An intelligent hot rolling process optimization method based on artificial labeling

Abstract

The present invention relates to an intelligent hot rolling process optimization method based on artificial labels, comprising performing labeling influencing factor analysis, specifying labeling judgment criteria, identifying labeling responsible persons, performing labeling process standard specifications, performing labeling data collection, data clarification and data pre-processing, constructing a machine learning solution, performing data training, inputting current equipment operating condition labels into an HMI, and optimizing process solution and application. The present invention utilizes a trained model to perform real-time online rolling force prediction, improves the rolling force prediction accuracy, fully considers the uncontrollable factors of the equipment, introduces the uncontrollable factors of the equipment into the machine learning algorithm through artificial labeling, is more in line with the actual situation, and makes the decision result more accurate, so that even when the training data is different from the latest state, better training results can be obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of hot-rolled strip steel, and in particular to an intelligent hot-rolling process optimization method based on artificial labels. Background Art

[0002] Hot-rolled strip production is an indispensable key process in the entire steel production process. From the perspective of the production process, hot-rolled strip production is a key link that connects the upstream and downstream. On the technical level, hot-rolled strip production involves multiple complex physical and chemical processes such as high-temperature heating, rolling deformation, and cooling control. The efficiency and quality of hot-rolled strip production also directly affect the economic benefits and market competitiveness of steel companies.

[0003] The core element of hot rolling production is the rolling mill, which provides the core capability for hot-rolled strip deformation: rolling force. Accurately predicting rolling force and optimizing rolling process parameters are crucial for ensuring strip quality and performance. Traditional rolling force calculation methods are typically based on theoretical models, but due to the complexity and uncertainty of the hot rolling process, these models often have large errors. Furthermore, as market requirements for strip quality and performance continue to increase, traditional process parameter adjustment methods based on experience and trial and error are no longer able to meet production needs.

[0004] Therefore, many researchers have proposed a variety of intelligent algorithms for predicting and calculating rolling force, such as the XGBoost algorithm used in CN110773580B and the random forest algorithm used in CN 111790762B. There are also many similar patents. These patents will perform data cleaning before modeling to obtain as much useful data as possible, and perform correlation analysis to obtain the influence weights of different influencing factors as much as possible. However, these patents do not distinguish the authenticity of the data when performing data analysis, especially for hot rolling mill systems composed of tens of thousands of spare parts and replacement parts. The production status of the system is constantly changing. For example, there is a resonance problem in the F1 frame during a certain data collection stage, and the rolling force of the frame is limited during production. However, these data are not marked as abnormal. Then these data will pollute the entire model training, and the model trained based on these data will not be able to guide the normal production process. Summary of the Invention

[0005] The present invention provides an intelligent hot rolling process optimization method based on artificial labels. During the historical data preparation stage, labeling influencing factor analysis is added, labeling judgment standards are formulated, labeling responsible persons are identified, and labeling process standard specification work is carried out. Then, a multi-objective intelligent optimization algorithm is used to construct a model and train data. Finally, the trained model is used to perform real-time online rolling force prediction, thereby improving the rolling force prediction accuracy, fully considering the uncontrollable factors of the equipment, being more in line with the actual situation, and making the decision results more accurate. Even when the training data is different from the latest state, better training results can be obtained.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] An intelligent hot rolling process optimization method based on manual labeling comprises the following steps:

[0008] S1. Label determination: Analyze factors affecting labeling, formulate labeling judgment criteria, identify the person responsible for labeling, and formulate labeling process standards and specifications;

[0009] S2. Data Collection: Collect hot-rolled strip production data for no less than three months and combine it with manual labeling to build a basic data platform;

[0010] S3. Data pre-processing: pre-process the collected production data;

[0011] S4. Build a machine learning solution: Select a machine learning solution with label scalability to classify the data and determine the training and validation datasets.

[0012] S5. Data training: Train the process problem data based on the machine learning algorithm to obtain the process model, and use the validation data set to verify the model results;

[0013] S6. Determine the current equipment working condition label: Based on the label annotation determined in S1, label the current equipment status;

[0014] S7, process solution when all equipment is currently operating normally: Use the machine learning model trained in S5 to solve the equipment operating condition labels in S6 and the production data to be produced to obtain the optimal process solution under the current equipment status.

