data driving, comprising: Step 1: Collect cooling process parameters, strip information parameters and cooling water temperature as input characteristic parameters, and use heat flux adaptive coefficient as output characteristic parameter. The input characteristic parameters and output characteristic parameters constitute the original data set. Step 2: Clean the original data set and divide the cleaned data set into training set, test set and validation set in a ratio of 7:2:1; Step 3: Construct a heat flux adaptive coefficient stacking model and use the chaos Harris Eagle optimization algorithm to obtain the optimal number and type of base learners; Step 4: Use the Chaos Harris Eagle Optimization Algorithm to optimize the hyperparameters of the obtained optimal base learner to obtain the optimal hyperparameter combination; Step 5: Input the training set into the heat flux density adaptive coefficient stacking model for training to obtain the trained heat flux density adaptive coefficient stacking model; Step 6: Use the test set to verify the generalization performance of the heat flux adaptive coefficient stacking model; Step 7: Collect the current strip information parameters and use the tested heat flux density adaptive coefficient stacking model to predict the current heat flux density adaptive coefficient of the strip.

[0007] The data-driven heat flux adaptive coefficient calculation method of the present invention has the following beneficial effects: The calculation method of the present invention is based on a large amount of historical production data, combined with the chaotic Harris Eagle optimization algorithm and the stacking model, to establish a prediction model for the heat flux adaptive coefficient. Compared with the case-based reasoning model used on site, the model proposed by the present invention has higher accuracy, improves the calculation accuracy of the heat transfer model of the post-rolling cooling process, and ensures the control accuracy and quality of the strip head temperature. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is a flow chart of a method for calculating heat flux adaptive coefficient based on data drive of the present invention; Figure 2 A graph showing changes in the first objective function during the optimization process of determining the optimal structure of the stacking model in an embodiment of the present invention; Figure 3 A diagram showing changes in the second objective function during the optimization process of determining stacking model hyperparameters in an embodiment of the present invention; Figure 4 It is a temperature control comparison diagram of the stacking model of the present invention and the field model within 50m of the head; Figure 5It is a temperature control comparison diagram of the stacking model of the present invention and the field model within 100m of the head. DETAILED DESCRIPTION

[0009] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0010] like Figure 1 As shown, a method for calculating heat flux adaptive coefficient based on data drive of the present invention comprises: Step 1: Collect cooling process parameters, strip information parameters and cooling water temperature as input feature parameters, and use heat flux adaptive coefficient as output feature parameter. The input feature parameters and output feature parameters constitute the original data set.

[0011] The cooling process parameters include: final rolling measured temperature, intermediate measured temperature, coiling measured temperature, rolling speed, upper valve cooling mode, lower valve cooling mode, upper start valve position, lower start valve position and cooling direction. The strip information parameters include: mass percentage of chemical composition, strip width and strip thickness. The 9 cooling process parameters, 18 strip information parameters and cooling water temperature, which are the 28 main factors affecting heat exchange, are used as input feature parameters. The heat flux adaptive coefficient is used as the output feature parameter, i.e., the sample label.

[0012] The step 1 is specifically as follows: Step 1.1: Collect characteristic parameters of the strip from the hot rolling production site. Since the function of the present invention is mainly to predict the strip head, the data collection position should be as close to the strip head as possible, but the influence of the poor shape of the strip head on the measured value needs to be considered. Considering the relationship between the two, the characteristic parameter collection position of the strip is calculated according to the following formula: ; ; In the formula, l b is the starting position of data collection, i.e. the length from the head of the strip, in m; l c Δ is the distance from the pyrometer to the coiler before coiling, in m; b is a fixed constant; h is the strip thickness, in m; l e The end position of data collection refers to the length from the head of the strip, in meters; N is the number of sampling points; v is the running speed of the strip, in m / s; Δ t is the sampling period, in seconds.

