Intelligent steel transfer control method and system for thick plate production line

Abstract

The invention provides an intelligent steel transfer control method and system for a thick plate production line. The method comprises the following steps: acquiring standard data of a rough rolling link of the thick plate production line and acquiring a standard image in a steel transfer area through image acquisition equipment; inputting the standard image into the target detection model, and outputting feature data; inputting the standard data and the characteristic data into a rotating speed prediction generation model, and outputting a rotating speed control curve of the steel rotating roller way; and a steel rotating request signal in the rolling process state data is extracted, a control command and a rotating speed given instruction of a steel rotating roller bed are generated according to the rotating speed control curve, and the steel rotating roller bed is controlled to drive the intermediate billet to rotate according to the rotating speed control curve. The rotating speed control curve is generated by extracting the standard data and the characteristic data in the rolling process state data, the steering command and the rotating speed given command are generated based on the rotating speed control curve, the steel rotating roller way is driven to operate according to the commands, and the problems that manual steel rotating is low in precision, lagged in response and low in efficiency are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of thick plate production line control, in particular to a thick plate production line intelligent steel transfer control method and system. BACKGROUND

[0002] In the thick plate rolling production process of the steel industry, the rough rolling link is the key process that determines the size accuracy and production efficiency of thick plates, and the steel transfer operation as the core process step of the rough rolling link directly affects the spread width effect and production line continuity of subsequent rolling.

[0003] Specifically, each billet needs to undergo 2-4 times of steel transfer operation in the rough rolling process. The principle is to control the steering and speed of the two groups of rotating roller tables on the steel transfer roller table to drive the billet to rotate 90° and then send it into the rolling mill, so as to realize accurate spread width of the billet in different directions and meet the multi-dimensional size process requirements of thick plate products.

[0004] Currently, the steel transfer control of the rough rolling link of the thick plate production line mainly relies on the traditional manual operation scheme. When the secondary control system calculates and determines that the steel transfer pass arrives and the billet meets the preset steel transfer basic conditions on the rotating roller table, the operator needs to observe the billet position, posture and roller table running state in real time through the monitoring screen, and then manually operate the control handle to adjust the steering and speed of the rotating roller table to complete the 90° rotation of the billet. Among them, the manual operation mode requires the operator to continuously focus on the monitoring screen and judge the adjustment time and parameters according to experience, which has a response lag. Especially in the high-paced rolling production, the lagged operation is easy to cause the steel transfer time to be prolonged, reduce the overall production rhythm of the rough rolling link, and is difficult to adapt to the high-efficiency production demand of the thick plate production line. SUMMARY

[0005] The present application provides a thick plate production line intelligent steel transfer control method and system to solve the problem of poor manual steel transfer efficiency.

[0006] In one aspect, the present application provides a thick plate production line intelligent steel transfer control method, comprising:

[0007] Obtaining rolling data of the rough rolling link of the thick plate production line; the rolling data includes intermediate billet data and rolling process state data;

[0008] Performing correlation analysis and preprocessing on the rolling data to obtain standard data;

[0009] Obtaining image data of the intermediate billet in the steel transfer area through an image acquisition device, and performing image transformation and preprocessing on the image data to obtain a standard image;

[0010] Inputting the standard image into a target detection model to obtain feature data including posture information, position information, size information and angle information of the intermediate billet output by the target detection model;

[0011] Input the standard data and the feature data into the speed prediction generation model to obtain the speed control curve of the steel conveyor output by the speed prediction generation model;

[0012] Extract the steel transfer request signal from the rolling process status data, and generate a first control command based on the speed control curve. The first control command is used to control the steel transfer roller table to drive the intermediate billet to rotate according to the speed control curve.

[0013] Optionally, the rolling data is subjected to correlation analysis and preprocessing to obtain standard data, including:

[0014] The acquired intermediate billet data and the rolling process status data are associated and matched according to billet number and pass information to establish initial standard data that identifies the correspondence between the intermediate billet data and the rolling process status data.

[0015] The initial standard data is subjected to noise identification and anomaly removal, and the format is standardized and the accuracy is calibrated to obtain standard data.

[0016] Optionally, image data of the intermediate billet within the steel transfer area is acquired using an image acquisition device, and the image data is transformed and preprocessed to obtain a standard image, including:

[0017] Video image data of intermediate billets in the steel transfer area is acquired by image acquisition equipment, and video frames are extracted as image data to be processed.

[0018] The image transformation is performed on the steel transfer region in the image data to be processed so that the pixel size of the steel transfer region matches the physical size of the intermediate billet;

[0019] The image data to be processed is scaled, denoised, and contrast optimized to obtain a standard image.

[0020] Optionally, the step of the target detection model outputting feature data includes:

[0021] The intermediate blank in the standard image is identified to obtain the identification result;

[0022] Based on the identification results, the spatial coordinates of the intermediate billet in the steel transfer area are calculated to obtain the position information;

[0023] Calculate the length and width of the intermediate blank to obtain dimensional information;

[0024] The angle information is obtained by determining the angle between the placement direction of the intermediate blank and the preset reference direction through the angle calculation method.