[0015] Furthermore, the step S1 includes the following steps:

[0016] S11. Analyze various production failures that may occur in hot rolling mills, including roll eccentricity, roll surface defects, roll bearing failures, hydraulic failures, temperature detection failures, uneven heating in the heating furnace, roller deviation, plate shape defects, and automation control system failures;

[0017] S12. Identify each fault based on the magnitude of the factors affecting production during the production process to determine whether it affects the production process;

[0018] S13. Introduce the purpose of data collection to the on-site operators, and make a preliminary quantitative determination based on the impact of the data on the rolling force;

[0019] S14. Make examination papers for process problems that need to be labeled and standardize the operators;

[0020] S15. Provide standard training to operators to ensure data accuracy during the data collection stage.

[0021] Furthermore, step S2 includes the following steps:

[0022] S21. The operator determines the device status during the data collection phase;

[0023] S22. After collecting data daily, the data is manually reviewed to determine its accuracy to ensure that the manually labeled data is consistent with the actual production situation.

[0024] Furthermore, step S3 includes the following steps:

[0025] S31. Initial attribute screening: Based on prior knowledge and data missing conditions, select attributes for subsequent modeling from the original data.

[0026] S32. Abnormal value processing: When a positive attribute has a non-positive value abnormality, all non-positive values ​​of this type of attribute are replaced with null values;

[0027] S33. Negative value processing: For attributes with both positive and negative values, convert all attributes of this type into absolute values.

[0028] Furthermore, the step S4 uses random forest theory to build a model, including the following steps:

[0029] S41. Randomly sample training subsets to build decision trees:

[0030] Randomly extract multiple sample subsets D from the original training dataset D i (i=1,2,……,n, where n is the number of decision trees) to construct a decision tree. For the sample subset D i , randomly select features from the subset for node splitting, there are m features in total, k features are randomly selected each time Used to evaluate the splitting criteria; the splitting criteria are used to determine the optimal splitting point. When splitting a node, the features and splitting points that minimize the sum of the Gini index of the child nodes after the split are selected for splitting;

[0031] S42. Ensemble decision trees for prediction: After building multiple decision trees, random forests integrate the prediction results of these decision trees to make the final prediction. For classification problems, the voting method is used, and for regression problems, the averaging method is used.

[0032] S44. Evaluate feature importance:

[0033] The average impurity reduction method is used to measure feature importance. For each decision tree, the impurity reduction brought by each feature when the node is split is calculated, and then the average is taken for all decision trees.

[0034] Furthermore, the classification problem in step S42 is:

[0035] For a new sample x, each decision tree will give a category prediction result, and the final prediction category is the category with the most votes. That is, if there are n decision trees, for category j, if a tree predicts category j, then the probability that the sample x belongs to category j is expressed as

[0036] Furthermore, the regression problem in step S42 is:

[0037] For a new sample x, each decision tree gives a prediction value, and the final prediction value is the average of the prediction values ​​of all decision trees. That is, if there are n decision trees, the prediction value of each tree for the sample is y i (i=1,2,3,……,n,), the final prediction value is

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] 1) Use the trained model to perform real-time online rolling force prediction, improving rolling force prediction accuracy. When optimizing the hot rolling intelligent process, the uncontrollable factors of the equipment are fully considered. Through manual labeling, the uncontrollable factors of the equipment are introduced into the machine learning algorithm, which is more in line with the actual situation and the decision results are more accurate.

[0040] 2) When applying the model, the uncontrollable factors of the equipment are fully considered. Even if the training data is different from the latest status, good training results can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a flow chart of a method according to an embodiment of the present invention.

[0042] Figure 2 Schematic diagram of data clarification and outlier removal according to an embodiment of the present invention.

[0043] Figure 3Schematic diagram of data pre-processing and data equalization according to an embodiment of the present invention.

[0044] Figure 4 2 is a schematic diagram of the random forest importance index analysis according to an embodiment of the present invention.