[0013] In this embodiment, l c Δ is the distance from the pyrometer to the coiler before coiling, and its value is 26.22 m. b The value is 0.08. N The value is 50. t The value is 0.2 s. Assuming that the thickness of a strip is 0.004 m and the running speed is constant at 7 m / s, the distances from the sampling start position and end position to the strip head are: ; ; Step 1.2: Calculate the characteristic parameters of the strip according to the following formula: ; In the formula, x j For the strip j The value of a feature; x k,j For the k The sampling point j Features M is the total number of input feature parameters and output feature parameters, and its value is 29.

[0014] Step 1.3: Repeat steps 1.1 and 1.2 to collect cooling process parameters, information parameters and cooling water temperature of different steel strips, so as to obtain multiple samples and form an original data set.

[0015] In this embodiment, the number of samples in the original data set is 60,000.

[0016] Step 2: Clean the original data set and divide the cleaned data set into training set, test set and validation set in a ratio of 7:2:1. Specifically: Step 2.1: Since the data collected from the hot rolling production site contains a lot of noise data, it is necessary to perform preliminary data cleaning on the original data set to remove abnormal data that is obviously inconsistent with the actual situation. Specifically, the abnormal data is removed according to the following formula: ; In the formula, x max and x min are the upper and lower limits of the characteristic parameters respectively; x i,j For the i The sample j Features n is the number of samples.

[0017] The upper and lower limits of the characteristic parameters in this embodiment are shown in Table 1: Table 1 Characteristic parameter setting range

[0018] Step 2.2: The data cleaned in step 2.1 is further cleaned using the Laida criterion to remove samples of characteristic parameters that do not conform to the following formula, so that the distribution of each characteristic parameter in the original data set is more consistent with the normal distribution: ; In the formula, For the j The average value of the features, σ For the j The standard deviation of a feature.

[0019] Step 2.3: Divide the cleaned dataset into training set, test set, and validation set in a ratio of 7:2:1. The training set is used for stacking model training, the test set is used for performance evaluation of the stacking model, and the validation set is used for verification when determining the hyperparameters to be optimized during the training process.

[0020] In this embodiment, after step 2.1 and step 2.2, the number of samples in the original data set changes from 60,000 to 54,518, and then after being divided into a training set, a test set, and a validation set according to a ratio of 7:2:1, the number of samples in each data set is 38,163, 10,903, and 5,452, respectively.

[0021] Step 3: Construct a heat flux adaptive coefficient stacking model and use the chaos Harris Eagle optimization algorithm to obtain the optimal number and type of base learners, specifically: Step 3.1: Take extreme random tree regressor (ETR), random forest (RF), extreme gradient boosting (XGBoost), categorical gradient boosting (CatBoost), gradient boosted decision tree (GBDT), histogram gradient boosted regressor (HGBR), bagging, adaptive boosting algorithm (AdaBoost), lightweight gradient boosting machine (LightGBM) and decision tree regressor (DTR) as candidate base learners.

[0022] Step 3.2: Use linear regression (LR) as the meta-learner.

[0023] Step 3.3: The number and type of base learners are taken as the parameters to be optimized.

[0024] In this embodiment, when determining the optimal stacking model structure, the parameters to be optimized and their search space are shown in Table 2: Table 2 Parameters to be optimized and search space

[0025] Step 3.4: The following formula is used as the first objective function to be optimized: ; In the formula, num is the number of base learners; type is the type of each base learner; m is the number of samples in the validation set; and y i Respectively i The model's predicted values ​​and actual values ​​(labels) for samples.

[0026] Step 3.5: Taking the validation set as input and taking the minimum value of the first objective function as the goal, the chaotic Harris Eagle optimization algorithm is used to optimize the parameters to be optimized in step 3.3. The optimal parameter combination is the optimal structure of the heat flux adaptive coefficient stacking model.

[0027] In this embodiment, when determining the optimal structure of the stacking model, the value of the first objective function changes with the iteration process as follows: Figure 2 As shown in Table 3, the parameter combination corresponding to the minimum value of the first objective function is the optimal parameter. Table 3 Optimal parameter values

[0028] Step 4: Use the Chaos Harris Eagle Optimization Algorithm to optimize the hyperparameters of the obtained optimal base learner to obtain the optimal hyperparameter combination, specifically: Step 4.1: Determine the hyperparameters to be optimized and the optimization interval according to the type of the optimal base learner obtained in step 3.

