Intelligent strip steel plate shape regulation and control method based on self-adaptive fusion model

Abstract

The intelligent strip steel plate shape regulation and control method based on the self-adaptive fusion model comprises the steps that 1, historical rolling production process data are collected and preprocessed; 2, constructing a rolling production process data set based on the preprocessed data; 3, plate shape quality classification is conducted according to the ratio of the strip steel convexity to the target thickness, and category labels are set; 4, establishing a self-adaptive strip steel outlet strip shape diagnosis model containing a plurality of classifiers based on the DS theory, and training the diagnosis model through the rolling production process data set; 5, inputting rolling production process data under a new rolling schedule into a classifier of the trained adaptive strip steel outlet strip shape diagnosis model to obtain a real-time classification prediction result of the strip steel, and if the prediction classification is under-convexity or over-convexity, executing the step 6; otherwise, the process parameters are not adjusted; and 6, according to a prediction result in the step 5, controlling ILQ to dynamically adjust and optimize process parameters based on an inverse linear quadratic form.

Description

Technical Field

[0001] The present invention belongs to the technical field of rolling control, and relates to an intelligent regulation method for strip shape based on an adaptive fusion model. Background Art

[0002] In the steel rolling process, strip crown is a key index to characterize product quality, and its stability and uniformity directly affect subsequent cold rolling and downstream processing technologies. Accurately predicting strip crown is crucial for improving rolling efficiency and product quality. Machine learning methods have been widely applied in rolling production due to their advantages in dealing with complex non-linear problems. However, due to the limitations of single machine learning algorithms, it is difficult to effectively integrate the advantages of different models, which limits the model accuracy and directly affects the model decision-making results. DS evidence theory has the ability to handle conflicts and uncertainties, and can fuse prediction results between different algorithms, thereby improving the credibility and consistency of predictions.

[0003] In order to improve the control accuracy of strip shape, domestic scholars have conducted in-depth research. Chinese Patent "CN116689506A Strip Shape Control Method" obtains the real-time strip shape value and the target strip shape value, and eliminates the deviation between the two by adjusting the roll inclination and the intermediate roll, improving the common rolling defects in the eighteen-high production line and enhancing the strip shape quality of strip rolling. Chinese Patent "CN117655118B Strip Shape Control Method and Device with Multi-Mode Fusion" constructs an integrated prediction model that fuses multiple intelligent algorithms, uses the method of error compensation to improve the model accuracy, and realizes the dynamic regulation of strip shape quality during the rolling process through crown feedback and flatness feedback, ensuring the strip shape quality of the entire length. Chinese Patent "CN101758084B Strip Shape Prediction Control Method with Model Adaptation" decomposes the strip shape pattern, establishes a strip shape control model using rolling force and incoming material crown, and dynamically corrects it according to historical data and real-time rolling parameters to eliminate transmission time delay, and adjusts the feedback controller in real time to achieve rapid dynamic control of strip shape.

[0004] The above methods have improved the strip shape quality control to a certain extent, but more rely on parameter setting and model fusion to optimize the control strategy, lacking a more refined feedback mechanism and working condition self-adaptability. In the actual production process, strip shape is affected by multiple dynamic factors, and traditional methods often control based on fixed parameters or simple models, making it difficult to respond to working condition fluctuations in real time. In addition, the existing strip shape control methods have not fully solved the problems of conflicts and uncertainties in data fusion, resulting in limited control accuracy and difficulty in fully adapting to complex rolling environments. Especially in the actual dynamic regulation process, the decision-making interpretability and feedback mechanism have not been effectively integrated, affecting the controllability and stability of the system, and ultimately resulting in limitations in the strip shape quality control effect. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a strip shape intelligent control method based on an adaptive fusion model, which formulates a variety of fusion strategies using the DS evidence theory to obtain a diagnosis result and decision logic with high credibility, and generates strip shape control feedback information to realize the intelligent control of the strip shape.

