Leveler control method, device, equipment and computer storage medium

Abstract

The application discloses a skin pass mill control method, device, equipment and computer storage medium, the method comprises the following steps: acquiring a calibration image of a strip steel, acquiring real-time flatness data of the strip steel by using a computer vision detection algorithm, acquiring historical data from a historical database of skin pass second communication and setting data of process parameters of the skin pass mill, determining a feedforward adjustment value for feedforward adjustment of the process parameters, and acquiring a feedback adjustment value for feedback adjustment of the process parameters, the application embodiment realizes the acquisition of real-time flatness data of the strip steel by combining the acquired calibration images before and after skin pass with the computer vision detection algorithm, the acquisition of historical data from the historical database and the setting data of process parameters of the skin pass mill, the calculation of the process parameter adjustment value of the skin pass mill in combination with the real-time flatness data, so as to realize the automatic control of the skin pass mill, reduce the probability of flatness defects of the strip steel in the skin pass process, and improve the surface quality of the strip steel after hot rolling.

Description

Technical Field

[0001] This application relates to the field of hot-rolled thin strip steel production technology, and in particular to a leveling machine control method, device, equipment, and computer storage medium. Background Technology

[0002] Cold rolling leveling is a rolling process that involves rolling recrystallized annealed strip steel with minimal plastic deformation (elongation typically 0.2%–3%) to eliminate yield plateaus, control flatness, and achieve the desired surface morphology. Leveling is an essential production step in both traditional and modern cold rolling processes. Since leveling is the final step that determines the flatness and mechanical properties of the finished strip steel, controlling the flatness of the leveling machine is of paramount importance for improving the quality of cold-rolled strip steel.

[0003] Currently, existing technologies for flatness and straightness control mainly rely on the quality parameters of the preceding process for prediction, and adjust the real-time flatness through the operator's human judgment. Therefore, human factors inevitably cause fluctuations in the final flatness, making it difficult to achieve precise control of flatness and accurately hit the target flatness at the end of the line. Summary of the Invention

[0004] This application provides a leveling machine control method, device, equipment, and computer storage medium, which can solve the problems of difficulty in controlling the straightness and low control accuracy of strip steel during leveling.

[0005] In a first aspect, embodiments of this application provide a leveling machine control method. The method includes: acquiring a calibration image of a strip steel, the calibration image including a first calibration image of the strip steel before leveling and a second calibration image after leveling; acquiring real-time flatness data of the strip steel using a computer vision detection algorithm based on the calibration image, the real-time flatness data including the first flatness data of the strip steel before leveling and the second flatness data of the strip steel after leveling, and temporarily storing the real-time flatness data in an online server; acquiring historical data and setting data of the leveling machine's process parameters from a historical database of the leveling secondary communication based on the first flatness data; determining a feedforward adjustment value for feedforward adjustment of the process parameters based on the historical data, setting data, and first flatness data, and obtaining a first process parameter based on the feedforward adjustment value; acquiring a feedback adjustment value for feedback adjustment of the process parameters based on the historical data, setting data, feedforward adjustment value, and second flatness data, and obtaining a second process parameter based on the feedback adjustment value; controlling the leveling machine to level the strip steel according to the second process parameter, and saving the second process parameter in the historical database.

[0006] In one possible implementation of this application embodiment, acquiring a calibration image of the strip, including a first calibration image of the strip before leveling and a second calibration image after leveling, includes: setting two camera groups on both sides of the leveling machine, the two camera groups including a first camera and a second camera, wherein the first camera is set above the strip and the second camera is set on one side of the strip; calibrating the production line environment for the first camera and the second camera; calibrating the width position and corrugation type of the strip for the first camera; and calibrating the corrugation height of the strip for the first camera.

[0007] In one possible implementation of this application embodiment, the real-time straightness data of the strip is obtained by using a computer vision detection algorithm based on the calibration image, including: obtaining the real-time wave type of the strip, which includes center wave, single-sided wave, double-sided wave and edge-center composite wave; and obtaining the real-time wave size of the strip, which includes wave position, wave width and wave height.

[0008] In one possible implementation of this application, real-time flatness data of the strip steel is obtained using a computer vision detection algorithm based on the calibration image. The computer vision detection algorithm includes: obtaining the current coil size information of the strip steel; generating multiple feature candidate boxes for the calibration image using a selective search algorithm; calculating feature information in the multiple feature candidate boxes through a convolutional layer; normalizing the feature information through a region of interest pooling layer; inputting the feature information into a fully connected layer and classifying the feature information using a normalization exponential function; and obtaining the defect information of the strip steel based on the current coil size information. The defect information includes defect type, defect location, and defect size and flatness.