[0045] Figure 5 This is a schematic diagram of the effect of optimizing the C40 mean value and the corresponding increase in the proportion of cold-rolled finished products less than 8um according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings:

[0047] See Figure 1 , is a flow chart of the method of the present invention. The present invention provides an intelligent hot rolling process optimization method based on artificial labels, comprising the following steps:

[0048] S1. Label determination: Analyze the factors affecting labeling, formulate labeling judgment standards, identify the person responsible for labeling, and implement labeling process standards and specifications; analyze various production failures that may be encountered by the hot rolling mill, determine the corresponding label categories for different failures, and introduce the purpose of data collection to the operators of the four shifts A, B, C, and D on site. Then, make a preliminary quantitative judgment based on the impact of the failure on the rolling force. The specific steps include the following:

[0049] S11. Analyze various production failures that may occur in hot rolling mills, such as roll eccentricity, roll surface defects, roll bearing failures, hydraulic failures, temperature detection failures, uneven heating of the heating furnace, roller deviation, plate shape defects, and automation control system failures;

[0050] S12. Identify each fault based on the magnitude of the factors influencing production during the production process to determine whether it affects the production process. For example, label the roll shifting function of the F1-F7 hot rolling mills as good or poor, and label the production stability of the rolling mills as good or poor.

[0051] S13. Explain the purpose of data collection to the operators of the four shifts A, B, C, and D on site, and make a preliminary quantitative determination based on the impact of the data on rolling force;

[0052] S14. Make examination papers for process problems that need to be labeled, and standardize the operators of the four shifts A, B, C, and D;

[0053] S15. Commend the operators of the four shifts A, B, C and D and explain that they strived for data accuracy during the data collection stage.

[0054] S2. Data Collection: Collect hot-rolled strip production data for no less than three months, including basic process data for heating furnaces, rough rolling, finishing rolling, layer cooling, coiling, etc., and combine it with manual labeling to build a basic data platform. The specific steps include the following:

[0055] S21. Operators from shifts A, B, C, and D determine the equipment status during the data collection phase. For example, shift A determines that the roller shifting function of the F4 stand is poor. To ensure normal production, they adjust the roller shifting capacity to other units. For another example, shift D determines that the resonance state of the F2 stand is poor. They shift the load distribution to the front and rear stands.

[0056] S22. After collecting data every day, the data must be manually reviewed to determine its accuracy to ensure that the manually labeled data is consistent with the actual production situation.

[0057] S3. Data pre-processing: Data pre-processing is performed on the collected production data, including data cleaning, data smoothing, and data standardization. Here, manual labeling retains both the original data and the labeled data. The steps include:

[0058] S31. Initial attribute screening: The original data contained a large number of attributes that were completely empty or all zero. Based on prior knowledge and data missing information, and relying on an expert database, we removed useless attributes from 2,270 attributes and screened out 144 attributes for subsequent modeling.

[0059] S32. Abnormal value processing: There are some attributes that must be positive values. If there are abnormal non-positive values, all non-positive values ​​of this type of attributes will be replaced with null values.

[0060] S33. Negative value processing: There are some attributes that have both positive and negative values. Convert all attributes of this type into absolute values.

[0061] S4. Build a machine learning solution: Select an appropriate machine learning solution based on the process optimization goal. The primary basis for selecting a solution is good label scalability, that is, when data outside the label appears, it can be expanded to a certain extent. The data is also classified and the training and validation data sets are determined. The following steps are included:

[0062] S41. Randomly sample training subsets to build decision trees:

[0063] Random forest randomly extracts multiple sample subsets D from the original training dataset D. i (i=1,2,...,n, where n is the number of decision trees) to construct a decision tree. The construction process of each decision tree is as follows:

[0064] For the sample subset D i, randomly select a part of the features from the subset for node splitting. Assuming there are m features in total, k features are randomly selected each time to evaluate the splitting criteria.

[0065] Use some splitting criteria, including information gain and Gini index, to determine the best splitting point. For example, for the Gini index, the calculation formula is: Where C is the number of categories, P i is the probability that the sample belongs to the i-th category. When splitting a node, the feature and splitting point that minimize the sum of the Gini index of the child nodes after the split are selected for splitting;

[0066] S42. Ensemble decision tree for prediction:

[0067] After building multiple decision trees, random forests make final predictions by integrating the prediction results of these decision trees. For classification problems, the voting method is used; for regression problems, the averaging method is used.