[0006] The present invention provides a strip shape intelligent control method based on an adaptive fusion model, comprising:

[0007] Step 1: Collect historical rolling production process data and perform preprocessing;

[0008] Step 2: Construct a rolling production process data set based on the preprocessed data;

[0009] Step 3: Classify the strip shape quality according to the ratio of the strip crown to the target thickness, including under-crown, qualified crown, and over-crown, and set category labels;

[0010] Step 4: Establish an adaptive strip exit shape diagnosis model containing multiple classifiers based on the DS theory, and train the diagnosis model with the rolling production process data set;

[0011] Step 5: Input the rolling process parameter data under the new rolling schedule into the classifier of the trained adaptive strip exit shape diagnosis model to obtain the real-time classification prediction result of the strip. If the predicted classification is under-crown or over-crown, execute Step 6; otherwise, do not adjust the process parameters;

[0012] Step 6: According to the prediction result of Step 5, dynamically adjust and optimize the process parameters based on the inverse linear quadratic (ILQ) control.

[0013] The strip shape intelligent control method based on an adaptive fusion model of the present invention has the following beneficial effects:

[0014] The regulation method of the present invention can effectively improve the intelligent regulation level of the strip rolling process through three major parts: data processing, prediction diagnosis, and dynamic process regulation. Compared with traditional methods, the present invention overcomes the problem of high precision requirements for parameter adjustment in traditional shape control by introducing multiple intelligent classifiers, DS evidence theory, and ILQ control. Specifically, first, multi-dimensional historical rolling process parameters are collected and processed, and shape classification criteria are set and category labels are set to provide high-precision shape prediction information. Subsequently, the DS evidence theory is used to fuse multi-source data (from multiple intelligent classifiers), effectively handling the uncertainty and conflict problems in the prediction process. Further, the SHAP model is used to clarify the decision logic, ensuring the interpretability and transparency of the regulation process and providing shape feedback information. Finally, the ILQ controller realizes the dynamic and precise regulation of the strip by real-time feedback and optimized adjustment of key parameters such as rolling force, bending roll force, and rolling speed through the feedback information. Compared with traditional control methods, the method of the present invention can adjust the control strategy in real time during the dynamic rolling process, significantly improving the stability and precision of the shape quality, optimizing the production efficiency, and reducing the energy consumption, thus significantly enhancing the intelligent and automated level of the strip production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flowchart of an intelligent strip shape regulation method based on an adaptive fusion model of the present invention;

[0016] Figure 2 is a waterfall chart of SHAP values in an embodiment of the present invention;

[0017] Figure 3 is a flowchart of dynamically adjusting and optimizing process parameters based on inverse linear quadratic form control ILQ in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The following will further describe in detail the specific embodiments of the present invention in conjunction with the drawings and embodiments. The following embodiments are used to illustrate the present invention but are not intended to limit the scope of the present invention.

[0019] As Figure 1 shown, an intelligent strip shape regulation method based on an adaptive fusion model of the present invention includes:

[0020] Step 1: Collect historical rolling production process data and perform preprocessing;

[0021] The rolling production process data includes: rolling force, bending roll force, roll shift amount, rolling speed, strip inlet temperature, strip inlet thickness, strip outlet temperature, strip outlet thickness, strip outlet width, and crown.

[0022] The preprocessing includes: data cleaning, missing value handling, outlier detection and handling to ensure the integrity and accuracy of the data.

[0023] Step 2: Construct a rolling production process dataset based on the preprocessed data and divide it into a training set and a test set in a ratio of 7:3.

[0024] In this embodiment, a large amount of representative data is collected from a certain 2160mm hot rolling production line. As shown in Table 1, the dataset includes a total of 2796 groups of samples.

[0025] Table 1. Rolling production process data

[0026]

[0027] Step 3: Classify the shape quality of the strip steel based on the ratio of strip steel crown to the target thickness, including under-crown, qualified crown, and over-crown, and set class labels, specifically:

[0028] When the ratio of strip steel crown to the target thickness is less than 0.8, it is defined as under-crown, and the class label is 0. When the ratio of strip steel crown to the target thickness is 0.8 - 1.8, it is defined as qualified crown, and the class label is 1. When the ratio of strip steel crown to the target thickness is greater than 1.8, it is defined as over-crown, and the class label is 2. In this embodiment, the target thickness is 3mm.