[0029] In this embodiment, the base learners are ETR, RF, CatBoost, and Bagging, and the meta learner is LR. The key hyperparameters and search spaces of each learner are shown in Table 4: Table 4 Hyperparameters and search space

[0030] Step 4.2: Take the coefficient of determination as the second objective function to be optimized: ; In the formula, R 2 is the determination coefficient, the closer it is to 1, the better the model fit is; m is the number of samples in the validation set; , yi and Respectively i The model predicted value, the actual value, and the average of the actual values ​​for samples.

[0031] Step 4.3: To obtain the maximum value of the second objective function, the Chaos Harris Eagle Optimization Algorithm is used to optimize the hyperparameters in step 4.1 to obtain the optimal hyperparameter combination.

[0032] In this embodiment, during the hyperparameter optimization process, the value of the second objective function changes with the iteration process as follows: Figure 3 As shown in Table 5, the parameter combination corresponding to the minimum value of the second objective function is the optimal parameter. Table 5 Hyperparameters and search space

[0033] Step 5: Input the training set into the heat flux density adaptive coefficient stacking model for training to obtain the trained heat flux density adaptive coefficient stacking model.

[0034] Step 6: Use the test set to verify the generalization performance of the heat flux adaptive coefficient stacking model, specifically: The coefficient of determination R 2 , mean square error (MSE) and mean absolute error (MAS) are used as evaluation indicators to verify the generalization performance of the heat flux adaptive coefficient stacking model obtained in step 4.

[0035] ; ; In the formula, p is the number of samples in the test set.

[0036] Step 7: Collect the current strip information parameters and use the tested heat flux density adaptive coefficient stacking model to predict the current heat flux density adaptive coefficient of the strip.

[0037] In this embodiment, the prediction results of each base learner, the prediction results of the on-site case-based reasoning model, and the prediction results of the stacking model of the present invention are compared as shown in Table 6: Table 6 Model performance evaluation

[0038] After adopting the stacking model established by the present invention, the temperature control accuracy within the head 50m and 100m of the production line is as follows: Figure 4 and Figure 5As shown in the figure, it can be clearly seen that compared with the case-based reasoning method used in the production line before, the stacking model proposed in the present invention has a higher prediction accuracy, which greatly improves the control accuracy of the head temperature. Among them, 88% and 93% of the samples within 50m and 100m of the head have a control accuracy of ±20℃, which is 26% and 24% higher than before.

[0039] The above description is only a preferred embodiment of the present invention and is not intended to limit the concept of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A data-driven heat flux adaptive coefficient calculation method, characterized in that: include: Step 1: Collect cooling process parameters, strip information parameters and cooling water temperature as input characteristic parameters, and use heat flux adaptive coefficient as output characteristic parameter. The input characteristic parameters and output characteristic parameters constitute the original data set. Step 2: Clean the original data set and divide the cleaned data set into training set, test set and validation set in a ratio of 7:2:1; Step 3: Construct a heat flux adaptive coefficient stacking model and use the chaos Harris Eagle optimization algorithm to obtain the optimal number and type of base learners; Step 4: Use the Chaos Harris Eagle Optimization Algorithm to optimize the hyperparameters of the obtained optimal base learner to obtain the optimal hyperparameter combination; Step 5: Input the training set into the heat flux density adaptive coefficient stacking model for training to obtain the trained heat flux density adaptive coefficient stacking model; Step 6: Use the test set to verify the generalization performance of the heat flux adaptive coefficient stacking model; Step 7: Collect the current strip information parameters and use the tested heat flux density adaptive coefficient stacking model to predict the current heat flux density adaptive coefficient of the strip.

2. The method for calculating heat flux adaptive coefficient based on data drive according to claim 1, characterized in that: The cooling process parameters include: final rolling measured temperature, intermediate measured temperature, coiling measured temperature, rolling speed, upper valve cooling mode, lower valve cooling mode, upper start valve position, lower start valve position and cooling direction; The strip steel information parameters include: mass percentage of chemical composition, strip steel width and strip steel thickness.