[0009] In one possible implementation of this application embodiment, historical data includes coil information, historical process parameters, and feature values ​​of historical process parameters. Based on the first flatness data, historical data and the setting data of the leveling machine's process parameters are obtained from the historical database of the leveling secondary communication, including: obtaining coil information of the strip steel based on the first flatness data, the coil information including coil number, hot rolling time, historical leveling times, steel grade, steel type, coil width, coil width classification, coil thickness, coil thickness classification, coil weight, coil length, and hot-rolled flatness; obtaining historical process parameters of the leveling machine based on the coil information, the historical process parameters including uncoiling tension, coiling tension, rough straightening roll gap, fine straightening roll gap, bending roll force, rolling speed, roll horizontal tilt angle, and rolling force; and obtaining feature values ​​based on the historical process parameters, the feature values ​​including the number of coils produced, average value, median, maximum value, minimum value, and standard deviation of flatness.

[0010] In one possible implementation of this application embodiment, the set data includes a unit adjustment value and an adjustment threshold. Based on the first flatness data, historical data and the set data of the process parameters of the leveling machine are obtained from the historical database of the leveling secondary communication. The method further includes: obtaining a unit adjustment value based on the historical process parameters, the unit adjustment value including a unit feedforward adjustment value and a unit feedback adjustment value; and obtaining an adjustment threshold based on the unit adjustment value, the adjustment threshold including the maximum and minimum values ​​of the historical process parameters.

[0011] In one implementation of this application, a feedforward adjustment value for feedforward adjustment of process parameters is determined based on historical data, set data, and first flatness data. A first process parameter is obtained based on the feedforward adjustment value. The first flatness data includes: setting the historical process parameters for feedforward within a cycle, where the cycle includes cycle time and time point; acquiring the first flatness data at the time point, where the first flatness data includes a first wave type and a first wave size; obtaining a unit feedforward adjustment value from the set data based on the coil information and the first wave type; obtaining a feedforward calculation value for feedforward adjustment based on the historical process parameters, characteristic values, the unit feedforward adjustment value, and the first wave size; determining the feedforward adjustment value based on the adjustment threshold and the feedforward calculation value; and adjusting the historical process parameters based on the feedforward adjustment value to obtain the first process parameter.

[0012] In one possible implementation of this application embodiment, obtaining a feedback adjustment value for adjusting process parameters based on historical data, set data, feedforward adjustment value, and second flatness data, and obtaining a second process parameter based on the feedback adjustment value, includes: setting a first process parameter within a cycle, wherein the cycle includes cycle time and time point; obtaining second flatness data at the time point, wherein the second flatness data includes a second wave type and a second wave size; obtaining a unit feedback adjustment value from the set data based on the steel coil information and the second wave type; obtaining a feedback calculation value for feedback adjustment based on the first process parameter, feedforward adjustment value, unit feedback adjustment value, and second wave size; determining the feedback adjustment value based on the adjustment threshold and the feedback calculation value; and adjusting the first process parameter based on the feedback adjustment value to obtain the second process parameter.

[0013] Secondly, embodiments of this application provide a leveling machine control device, comprising: an image acquisition and calibration module for acquiring calibration images of a strip steel, the calibration images including a first calibration image of the strip steel before leveling and a second calibration image of the strip steel after leveling; a recognition algorithm module for acquiring real-time flatness data of the strip steel based on the calibration images using a computer vision detection algorithm, the real-time flatness data including the first flatness data of the strip steel before leveling and the second flatness data of the strip steel after leveling, and temporarily storing the real-time flatness data in an online server; and a data acquisition module for acquiring data from the leveling secondary communication based on the first flatness data. The system retrieves historical data and the setting data of the process parameters of the leveling machine from the historical database; the data adjustment module is used to determine the feedforward adjustment value for the process parameters based on the historical data, the setting data, and the first flatness data, and obtain the first process parameter based on the feedforward adjustment value; and, based on the historical data, the setting data, the feedforward adjustment value, and the second flatness data, obtains the feedback adjustment value for the process parameters, and obtains the second process parameter based on the feedback adjustment value; the control module controls the leveling machine to level the strip steel according to the second process parameter, and saves the second process parameter in the historical database.

[0014] Thirdly, embodiments of this application provide a control device, the device including: a processor and a memory storing computer program instructions;

[0015] When the processor executes computer program instructions, it implements the leveling machine control method as described in the first aspect.

[0016] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the flattening machine control method as described in the first aspect.