[0029] Step 4: Establish an adaptive strip steel exit shape diagnosis model containing multiple classifiers based on the DS theory and train the diagnosis model through the rolling production process dataset, specifically:

[0030] Step 4.1: Select multiple intelligent classifiers, divide the rolling production process dataset into a training set and a test set in a ratio of 7:3, train each classifier, and calculate the posterior probability and reliability factor of each group of samples based on each classifier according to the shape quality classification results, specifically:

[0031] Step 4.11: Select multiple intelligent classifiers, including Random Forest RF, Light Gradient Boosting Machine LightGBM, Extreme Gradient Boosting XGBoost, K-Nearest Neighbor KNN, Decision Tree DT, and Gradient Boosting Decision Tree GBDT.

[0032] During specific implementation, the 2796 groups of samples are divided into a training set and a test set in a ratio of 7:3, and finally 29-dimensional data is determined as the input of the classifier. As shown in Table 1, the strip steel crown is used as the output of the classifier.

[0033] Step 4.12: Take the rolling force, roll bending force, roll shifting amount, rolling speed, strip inlet temperature, strip inlet thickness, strip outlet temperature, strip outlet thickness, and strip outlet width as the inputs of the classifier, and the crown as the output of the classifier; Each training uses 80% of the randomly selected training set data, and the test set remains unchanged. Conduct 10 independent trainings and predictions for each classifier.

[0034] Step 4.13: For each group of samples, count the number of occurrences of each category in 10 predictions to calculate the posterior probability P(A i |X), and the calculation formula is:

[0035]

[0036] where i = 0, 1, 2, A i represents the category as i; represents the number of occurrences of category i in 10 predictions. For example, if category 0 appears 3 times, then P(A0|X) = 0.3.

[0037] Step 4.14: The reliability factor evaluates the performance of the classifier through the statistical information of the confusion matrix . The confusion matrix is an H×H matrix, where H represents the number of categories, and the form of the confusion matrix is as follows:

[0038]

[0039] where C ij represents the number of samples that are actually category j but predicted as category i. Calculate the correct recognition ability of the classifier when predicting a certain category through the following formula:

[0040]

[0041] where C ii represents the number of samples correctly recognized as category i by the classifier; represents the number of samples that are actually category i; Denote R i as Rf(A i ), as the reliability factor of the classifier.

[0042] Step 4.2: Calculate the mass function of each classifier according to the posterior probability and the reliability factor, use the DS evidence theory to fuse the mass functions of multiple classifiers, and select the category with the largest synthesized mass function as the final shape diagnosis result; If there is a deviation between the shape diagnosis result and the actual category, then add a fuzzy membership function to adjust the weights of each classifier in the synthesis process, update the mass functions of each classifier, and re-conduct the mass function fusion until the diagnosis result is consistent with the actual classification, and select the classifier in the combination as the final classifier of the adaptive strip outlet shape diagnosis model, specifically:

[0043] Step 4.21: Define the recognition framework Θ = {under-convexity, qualified convexity, over-convexity}, corresponding to the class labels 0, 1, and 2 respectively.

[0044] Step 4.22: Calculate the mass function of each classifier according to the posterior probability and the reliability factor:

[0045] m r (A i ) = P(A i |X) × Rf(A i )

[0046] where m r (A i ) represents the mass function that the sample belongs to the class i predicted by the r-th classifier.

[0047] Step 4.23: Combine the mass functions of multiple classifiers to obtain the final credibility. For the mass functions of two classifiers, the Dempster combination rule is:

[0048]

[0049] where A represents the combined class, B ∈ Θ, and C ∈ Θ.

[0050] Step 4.24: Combine multiple classifiers and calculate the combined mass function of each combination. Select the class with the largest combined mass function as the final strip shape diagnosis result, and select the classifier in the combination as the final classifier of the adaptive strip exit shape diagnosis model.

[0051] Step 4.25: If there is a deviation between the strip shape diagnosis result and the actual class, add a fuzzy membership function to adjust the weights of each classifier in the combination process, and update the mass function of each classifier in real time to m r (A i )':

[0052] m r (A i )' = μ(x) · m r (A i )

[0053] where μ(x) represents the membership degree of x:

[0054]

[0055] where x = P(A i|X), where a, b, and c represent fuzzy parameters, and a, b, c ∈ [0, 1] (a < b < c); the values of a, b, and c are adjusted according to the deviation between the prediction result and the actual category until the diagnostic result is consistent with the actual category.

[0056] Repeat step 4.23, and use the updated quality function for DS synthesis to obtain a new diagnostic result.