[0068] Classification problem: For a new sample x, each decision tree will give a category prediction result, and the final prediction category is the category with the most votes. That is, if there are n decision trees, for category j, if a tree predicts category j, then the probability that the sample x belongs to category j can be expressed as

[0069] Regression problem: For a new sample x, each decision tree gives a prediction value, and the final prediction value is the average of the prediction values ​​of all decision trees. That is, if there are n decision trees, the prediction value of each tree for the sample is y i (i=1,2,3,……,n,), the final prediction value is

[0070] S5. Data training: Train the data of process problems based on machine learning algorithms to obtain accurate process models, and use validation data sets to verify the model results.

[0071] S6. Based on the labeling determined in S1, label the current state of the device.

[0072] S7, process solution when all equipment is currently operating normally: Use the machine learning model trained in S5 to solve the equipment operating condition labels in S6 and the production data to be produced to obtain the optimal process solution under the current equipment status.

[0073] The following examples are implemented under the premise of the technical solution of the present invention, and provide detailed implementation methods and specific operating processes, but the scope of protection of the present invention is not limited to the following examples. The methods used in the following examples are conventional methods unless otherwise specified.

[0074] [Example]

[0075] In a steel plant producing hot-rolled silicon steel crown control, in order to reduce the actual control value of silicon steel crown C40, and thereby increase the proportion of cold-rolled finished products C15 ≤ 8um; an intelligent hot rolling process optimization method based on manual labeling specifically includes the following steps:

[0076] S1. Various factors affecting the convexity of hot-rolled silicon steel are sorted out to determine important influencing factors such as rolling force, load distribution, and proportional convexity distribution. Among them, there are uncontrollable factors such as roll eccentricity, roll surface defects, roll bearing failure, hydraulic failure, temperature detection failure, uneven heating of the heating furnace, roller deviation, plate shape defects, and automatic control system failure. These factors cannot be entered into the machine learning system in a digitized way, and they need to be manually labeled to introduce their influencing factors into the machine learning system.

[0077] S2. After quantifying factors such as roll shifting, rolling force, and motor power, corresponding labels are developed and the data collection phase begins. This phase requires consideration of both data accuracy and manual judgment accuracy. For example, if Team A determines that the roll shifting function of the F4 stand is abnormal, to ensure normal production, the roll shifting capacity is adjusted to other stands. At this time, the roll shifting capacity of the F4 stand will be labeled as poor.

[0078] After collecting data every day, the data must be manually reviewed to determine its accuracy to ensure that the manually labeled data is consistent with the actual production situation.

[0079] S3. Preprocess the collected production data, such as Figure 2 As shown in , some abnormal values ​​in the material are eliminated; Figure 3 The data shown are equalized.

[0080] S4. Construct a machine learning solution. Here we choose the random forest theory: Random forest randomly extracts multiple sample subsets D from the original training data set D. i (i=1,2,...,n, where n is the number of decision trees) to construct a decision tree. The construction process of each decision tree is as follows:

[0081] For the sample subset D i , randomly select a part of the features from the subset for node splitting. Assuming there are m features in total, k features are randomly selected each time to evaluate the splitting criteria.

[0082] Use some splitting criteria, including information gain and Gini index, to determine the best splitting point. For the Gini index, the calculation formula is: Where C is the number of categories, Pi is the probability that the sample belongs to the i-th category. When splitting a node, the feature and splitting point that minimize the sum of the Gini index of the child nodes after the split are selected for splitting;

[0083] After building multiple decision trees, random forests make final predictions by integrating the prediction results of these decision trees. For classification problems, the voting method is used; for regression problems, the averaging method is used.

[0084] Classification problem: For a new sample x, each decision tree will give a category prediction result, and the final prediction category is the category with the most votes. That is, if there are n decision trees, for category j, if a tree predicts category j, then the probability that the sample x belongs to category j can be expressed as

[0085] Regression problem: For a new sample x, each decision tree gives a prediction value, and the final prediction value is the average of the prediction values ​​of all decision trees. That is, if there are n decision trees, the prediction value of each tree for the sample is: y i (i=1,2,3,……,n,), the final prediction value is

[0086] S5. Use random forest theory to train the data and obtain the corresponding training model, as shown in the following example: Figure 4 The conclusions of the importance index analysis are shown.

[0087] S6. The operator first determines the data label and inputs it into the HMI interactive interface, and the model automatically reads the data from the interactive interface.