[0017] The leveling machine control method, apparatus, equipment, and computer storage medium of the embodiments of this application have at least the following beneficial effects:

[0018] In the leveling machine control method provided in this application embodiment, real-time flatness data of the strip steel is obtained by combining the calibration images before and after leveling with computer vision detection algorithms. Then, by combining historical data obtained from the historical database and the setting data of the leveling machine's process parameters with the real-time flatness data, the adjustment values ​​of the key leveling process parameters of the leveling machine are calculated, thereby realizing automatic control of the leveling machine, thereby reducing the probability of flatness defects in the strip steel during the leveling process and improving the surface quality of the strip steel after hot rolling. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the leveling system structure of a leveling machine control method provided in one embodiment of this application;

[0021] Figure 2 This is a schematic flowchart of the leveling machine control method provided in the embodiments of this application;

[0022] Figure 3 yes Figure 2 A flowchart illustrating the specific implementation method of S100 in China;

[0023] Figure 4 yes Figure 2 A flowchart illustrating the specific implementation of the S200;

[0024] Figure 5 yes Figure 2 A flowchart illustrating the specific implementation method of S300 in China;

[0025] Figure 6 yes Figure 2 A flowchart illustrating the specific implementation of the S400 in China;

[0026] Figure 7 This is a schematic diagram of the structure of a control device provided in one embodiment of this application;

[0027] Figure 8 This is one of the structural schematic diagrams of the image acquisition and calibration module of the control device provided in the embodiments of this application;

[0028] Figure 9 This is a second schematic diagram of the structure of the image acquisition and calibration module of the control device provided in the embodiments of this application;

[0029] Figure 10 This is a schematic diagram of the structure of a control device provided in one embodiment of this application. Detailed Implementation

[0030] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0031] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0032] The Faster R-CNN is a classic two-stage object detection algorithm. It uses a Region Proposal Network (RPN) to generate anchor points on the feature map and further generate bounding boxes, which are then fed into the Fast R-CNN network, reducing the computational cost of generating object candidate regions (Ren et al., 2015). Compared to its predecessor, Fast R-CNN, Faster R-CNN integrates the four steps of object detection (feature extraction, candidate region formation, object classification, and bounding box regression) into a single network framework, further improving the efficiency of object detection.

[0033] Faster R-CNN consists of three parts: the backbone feature extraction network, the region proposal network (RPN), and Fast R-CNN. During training, the image is first processed by the backbone feature extraction network to extract features, and then a series of convolution and pooling operations are used to obtain feature maps. In the Region Proposal Network (RPN), the Anchors Generator uses a 3×3 sliding window to slide across the feature map, generating 15 different types of anchors centered on each pixel at five scales (32×32, 64×64, 128×128, 256×256, 512×512) and three aspect ratios (1:2, 1:1, 2:1). Then, two 1×1 convolutions are used to obtain the corresponding target scores (foreground / background distinction) and bounding box regression parameters (for calculating lesion coordinates). The bounding box regression parameters fine-tune the anchors, and the anchors are sorted according to the target scores. The top 2000 anchors are then mapped back to the original image, and non-maximum suppression (NMS) is used to filter the anchors to obtain the proposed bounding boxes. Simultaneously, the RPN selects positive and negative samples for loss calculation by calculating whether the Intersection Over Union (IOU) between the anchor and the ground truth bounding box reaches an IOU threshold. Softmax is used to calculate the classification loss between the sample and the ground truth bounding box.

[0034] Currently, with the increasing demands for flatness in the automotive and construction machinery industries, more and more thin-gauge strips are incorporating flattening processes to ensure flatness after flattening. Although the flattening process can use a two-level flattening model to calculate the rolling force and bending force during the flattening process, the calculated rolling force and bending force have poor accuracy and are basically not used as a reference in the flattening process.

[0035] The inventors of this application considered that current flatness and straightness control mainly relies on the quality parameters of the preceding process for prediction, resulting in low accuracy of the flatness process model. Traditional online flatness control mainly relies on the operator's human judgment to adjust the real-time flatness. Human factors inevitably cause fluctuations in the final flatness, making it difficult to achieve precise flatness control and accurate hit of the offline flatness. In order to achieve high stability of flatness and straightness, it is necessary to improve the real-time detection and control capability of flatness.

[0036] To address the problems of the prior art, embodiments of this application provide a leveling machine control method, apparatus, device, and computer storage medium. The leveling machine control method provided in this application embodiment will be described first below.

[0037] Figure 2 A schematic flowchart of a leveling machine control method provided in one embodiment of this application is shown.

[0038] like Figure 2 As shown in the figure, this application provides a leveling machine control method, which may include the following steps:

[0039] S100. Obtain the calibration image of strip 50. The calibration image includes a first calibration image of strip 50 before leveling and a second calibration image after leveling.

[0040] Camera calibration refers to the process of solving for camera parameters. In image measurement and machine vision applications, in order to determine the relationship between the three-dimensional geometric position of a point on the surface of a spatial object and its corresponding point in the image, it is necessary to establish a geometric model of camera imaging. This geometric model is the camera parameters. Camera calibration can be a traditional camera calibration method, an active vision camera calibration method, a camera self-calibration method, a zero-loss camera calibration method, and other calibration methods.