[0057] Step 5: Input the rolling production process data under the new rolling schedule into the classifier of the trained adaptive strip exit shape diagnostic model to obtain the real-time classification prediction result of the strip. If the predicted classification is under-crown or over-crown, execute step 6; otherwise, do not adjust the process parameters.

[0058] Step 6: Based on the prediction result of step 5, dynamically adjust and optimize the process parameters based on the inverse linear quadratic (ILQ) control, as Figure 3 shown, specifically:

[0059] Step 6.1: Input the rolling production process data under the new rolling schedule and the predicted crown output by the classifier of the diagnostic model into the SHAP model to obtain the SHAP values of each rolling production process data. As Figure 2 shown, the horizontal axis represents the target value, and the vertical axis represents the names and values of different input features input into the SHAP model. The figure shows the first 9 items of data output by the SHAP model, including: F6 roll bending force, F6 rolling force, F5 rolling speed, strip width, F6 roll shifting amount, F5 roll bending force, F4 rolling speed, F4 rolling force, and F5 roll shifting amount. In this embodiment, a SHAP waterfall plot at a certain moment is obtained. For example, the SHAP value of the F6 roll bending force is +2.59, indicating that the F6 roll bending force is the main influencing factor of the current strip shape, and an increase in this value will cause an increase in the strip shape deviation.

[0060] Step 6.2: Establish the ILQ system state equation to clarify the relationship between the state variables and the control input, providing a mathematical model for the design of the ILQ controller:

[0061]

[0062] Among them, X is the state variable, that is, the rolling production process data under the new rolling schedule; U is the control input, that is, the adjustment amount of the key rolling parameters. The key rolling parameters are the rolling production process data among the first 9 items output by the SHAP model except for the roll shifting amount and strip width. In this embodiment, the control input is the adjustment amounts of the F6 roll bending force, F6 rolling force, F5 rolling speed, F5 roll bending force, F4 rolling speed, and F4 rolling force. E is the control matrix, which reflects the influence of the control input on the system state. For the convenience of engineering implementation, it is considered that each control input only acts on the corresponding state variable, and it is set as the identity matrix. D is the system state matrix, which describes the dynamic relationship between state variables. The autocorrelation coefficient and cross-correlation coefficient of the state variables are obtained through regression analysis of historical data and are used as the elements on the diagonal and the remaining adjacent elements respectively.

[0063] Step 6.3: The goal of the ILQ controller is to minimize the deviation of the target variable, that is, to minimize the state deviation X T QX and the control cost U T RU, to ensure that the system achieves optimal performance under dynamic working conditions. The objective function J is defined as:

[0064]

[0065] Among them, Q is the state weight matrix, which is a diagonal matrix, and the elements on the diagonal are the SHAP values of the key rolling parameters, used to adjust the importance of each state variable; R is the control input weight matrix, used to adjust the control strength of each state variable. To prevent drastic adjustment, it is set as the identity matrix.

[0066] Transform the minimization problem of the objective function into the solution of the Riccati equation:

[0067] P = D T PD - D T PE(R + E T PE) -1 E T PD + Q

[0068] Calculate the ILQ feedback gain matrix K through the solution P of the Riccati equation:

[0069] K = (R + E T PE) -1 E T PD

[0070] Calculate the adjustment amount U of the rolling process parameters to achieve real-time adjustment of the system state:

[0071] U = -KX

[0072] In this embodiment, the known current rolling process parameters are as follows: the F6 roll bending force is 550 kN, the F6 rolling force is 9549.411 kN, the F5 rolling speed is 7.835 m / s, the strip width is 1276.348 mm, the F5 roll bending force is 480 kN, the F4 rolling speed is 5.621 m / s, and the F4 rolling force is 12069.873 kN. Then:

[0073]

[0074] According to the feedback information generated in step 4.1, construct Q and R as follows:

[0075]

[0076] The adjusted rolling parameters are as follows: the F6 roll bending force is 495 kN, the F6 rolling force is 8594.47 kN, the F5 rolling speed is 7.0515 m / s, the strip width is 1276.348 mm, the F5 roll bending force is 432 kN, the F4 rolling speed is 5.0589 m / s, and the F4 rolling force is 10862.89 kN.