[0088] S7. When the model is ready to guide on-site production, the optimization goal is to minimize the convexity of each strip. The model trained in step S5 is applied to the site. The operator first judges the data label and inputs it into the HMI interactive interface. The model automatically reads the data from the interactive interface and performs optimization. The optimization result is as follows: Figure 5 As shown in the figure, the C40 control target of hot-rolled silicon steel gradually decreases, the smaller the better, and the corresponding proportion of cold-rolled finished products less than 8um gradually increases.

Claims

1. An intelligent hot rolling process optimization method based on artificial labels, characterized in that: The steps include: S1. Label determination: Analyze the factors affecting labeling, formulate labeling judgment standards, identify the person responsible for labeling, and implement labeling process standards and specifications. The specific steps include the following: S11. Analyze various production failures that may occur in hot rolling mills, including roll eccentricity, roll surface defects, roll bearing failures, hydraulic failures, temperature detection failures, uneven heating in the heating furnace, roller deviation, plate shape defects, and automation control system failures; S12. Identify each fault based on the magnitude of the factors affecting production during the production process to determine whether it affects the production process; S13. Introduce the purpose of data collection to the on-site operators, and make a preliminary quantitative determination based on the impact of the data on the rolling force; S14. Make examination papers for process problems that need to be labeled and standardize the operators; S15. Provide standard training to operators to ensure data accuracy during the data collection stage; S2. Data Collection: Collect hot-rolled strip production data for no less than three months and combine it with manual labeling to build a basic data platform. The specific steps include the following: S21. The operator determines the device status during the data collection phase; S22. After daily data collection, manually review the data to determine its accuracy and ensure that the manually labeled data is consistent with actual production conditions. S3. Data pre-processing: Data pre-processing is performed on the collected production data, specifically including the following steps: S31. Initial attribute screening: Based on prior knowledge and data missing conditions, select attributes for subsequent modeling from the original data; S32. Abnormal value processing: When a positive attribute has a non-positive value abnormality, all non-positive values ​​of this type of attribute are replaced with null values; S33, negative value processing: For attributes with both positive and negative values, convert all attributes of this type into absolute values; S4. Build a machine learning solution: Select a machine learning solution with label scalability to classify the data and determine the training and validation datasets. The specific steps include the following: S41. Randomly sample training subsets to build decision trees: Randomly extract multiple sample subsets D from the original training dataset D i (i=1,2,……,n, where n is the number of decision trees) to build a decision tree. For the sample subset D i , randomly select features from the subset for node splitting, there are m features in total, k features are randomly selected each time , used to evaluate the splitting criteria; the splitting criteria are used to determine the optimal splitting point. When splitting a node, the feature and splitting point that minimize the sum of the Gini index of the child nodes after the split are selected for splitting; S42. Ensemble decision trees for prediction: After building multiple decision trees, random forests integrate the prediction results of these decision trees to make the final prediction. For classification problems, the voting method is used, and for regression problems, the averaging method is used. S44. Evaluate feature importance: The average impurity reduction method is used to measure feature importance. For each decision tree, the impurity reduction brought by each feature when the node is split is calculated, and then the average is calculated for all decision trees. S5. Data training: Train the process problem data based on the machine learning algorithm to obtain the process model, and use the validation data set to verify the model results; S6. Determine the current equipment working condition label: Based on the label annotation determined in S1, label the current equipment status; S7, process solution when all equipment is currently operating normally: Use the machine learning model trained in S5 to solve the equipment operating condition labels in S6 and the production data to be produced to obtain the optimal process solution under the current equipment status.

2. The intelligent hot rolling process optimization method based on artificial labels according to claim 1 is characterized in that: The classification problem in step S42 is: For a new sample , each decision tree will give a category prediction result, and the final prediction category is the category with the most votes, that is, if there is A decision tree, for the category , if a tree predicts the category ,but The sample belongs to the category The probability of .

3. The intelligent hot rolling process optimization method based on artificial labels according to claim 1 is characterized in that: The regression problem in step S42 is: For a new sample , each decision tree gives a prediction value, and the final prediction value is the average of all decision tree prediction values. That is, if there is decision trees, each tree's prediction value for the sample is (i=1,2,3,……,n,), the final prediction value is .

Images