[0041] In this embodiment of the application, by obtaining flatness calibration images of strip steel 50 before and after flattening by the flattening machine 10 through camera calibration, flatness problems that are difficult to detect by visual means alone can be detected, and the flatness problems are reflected by the calibration images, thereby improving the accuracy of improving the flatness of strip steel 50 by controlling the flattening machine 10.

[0042] S200. Based on the calibration image, a computer vision detection algorithm is used to obtain the real-time flatness data of the strip 50. The real-time flatness data includes the first flatness data of the strip 50 before flattening and the second flatness data of the strip 50 after flattening. The real-time flatness data is then temporarily stored in the online server.

[0043] After camera calibration of the real-time images of strip 50 before and after leveling, real-time calibration images are obtained. Then, the real-time straightness data of strip 50 is obtained through computer vision detection algorithm. The real-time straightness data includes the first straightness data of strip 50 before leveling and the second straightness data after leveling. The computer vision detection algorithm can be the Faster R-CNN object detection algorithm.

[0044] S300: Based on the first flatness data, retrieve historical data and the setting data of the process parameters of the leveling machine 10 from the historical database of the leveling secondary communication.

[0045] The historical database of the leveling secondary communication 40 stores all historical data and corresponding setting data of the process parameters of the leveling machine 10. In this embodiment, the corresponding historical data and the corresponding process parameter setting data are retrieved from the historical database using the first flatness data of the strip 50 before leveling, simplifying the control process of the leveling machine 10 and improving work efficiency. The process parameter setting values ​​for this leveling operation are obtained based on the real-time flatness data. Then, combined with the corresponding historical data, setting values, and first flatness data, the process parameters of the leveling machine 10 are adjusted to obtain feedforward adjustment values. The process parameters of the leveling machine 10 are pre-adjusted using the feedforward adjustment values ​​to improve the accuracy of the process parameters of the leveling machine 10, thereby improving the flatness of the strip 50 after leveling.

[0046] S400. Based on historical data, set data, and first flatness data, determine the feedforward adjustment value for feedforward adjustment of process parameters, and obtain the first process parameter based on the feedforward adjustment value.

[0047] Before the strip 50 is leveled, the first flatness data of the strip 50 calculated by the computer vision detection algorithm is combined with the retrieved historical data and set data to adjust the process parameters of the leveling machine 10. The flatness of the strip 50 after being leveled by the leveling machine 10 can be further improved through the feedforward adjustment. At the same time, the automatic control of the leveling machine 10 is realized by combining the computer vision detection algorithm with the feedforward adjustment, thereby improving the working efficiency of the leveling machine 10.

[0048] S500: Based on historical data, set data, feedforward adjustment value, and second flatness data, obtain feedback adjustment value for feedback adjustment of process parameters, and obtain the second process parameter based on the feedback adjustment value.

[0049] By combining the feedforward adjustment value with historical data, set data, and second flatness data, a feedback adjustment value for the first process parameter is obtained. This allows for real-time feedback adjustment of the first process parameter of the leveling machine 10 during the current leveling process. The second flatness data reflects the real-time flatness defects of the strip 50 after leveling. In this embodiment, the leveling machine 10 operates in two control stages. The first stage involves pre-adjusting the process parameters before leveling through feedforward adjustment to improve the flatness of the strip 50. The second stage involves the leveling machine 10 leveling the strip 50 under the first process parameter setting, obtaining the second flatness data of the strip 50, and further adjusting the first process parameter in real-time to address flatness defects after leveling. This further improves the accuracy of the process parameters and enhances the flatness of the strip 50 after leveling by the leveling machine 10, thus achieving automatic control of the leveling machine 10.

[0050] S600: Control the leveling machine 10 to level the strip 50 according to the second process parameters, and save the second process parameters in the historical database.

[0051] The specific implementation methods of each of the above steps will be described in detail below.

[0052] In some embodiments, see Figure 3 The above step S100 obtains the calibration image of the strip 50. The calibration image includes the first calibration image of the strip 50 before leveling and the second calibration image after leveling. Specifically, the following steps can be performed:

[0053] S101. Two camera groups 20 are set on both sides of the leveling machine 10. The two camera groups 20 include a first camera 20a and a second camera 20b. The first camera 20a is set above the strip 50, and the second camera 20b is set on one side of the strip 50.

[0054] S102. Calibrate the production line environment for the first camera 20a and the second camera 20b;

[0055] S103. The width position and wave type of the strip 50 of the first camera 20a are calibrated.

[0056] S104. Calibrate the wave height of the strip steel 50 for the first camera 20a.