[0077] The ILQ controller is connected to each actuator through the PLC and adjusts the actions of the U regulating mechanism. The actuators include hydraulic cylinders and main motors; the hydraulic cylinders are used to adjust the magnitudes of the rolling force and the roll bending force, and the main motors are used to adjust the magnitude of the rolling speed.

[0078] The above are only the preferred embodiments of the present invention and are not intended to limit the idea of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An intelligent shape control method for strip steel based on an adaptive fusion model, characterized in that Including: Step 1: Collect historical rolling production process data and perform preprocessing; Step 2: Based on the preprocessed data, construct a rolling production process dataset and divide it into a training set and a test set according to a ratio of 7:3; Step 3: Classify the strip shape quality by the ratio of strip crown to target thickness, including under-crown, qualified crown and over-crown, and set class labels; Step 4: Establish an adaptive strip exit shape diagnosis model containing multiple classifiers based on the DS theory, and train the diagnosis model through the rolling production process dataset; Step 5: Input the rolling production process data under the new rolling schedule into the classifier of the trained adaptive strip exit shape diagnosis model to obtain the real-time classification prediction result of the strip. If the predicted classification is under-crown or over-crown, then execute Step 6; otherwise, do not adjust the process parameters; Step 6: According to the real-time classification prediction result of Step 5, dynamically adjust and optimize the process parameters based on the inverse linear quadratic control ILQ.

2. The intelligent shape control method for strip steel based on the adaptive fusion model according to claim 1, characterized in that The rolling production process data includes: rolling force, bending roll force, roll shift amount, rolling speed, strip inlet temperature, strip inlet thickness, strip outlet temperature, strip outlet thickness, strip outlet width and crown.

3. The intelligent shape control method for strip steel based on the adaptive fusion model according to claim 1, wherein, The preprocessing in Step 1 includes: data cleaning, missing value processing, outlier detection and processing to ensure the integrity and accuracy of the data.

4. The strip shape intelligent control method based on the adaptive fusion model according to claim 1, characterized in that The specific content of Step 3 is: When the ratio of strip crown to target thickness is less than 0.8, it is defined as under-crown, and the class label is 0; When the ratio of strip crown to target thickness is 0.8 - 1.8, it is defined as qualified crown, and the class label is 1; When the ratio of strip crown to target thickness is greater than 1.8, it is defined as over-crown, and the class label is 2.

5. The intelligent strip shape control method based on the adaptive fusion model according to claim 1, characterized in that, The specific content of Step 4 is: Step 4.1: Select multiple intelligent classifiers, and divide the rolling production process dataset into a training set and a test set according to a ratio of 7:

3. Train each classifier, and calculate the posterior probability and reliability factor of each group of samples based on each classifier according to the strip shape quality classification result; Step 4.2: Calculate the mass function of each classifier according to the posterior probability and reliability factor, use the DS evidence theory to fuse the mass functions of multiple classifiers, and select the class with the largest mass function after synthesis as the final strip shape diagnosis result; If there is a deviation between the strip shape diagnosis result and the actual class, then add a fuzzy membership function to adjust the weights of each classifier in the synthesis process, update the mass function of each classifier, and re-fuse the mass function until the diagnosis result is consistent with the actual classification, and select the classifier in the combination as the final classifier of the adaptive strip exit shape diagnosis model.

6. The intelligent strip shape control method based on the adaptive fusion model according to claim 5, wherein, The specific content of Step 4.1 is: Step 4.11: Select multiple intelligent classifiers, including Random Forest (RF), Light Gradient Boosting Machine (LightGBM), Extreme Gradient Boosting (XGBoost), K-Nearest Neighbor (KNN), Decision Tree (DT) and Gradient Boosting Decision Tree (GBDT); Step 4.12: Take the rolling force, roll bending force, roll shifting amount, rolling speed, strip inlet temperature, strip inlet thickness, strip outlet temperature, strip outlet thickness, and strip outlet width as the inputs of the classifier, and the crown as the output of the classifier; Each training uses 80% of the training set data randomly selected, and the test set remains unchanged. Conduct 10 independent trainings and predictions for each classifier. Step 4.13: For each group of samples, count the number of times each category appears in 10 predictions to calculate the posterior probability P(A i |X), and the calculation formula is: where \(i = 0, 1, 2\), \(A\) i represents the category \(i\); represents the number of times the category \(i\) appears in 10 predictions; Step 4.14: The reliability factor evaluates the performance of the classifier through the statistical information of the confusion matrix where the confusion matrix is an H×H matrix, H representing the number of classes, and the form of the confusion matrix is as follows: Among them, C ij represents the number of samples that are actually of class j but predicted as class i. The correct recognition ability of the classifier when predicting a certain class is calculated by the following formula: Among them, C ii represents the number of samples correctly identified as class i by the classifier; represents the number of samples actually belonging to class i; Let R i be denoted as Rf(A i ), as the reliability factor of the classifier.