[0025] By combining the position information and the angle information, the tilt and offset state of the intermediate billet is determined to obtain the attitude information;

[0026] The location information, size information, angle information, and attitude information are integrated to form feature data and then output.

[0027] Optionally, the step of generating the rotational speed control curve of the steel conveyor output by the rotational speed prediction model includes:

[0028] Based on the steel transfer data stored in the historical steel transfer database, the force balance and motion law of the intermediate billet during rotation are analyzed based on the standard data and the feature data, and the speed values ​​of the steel transfer roller table matched at different times during the steel transfer process are generated.

[0029] Based on the speed value, a speed control curve for the steel transfer roller conveyor is generated.

[0030] Optionally, it also includes:

[0031] In the speed prediction generation model, a safe steel-turning zone for the steel-turning roller conveyor is pre-defined. Combining the pusher position, intermediate billet size data in the standard data, and the position information in the feature data, it is determined whether the intermediate billet is in the safe steel-turning zone.

[0032] If the intermediate billet is not in the safe steel transfer zone, the position adjustment parameters are generated by the rotation speed prediction generation model to determine the adjustment direction and adjustment range of the steel transfer roller table.

[0033] Optionally, the steel transfer request signal is extracted from the rolling process status data, and a first control command is generated based on the speed control curve, including:

[0034] Based on the speed values ​​and speed change trends at each time point in the speed control curve, a steering control command for the steel conveyor is generated.

[0035] The speed control curve is split according to a time series to generate speed command corresponding to each time period; the speed command and the steering control command are the first control commands.

[0036] The first control command is sent to the drive device of the steel transfer roller conveyor, which drives the intermediate billet to rotate in a preset direction.

[0037] Optionally, it also includes:

[0038] If the intermediate billet deviates from the safe steel transfer area and / or the rotation angle of the intermediate billet is not within the preset range, the rotation speed command will be dynamically adjusted.

[0039] Optionally, it also includes:

[0040] If the rotation angle of the intermediate billet reaches the preset steel transfer completion angle range, a steel transfer completion signal is generated to control the drive device to stop the operation of the steel transfer roller conveyor.

[0041] On the other hand, this application provides an intelligent steel transfer control system for a thick plate production line, applied to the intelligent steel transfer control method for a thick plate production line described in the first aspect, comprising:

[0042] The data acquisition module is used to acquire rolling data from the roughing stage of the thick plate production line; the rolling data includes intermediate billet data and rolling process status data.

[0043] The data processing module is used to perform correlation analysis and preprocessing on the rolling data to obtain standard data;

[0044] Image data of intermediate billets in the steel transfer area are acquired using an image acquisition device, and the image data is transformed and preprocessed to obtain a standard image.

[0045] The target detection module is used to input the standard image into the target detection model to obtain feature data output by the target detection model, including the pose information, position information, size information and angle information of the intermediate blank;

[0046] The speed prediction generation module is used to input the standard data and the feature data into the speed prediction generation model to obtain the speed control curve of the steel conveyor output by the speed prediction generation model.

[0047] The execution drive module is used to extract the steel transfer request signal from the rolling process status data and generate a first control command according to the speed control curve. The first control command is used to control the steel transfer roller table to drive the intermediate billet to rotate according to the speed control curve.

[0048] As can be seen from the above technical solutions, this application provides an intelligent steel transfer control method and system for a thick plate production line. The method includes: acquiring rolling data of the roughing stage of the thick plate production line; the rolling data includes intermediate billet data and rolling process status data; performing correlation analysis and preprocessing on the rolling data to obtain standard data; acquiring image data of the intermediate billet in the steel transfer area through an image acquisition device, and performing image transformation and preprocessing on the image data to obtain a standard image; inputting the standard image to a target detection model, and outputting feature data including the attitude information, position information, size information, and angle information of the intermediate billet; inputting the standard data and the feature data to a speed prediction generation model, and outputting the speed control curve of the steel transfer roller table; extracting the steel transfer request signal from the rolling process status data, and generating control commands and speed setpoint instructions for the steel transfer roller table according to the speed control curve, and controlling the steel transfer roller table to drive the intermediate billet to rotate according to the speed control curve. This application solves the problems of low accuracy, slow response, and low efficiency of manual steel transfer by extracting the steel transfer request signal from the rolling process status data and judging the start-up conditions by combining the intermediate billet position and posture characteristics. Once the conditions are met, a steering command and a time-limited speed setting instruction are generated based on the speed control curve, and the roller conveyor is driven to run according to the instructions. Attached Figure Description

[0049] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 A flowchart of the intelligent steel transfer control method for thick plate production lines provided in this application embodiment;

[0051] Figure 2 This is a schematic diagram of the intelligent steel transfer control system for a thick plate production line provided in an embodiment of this application. Detailed Implementation

[0052] The embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described below do not represent all embodiments consistent with this application. They are merely examples of systems and methods consistent with some aspects of this application as detailed in the claims.