[0057] In the embodiments of this application, such as Figure 8 as well as Figure 9 As shown, two camera groups 20 are arranged on both sides of the leveling machine 10. One camera group 20 is installed between the uncoiler 60 and the leveling machine 10, and the other camera group 20 is installed between the leveling machine 10 and the winding machine 70. Each camera group 20 includes a first camera 20a and a second camera 20b. The first camera 20a is located above the strip 50, and the second camera 20b is located on the side of the strip 50. The camera group 20 is used to obtain real-time video images of the strip 50. The camera group 20 can be set to extract images every 5 frames to obtain real-time image information of the strip 50 before and after leveling. By obtaining the real-time image information of the strip 50 before and after leveling, the real-time flatness defects of the strip 50 can be obtained, which facilitates timely feedforward and feedback adjustments to the process parameters set on the leveling machine 10, thereby improving the flatness of the strip 50 and realizing automated adjustment and control.

[0058] In some embodiments, step S200 above, which uses a computer vision detection algorithm to obtain real-time flatness data of the strip 50 based on the calibration image, specifically includes the following steps:

[0059] Obtain the real-time wave pattern type of strip steel 50, which includes medium wave, single-sided wave, double-sided wave, and combined edge and medium wave;

[0060] Obtain the real-time wave size of strip 50, which includes wave position, wave width, and wave height.

[0061] Computer vision detection algorithms are used to calculate the real-time straightness data of the calibrated strip 50 from its calibration image. This real-time straightness data includes center waves, single-sided waves, double-sided waves, and combined edge-center waves. Wave size includes wave position, wave width, and wave height. In other words, real-time data on the wave type, position, width, and height of the strip 50 are obtained. This real-time data allows for the identification of straightness defects in the strip 50, thereby improving the accuracy of feedforward and feedback adjustments.

[0062] In some embodiments, see Figure 4 In step S200 above, based on the calibration image, a computer vision detection algorithm is used to obtain the real-time straightness data of the strip 50. The computer vision detection algorithm may specifically include the following steps:

[0063] S201. Obtain the current coil size information of strip steel 50.

[0064] S202. Use a selective search algorithm to generate multiple feature candidate boxes for the calibration image;

[0065] S203. Feature information in multiple feature candidate boxes is obtained through convolutional layers;

[0066] S204. Normalize the feature information through the region of interest pooling layer;

[0067] S205. Input the feature information into the fully connected layer and classify the feature information using the normalized exponential function;

[0068] S206. Based on the current steel coil size information, obtain the defect information of strip 50. The defect information includes the defect type, defect location, and defect size.

[0069] The computer vision detection algorithm can employ the Faster R-CNN object detection algorithm. The steel coil size information includes coil number, width, thickness, and length. Obtaining the steel coil size information of the strip 50 currently being leveled helps the Faster R-CNN object detection algorithm obtain accurate calculation results. The Faster R-CNN object detection algorithm can then determine the defect location, size, and waviness type of the strip 50, facilitating the automatic adjustment of the leveling process parameters of the leveling machine 10.

[0070] In some embodiments, the historical data in step S300 above includes steel coil information, historical process parameters, and characteristic values ​​of historical process parameters; based on the first flatness data, historical data and the setting data of the process parameters of the leveling machine 10 are obtained from the historical database of the leveling secondary communication, specifically including the following steps:

[0071] Based on the first flatness data, obtain the coil information of strip 50. The coil information includes coil number, hot rolling time, historical flattening times, steel grade, steel class, coil width, coil width classification, coil thickness, coil thickness classification, coil weight, coil length, and hot rolling flatness.

[0072] Based on the steel coil information, the historical process parameters of the leveling mill 10 are obtained. The historical process parameters include uncoiling tension, coiling tension, coarse straightening roll gap, fine straightening roll gap, bending roll force, rolling speed, roll horizontal tilt angle, and rolling force.

[0073] Characteristic values ​​are obtained based on historical process parameters. These characteristic values ​​include the number of rolls produced, average value, median, maximum value, minimum value, and standard deviation.

[0074] The first flatness data of strip 50 is obtained by the Faster R-CNN object detection algorithm. Then, based on the first flatness data, the corresponding coil information, historical process parameters, and feature values ​​of historical process parameters are searched from the historical database. This avoids resetting the process parameters of strip 50 and improves the efficiency and accuracy of the flattening work. In this embodiment, the median, or MeD, is selected as the feature value of the historical data. i , where i represents uncoiling tension, winding tension, coarse straightening roll gap, fine straightening roll gap, bending roll force, rolling speed, roll horizontal tilt angle, and rolling force, respectively.

[0075] In some embodiments, the data set in step S300 above includes a unit adjustment value and an adjustment threshold. Based on the first flatness data, historical data and the setting data of the process parameters of the leveling machine 10 are obtained from the historical database of the leveling secondary communication. Obtaining the setting data specifically includes performing the following steps:

[0076] The unit adjustment value is obtained based on historical process parameters. The unit adjustment value includes the unit feedforward adjustment value and the unit feedback adjustment value.

[0077] The adjustment threshold is obtained based on the unit adjustment value. The adjustment threshold includes the maximum and minimum values ​​of historical process parameters.