3. The method for calculating heat flux adaptive coefficient based on data drive according to claim 1, characterized in that: The step 1 is specifically as follows: Step 1.1: Collect the characteristic parameters of the strip from the hot rolling production site, and calculate the collection position of the characteristic parameters of the strip according to the following formula: ； ； In the formula, l b is the starting position of data collection, i.e. the length from the head of the strip, in m; l c is the distance from the pyrometer to the coiler before coiling, in meters; Δ b is a fixed constant; h is the strip thickness, in m; l e The end position of data collection refers to the length from the head of the strip, in meters; N is the number of sampling points; v is the running speed of the strip, in m / s; Δ t is the sampling period, unit is s; Step 1.2: Calculate the characteristic parameters of the strip according to the following formula: ； In the formula, x j For strip steel j The value of a feature; x k,j For the k The sampling point j Features M is the total number of input feature parameters and output feature parameters; Step 1.3: Repeat steps 1.1 and 1.2 to collect cooling process parameters, information parameters and cooling water temperature of different steel strips, so as to obtain multiple samples and form an original data set.

4. The method for calculating heat flux adaptive coefficient based on data drive according to claim 1, characterized in that: The step 2 is specifically as follows: Step 2.1: Preliminarily eliminate abnormal data according to the following formula: ； In the formula, x max and x min are the upper and lower limits of the characteristic parameters respectively; x i,j For the i The sample j Features n is the number of samples; Step 2.2: The data cleaned in step 2.1 is further cleaned using the Laida criterion to remove samples of characteristic parameters that do not conform to the following formula, so that the distribution of each characteristic parameter in the original data set is more consistent with the normal distribution: ； In the formula, For the j The average value of the features, σ For the j The standard deviation of the feature; Step 2.3: Divide the cleaned dataset into training set, test set, and validation set in a ratio of 7:2:

1.

5. The method for calculating heat flux adaptive coefficient based on data drive according to claim 1, characterized in that: The step 3 is specifically as follows: Step 3.1: Construct a candidate pool of base learners: extreme random tree regressor, random forest, extreme gradient boosting, categorical gradient boosting, gradient boosted decision tree, histogram gradient boosting regressor, bagging, adaptive boosting algorithm, lightweight gradient boosting machine and decision tree regressor as candidate base learners; Step 3.2: Use linear regression as a meta-learner; Step 3.3: The number and type of base learners are taken as the parameters to be optimized; Step 3.4: The following formula is used as the first objective function to be optimized: ； In the formula, num is the number of base learners; type is the type of each base learner; m is the number of samples in the validation set; and y i Respectively i The model predicted value and actual value of samples; Step 3.5: Taking the validation set as input and taking the minimum value of the first objective function as the goal, the chaotic Harris Eagle optimization algorithm is used to optimize the parameters to be optimized in step 3.

3. The optimal parameter combination is the optimal structure of the heat flux adaptive coefficient stacking model.

6. The method for calculating heat flux adaptive coefficient based on data drive according to claim 1, characterized in that: The step 4 is specifically as follows: Step 4.1: Determine the hyperparameters to be optimized and the optimization interval according to the type of the optimal base learner obtained in step 3; Step 4.2: Take the coefficient of determination as the second objective function to be optimized: ； In the formula, R 2 is the coefficient of determination, the closer it is to 1, the better the model fit; m is the number of samples in the validation set; , y i and Respectively i The model predicted value, actual value and the average of the actual value of samples; Step 4.3: To obtain the maximum value of the second objective function, the Chaos Harris Eagle Optimization Algorithm is used to optimize the hyperparameters in step 4.1 to obtain the optimal hyperparameter combination.

7. The method for calculating heat flux adaptive coefficient based on data drive according to claim 1, characterized in that: The step 6 is specifically as follows: The coefficient of determination R 2 , mean square error (MSE) and mean absolute error (MAS) are used as evaluation indicators to verify the generalization performance of the heat flux adaptive coefficient stacking model obtained in step 5.

Images