7. The strip shape intelligent control method based on the adaptive fusion model according to claim 6, wherein, The specific content of step 4.2 is as follows: Step 4.21: Define the recognition framework Θ = {under crown, qualified crown, over crown}, corresponding to the category labels 0, 1, and 2 respectively; Step 4.22: Calculate the mass functions of each classifier according to the posterior probability and reliability factor: m r (A i ) = P(A i |X) × Rf(A i ) where m r (A i ) represents the mass function that the sample belongs to the class i with the prediction result of the r-th classifier; Step 4.23: Synthesize the mass functions of multiple classifiers to obtain the final credibility. For the mass functions of two classifiers, the Dempster synthesis rule is: where A represents the synthesized category, B ∈ Θ, C ∈ Θ; Step 4.24: Combine multiple classifiers and calculate the synthesized mass function of each combination. Select the category with the largest synthesized mass function as the final strip shape diagnosis result, and select the classifier in the combination as the final classifier of the adaptive strip outlet strip shape diagnosis model; Step 4.25: If there is a deviation between the shape diagnosis result and the actual category, then add a fuzzy membership function to adjust the weights of each classifier during the synthesis process, and update the quality function of each classifier in real time to m r (A i )': m r (A i )' = μ(x)·m r (A i ) where μ(x) represents the membership degree of x: where x = P(A i |X), a, b, and c represent fuzzy parameters, a, b, c ∈ [0, 1] (a < b < c); the values of a, b, and c are adjusted according to the deviation between the prediction result and the actual category until the diagnostic result is consistent with the actual category; Repeat step 4.23, use the updated mass function for DS synthesis to obtain a new diagnosis result.

8. The intelligent shape control method for strip steel based on the adaptive fusion model according to claim 1, wherein The specific content of step 6 is as follows: Step 6.1: Input the rolling production process data under the new rolling schedule and the predicted crown output by the classifier of the diagnosis model into the SHAP model to obtain the SHAP values of each rolling production process data; Step 6.2: Establish the ILQ system state equation to clarify the relationship between the state variables and the control input, and provide a mathematical model for the design of the ILQ controller: where X is the state variable, that is, the rolling production process data under the new rolling schedule; U is the control input, that is, the adjustment amount of the key rolling parameters. The key rolling parameters are the rolling production process data except the roll shifting amount and strip width among the first 9 items output by the SHAP model; D is the system state matrix, describing the dynamic relationship between the state variables; E is the control matrix, describing the influence of the control variables on the system; Step 6.3: The goal of the ILQ controller is to minimize the deviation of the target variable, that is, to minimize the state deviation X T QX and the control cost U T RU, ensuring that the system achieves optimal performance under dynamically changing operating conditions. The objective function J is defined as: where Q is the state weight matrix, which is a diagonal matrix, and the elements on the diagonal are the SHAP values of the key rolling parameters, used to adjust the importance of each state variable; R is the control input weight matrix, used to balance the control energy consumption and system performance; Convert the minimization problem of the objective function into the solution of the Riccati equation: P = D T PD - D T PE(R + E T PE) -1 E T PD + Q Calculate the ILQ feedback gain matrix K through the solution P of the Riccati equation: K = (R + E T PE) -1 E T PD Calculate the adjustment amount U of the rolling process parameters to achieve real-time adjustment of the system state: U = -KX The ILQ controller is connected to each actuator through the PLC, and based on U, the actions of the actuators are adjusted. The actuators include hydraulic cylinders and main motors; The hydraulic cylinder is used to adjust the magnitude of the rolling force and roll bending force, and the main motor is used to adjust the magnitude of the rolling speed.

Images