[0078] In this embodiment of the application, the maximum value of the adjustment threshold is denoted as Upi, and the minimum value is Low. i .

[0079] In some embodiments, see Figure 5 In step 400 above, based on historical data, set data, and the first flatness data, a feedforward adjustment value is determined for the feedforward adjustment of the process parameters. The first process parameter is obtained based on the feedforward adjustment value. The feedforward setting specifically includes the following steps:

[0080] S401. Feedforward setting of historical process parameters within a cycle, wherein the cycle includes cycle time and time point.

[0081] S402. Obtain the first flatness data at the time point, wherein the first flatness data includes the first wave type and the first wave size.

[0082] S403. Based on the steel coil information and the first wave type, obtain the unit feedforward adjustment value in the set data.

[0083] S404. Based on historical process parameters, characteristic values, unit feedforward adjustment values, and the size of the first wave shape, obtain the feedforward calculation value for feedforward adjustment.

[0084] In this embodiment, the feedforward cycle time is set to 3 seconds, the current time point is denoted as T, and the feedforward calculation value FF0 is calculated based on the current leveling process parameters, eigenvalues, unit feedforward adjustment value, and the size of the first wave. i The feedforward calculated value FF0 i Satisfies the calculation formula:

[0085] FF0 i =Med i +UnitFF i ×ValueFF Equation (1)

[0086] Among them, MeD i For eigenvalues, UnitFF i ValueFF is the unit feedforward adjustment value, and ValueFF is the size of the first wave.

[0087] S405. Determine the feedforward adjustment value based on the adjustment threshold and the feedforward calculated value.

[0088] Based on the adjusted thresholds (Upi, Lowi) and the feedforward calculated value FF0 i The final feedforward adjustment value FF is obtained. i The feedforward adjustment value FF i The relation is satisfied as follows:

[0089]

[0090] Among them, Up i To adjust the maximum value of the threshold, Low i To adjust the minimum value of the threshold.

[0091] S406. Based on the feedforward adjustment value, adjust the historical process parameters to obtain the first process parameters.

[0092] The feedforward adjustment value is calculated using Equation 2. The feedforward adjustment value is used to set the historical process parameters of the leveling machine 10 before the strip 50 is leveled, so as to improve the flatness of the strip 50 after leveling. The feedforward adjustment value is obtained through historical data of process parameters, set data and first flatness data. It can make targeted real-time adjustments to the information of the strip 50 in this leveling operation, and realize automatic control.

[0093] In some embodiments, see Figure 6 The above step S500 obtains the feedback adjustment value for adjusting the process parameters based on historical data, set data, feedforward adjustment value, and second flatness data, and obtains the second process parameter based on the feedback adjustment value. Specifically, it includes performing the following steps:

[0094] S501. Feedback setting of the first process parameter within a cycle, wherein the cycle includes cycle time and time point.

[0095] S502. Obtain the second flatness data at the time point, wherein the second flatness data includes the second wave type and the second wave size.

[0096] S503. Based on the steel coil information and the second wave type, obtain the unit feedback adjustment value from the set data.

[0097] S504. Based on the first process parameter, the feedforward adjustment value, the unit feedback adjustment value, and the second wave size, obtain the feedback calculation value for feedback adjustment.

[0098] In this embodiment, the cycle time of the feedback adjustment is set to 3 seconds, the current time point is denoted as T, and the feedback calculation value FB0 is calculated based on the current leveling process parameters, the feedforward adjustment value obtained in the previous calculation, the unit feedback adjustment value, and the size of the second wave. i The feedback calculation value FB0 i Satisfies the calculation formula:

[0099] FB0 i =Med i +UnitFB i ×ValueFB formula (3)

[0100] Among them, Med i For eigenvalues, UnitFB i The unit feedback adjustment value is ValueFB, which represents the size of the second wave.

[0101] S505. Determine the feedback adjustment value based on the adjustment threshold and the feedback calculation value.

[0102] S506. Based on the feedback adjustment value, adjust the first process parameter to obtain the second process parameter.

[0103] Based on the adjusted thresholds (Upi, Lowi) and the feedback calculation value FB0 i The final feedback adjustment value FB is obtained. i The feedback adjustment value FB i The relation is satisfied as follows:

[0104]

[0105] The feedback adjustment value is calculated using Equation 4. This feedback adjustment value is used to set the leveling process parameters of the leveling machine 10 based on the second flatness data of the strip 50 after it has been leveled by the leveling machine 10. The feedback adjustment value is obtained by using historical data of the leveling process parameters, the previous feedforward adjustment value, and the second flatness data. Therefore, the leveling process parameters of the leveling machine 10 can be adjusted in real time within one cycle based on the real-time flatness of the strip 50 after leveling, making the leveling process parameters more in line with the flatness requirements of the strip 50 and further improving the flatness of the strip 50 after leveling.

[0106] Please see Figure 1 and Figure 7 This application provides a control device for a leveling machine 10, comprising: an image acquisition and calibration module A100, used to acquire calibration images of a strip 50, the calibration images including a first calibration image of the strip 50 before leveling and a second calibration image after leveling; a recognition algorithm module A200, used to acquire real-time flatness data of the strip 50 based on the calibration images using a computer vision detection algorithm, the real-time flatness data including the first flatness data of the strip 50 before leveling and the second flatness data of the strip 50 after leveling, and temporarily storing the real-time flatness data in an online server; and a data acquisition module A300, used to acquire data from the leveling machine based on the first flatness data. The secondary communication historical database retrieves historical data and the setting data of the process parameters of the leveling machine 10; the data adjustment module A400 is used to determine the feedforward adjustment value for feedforward adjustment of the process parameters based on the historical data, the setting data, and the first flatness data, and obtain the first process parameter based on the feedforward adjustment value; and, based on the historical data, the setting data, the feedforward adjustment value, and the second flatness data, obtain the feedback adjustment value for feedback adjustment of the process parameters, and obtain the second process parameter based on the feedback adjustment value; the control module A500 controls the leveling machine 10 to level the strip 50 according to the second process parameter, and saves the second process parameter in the historical database.

[0107] Figure 10 A schematic diagram of the hardware structure of the control device provided in an embodiment of this application is shown.

[0108] The control device may include a processor 301 and a memory 302 storing computer program instructions.

[0109] Specifically, the processor 301 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0110] Memory 302 may include mass storage for data or instructions. For example, and not limitingly, memory 302 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 302 may include removable or non-removable (or fixed) media. Where appropriate, memory 302 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 302 is non-volatile solid-state memory.

[0111] In a particular embodiment, memory 302 includes read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0112] Memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of this application.

[0113] The processor 301 implements any of the control methods described in the above embodiments by reading and executing computer program instructions stored in the memory 302.

[0114] In one example, the control device may also include a communication interface 303 and a bus 310. For example, Figure 3 As shown, the processor 301, memory 302, and communication interface 303 are connected through bus 310 and complete communication with each other.

[0115] The communication interface 303 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0116] Bus 310 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 310 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0117] Furthermore, in conjunction with the control methods described in the above embodiments, this application embodiment can provide a computer storage medium for implementation. This computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the control methods described in the above embodiments.

[0118] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0119] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0120] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0121] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0122] The above are merely specific embodiments of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for controlling a leveling machine, characterized in that the method... include: Two camera groups are set on both sides of the leveling machine. The two camera groups include a first camera and a second camera. The first camera is set above the strip, and the second camera is set on one side of the strip. The production line environment was calibrated for both the first and second cameras. The width, position, and wave type of the strip are calibrated for the first camera; The height of the corrugated strip of steel is calibrated for the first camera; Obtain calibration images of the strip steel, including a first calibration image of the strip steel before leveling and a second calibration image of the strip steel after leveling; Based on the calibration image, a computer vision detection algorithm is used to obtain the real-time flatness data of the strip. The real-time flatness data includes the first flatness data of the strip before flattening and the second flatness data of the strip after flattening. The real-time flatness data is then temporarily stored in an online server. Based on the first flatness data, historical data and the setting data of the leveling machine's process parameters are obtained from the historical database of the leveling secondary communication. The historical data includes steel coil information, historical process parameters, and characteristic values ​​of historical process parameters. The setting data includes unit adjustment values ​​and adjustment thresholds. Based on the historical data, the set data, and the first flatness data, a feedforward adjustment value is determined for the feedforward adjustment of the process parameters, and the first process parameter is obtained based on the feedforward adjustment value. The step of determining a feedforward adjustment value for adjusting the process parameters based on the historical data, the set data, and the first flatness data, and obtaining the first process parameter based on the feedforward adjustment value, includes: The unit adjustment value is obtained based on historical process parameters. The unit adjustment value includes the unit feedforward adjustment value and the unit feedback adjustment value. The adjustment threshold is obtained based on the unit adjustment value. The adjustment threshold includes the maximum and minimum values ​​of historical process parameters. The historical process parameters are fed forward within a cycle, where the cycle includes the cycle time and the time point. Obtain the first flatness data at a given time point, where the first flatness data includes the first wave type and the first wave size; Based on the information about the steel coil and the type of the first wave, the unit feedforward adjustment value is obtained from the set data; Based on historical process parameters, eigenvalues, unit feedforward adjustment values, and the size of the first wave, the calculated feedforward adjustment value is obtained, and the calculation formula is as follows: Where FF0i is the feedforward calculated value, and MeD i For eigenvalues, UnitFF i The unit feedforward adjustment value, ValueFF is the size of the first wave; The feedforward adjustment value is determined based on the adjustment threshold and the feedforward calculated value. The calculation formula is as follows: Where FFi is the feedforward adjustment value, Up i To adjust the maximum value of the threshold, Low i To adjust the minimum value of the threshold; The first process parameter is obtained by adjusting the historical process parameters based on the feedforward adjustment value. Based on historical data, set data, feedforward adjustment values, and second flatness data, obtain feedback adjustment values ​​for adjusting process parameters, and obtain the second process parameters based on the feedback adjustment values. The leveling machine is controlled to level the strip steel according to the second process parameters, and the second process parameters are saved in the historical database.

2. The leveling machine control method according to claim 1, characterized in that, Based on the calibration image, real-time straightness data of the strip steel is obtained using computer vision detection algorithms, including: Obtain the real-time wave pattern type of the strip steel. The real-time wave pattern types include medium wave, single-sided wave, double-sided wave, and combined edge and medium wave. Obtain the real-time wave size of the strip steel, which includes the wave position, wave width, and wave height.

3. The leveling machine control method according to claim 1, characterized in that, Based on the calibration image, real-time straightness data of the strip steel is obtained using a computer vision detection algorithm. The computer vision detection algorithm includes: Obtain the current coil size information of the strip steel; Multiple feature candidate boxes are generated for the calibration image using a selective search algorithm; Feature information from multiple feature candidate boxes is obtained through convolutional layers; Feature information is normalized using a region-of-interest pooling layer; The feature information is input into the fully connected layer, and the feature information is classified using the normalized exponential function; Based on the current steel coil size information, obtain the strip steel defect information, which includes the defect type, defect location, and defect size.

4. The leveling machine control method according to claim 1, characterized in that, Based on the first flatness data, historical data and the setting data of the leveling machine's process parameters are obtained from the historical database of the leveling secondary communication, including: Based on the first flatness data, obtain the steel coil information of the strip steel. The steel coil information includes steel coil number, hot rolling time, historical flattening times, steel grade, steel type, steel coil width, steel coil width classification, steel coil thickness, steel coil thickness classification, steel coil weight, steel coil length, and hot rolling flatness. Based on the steel coil information, obtain the historical process parameters of the leveling mill. The historical process parameters include uncoiling tension, coiling tension, coarse straightening roll gap, fine straightening roll gap, bending roll force, rolling speed, roll horizontal tilt angle, and rolling force. Characteristic values ​​are obtained based on historical process parameters. These characteristic values ​​include the number of rolls produced, average value, median, maximum value, minimum value, and standard deviation.

5. The leveling machine control method according to claim 1, characterized in that, Based on historical data, set data, feedforward adjustment values, and second flatness data, feedback adjustment values ​​are obtained to adjust the process parameters. The second process parameter is then derived from these feedback adjustment values, including: The first process parameter is set by feedback within a cycle, where the cycle includes the cycle time and the time point; Obtain the second flatness data at the time point, where the second flatness data includes the second wave type and the second wave size; Based on the information about the steel coil and the type of the second wave, the unit feedback adjustment value is obtained from the set data; Based on the first process parameter, the feedforward adjustment value, the unit feedback adjustment value, and the second wave size, the feedback calculation value of the feedback adjustment is obtained; The feedback adjustment value is determined based on the adjustment threshold and the feedback calculation value; Based on the feedback adjustment value, the first process parameter is adjusted to obtain the second process parameter.

6. A leveling machine control device, characterized in that, The apparatus is used to implement the leveling machine control method as described in any one of claims 1 to 5, and the apparatus comprises: The image acquisition and calibration module is used to acquire calibration images of the strip steel, including a first calibration image of the strip steel before leveling and a second calibration image of the strip steel after leveling. The recognition algorithm module is used to obtain real-time flatness data of the strip steel based on the calibration image using a computer vision detection algorithm. The real-time flatness data includes the first flatness data of the strip steel before flattening and the second flatness data of the strip steel after flattening. The real-time flatness data is temporarily stored in the online server. The data acquisition module is used to acquire historical data and the setting data of the process parameters of the leveling machine from the historical database of the leveling secondary communication based on the first flatness data. The data adjustment module is used to determine the feedforward adjustment value for feedforward adjustment of process parameters based on historical data, set data, and first flatness data, and to obtain the first process parameter based on the feedforward adjustment value; and to obtain the feedback adjustment value for feedback adjustment of process parameters based on historical data, set data, feedforward adjustment value, and second flatness data, and to obtain the second process parameter based on the feedback adjustment value. The control module controls the leveling machine to level the strip steel according to the second process parameters, and saves the second process parameters in the historical database.

7. A control device, characterized in that, The equipment includes: Processor and memory storing computer program instructions; When the processor executes computer program instructions, it implements the leveling machine control method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer program instructions, which, when executed by a processor, implement the leveling machine control method as described in any one of claims 1 to 5.

Images