[0053] In the steel industry's production system, heavy plate production lines are specialized production lines dedicated to rolling thicker steel plates. They are widely used in high-end equipment manufacturing and infrastructure construction fields such as shipbuilding, bridges, engineering machinery, and pressure vessels, and have stringent requirements for the dimensional accuracy, mechanical properties, and surface quality of the steel plates. Heavy plate production lines typically encompass core processes such as billet heating, rough rolling, finish rolling, cooling, straightening, and shearing. Among these, the rough rolling stage is the key step that determines the initial rolling accuracy of the heavy plate and its compatibility with subsequent processes. The steel transfer operation is the core technological action in the rough rolling stage that enables the billet to widen in multiple directions and meet the dimensional requirements of different specifications of heavy plates.

[0054] From the perspective of the steel-turning process requirements in the roughing rolling stage, each steel billet needs to undergo 2-4 steel-turning operations during the roughing rolling process, according to the rolling process requirements. The purpose is to control the direction and speed of the two sets of rotating rollers on the steel-turning roller table, causing the steel billet to rotate 90° around the center before being fed into the rolling mill. This achieves alternating widening of the steel billet in both the longitudinal and transverse directions, ensuring that the final thick plate product meets the preset dimensional standards in both length and width. The accuracy and efficiency of the steel-turning operation directly affect the rolling rhythm and billet dimensional accuracy in the roughing rolling stage.

[0055] In heavy plate production lines, the roughing and billet transfer process has long relied on manual control. When the secondary control system calculates and determines that the transfer pass has been reached, and the billet meets the basic transfer conditions on the rotating conveyor, operators must observe the billet's position, posture, and the conveyor's operating status in real time via a monitoring screen. They then manually operate the control handle to adjust the rotation direction and speed of the rotating conveyor to complete the billet's 90° rotation. This manual control method not only relies on the operator's experience and judgment, resulting in inconsistent operational precision and lag, but also requires operators to maintain a high level of concentration for extended periods, leading to poor efficiency in manual billet transfer.

[0056] To solve the problem of low efficiency in manual steel transfer, such as Figure 1 As shown, some embodiments of this application provide an intelligent steel transfer control method for a thick plate production line, including:

[0057] S101. Obtain rolling data for the roughing stage of the thick plate production line.

[0058] The rolling data includes intermediate billet data and rolling process status data. Specifically, the intermediate billet data is the attribute information of the intermediate billet to be transferred in the roughing stage, including at least the inlet intermediate billet size, the outlet intermediate billet size, and the billet number. The intermediate billet data reflects the physical specifications and identification of the intermediate billet and is the basis for judging the suitability for transfer and calculating the rotation speed parameters.

[0059] The rolling process status data includes at least the pass information, roll gap and rolling speed, speed and forward / reverse rotation of the inlet and outlet rotating rollers, steel transfer request for the steel transfer pass, and pusher position. The rolling process status data reflects the production line operation status in real time, providing real-time operating condition support for the start judgment, position adjustment and speed control of the steel transfer process.

[0060] S102. Perform correlation analysis and preprocessing on the rolling data to obtain standard data;

[0061] Standard data is a collection of rolling data that has been cleaned, transformed, and standardized according to preset rules. Among these, correlation analysis is used to verify the completeness and consistency of the rolling data, ensuring the logical relationships between data are correct; preprocessing addresses data format, units, and outliers to eliminate data noise and improve data quality.

[0062] Specifically, the standard data for correlation analysis and preprocessing of rolling data includes:

[0063] The acquired intermediate billet data and rolling process status data are associated and matched according to billet number and pass information to establish the correspondence between the intermediate billet data and the rolling process status data, and the initial standard data is obtained.

[0064] Noise identification and anomaly removal are performed on the initial standard data, and the format is standardized and the accuracy is calibrated to obtain standard data.

[0065] Since the intermediate billet data and the rolling process status data record information from two dimensions, namely intermediate billet attributes and production line dynamic conditions, respectively, it is necessary to use billet number and pass information as association fields to match the two types of data for the same billet under the same rolling pass, so as to ensure that each set of data can correspond to a specific rolling stage of a specific billet, thereby forming initial standard data containing complete association logic.

[0066] Next, for any noise data that may exist in the initial standard data, such as abnormal values ​​of roller speed caused by equipment signal fluctuations, or abnormal data such as billet size detection deviations exceeding the allowable range of the process, the data is identified and removed through preset process thresholds and data fluctuation range rules to ensure the authenticity of the data.

[0067] In addition, the initial standard data after screening are standardized in format and calibrated in accuracy. For example, the size data and speed data output by different acquisition devices are standardized to preset numerical units and formats. The size data of intermediate billets are corrected according to the accuracy deviation of the detection equipment, and finally standard data with structural specifications and accurate data are obtained.

[0068] S103. Acquire image data of intermediate billets in the steel transfer area through image acquisition equipment, and perform image transformation and preprocessing on the image data to obtain a standard image;

[0069] Image transformation primarily involves rotating and scaling the image data to ensure that the intermediate blank in the image is in a preset standard position and size for easy and uniform processing. Preprocessing steps include noise reduction, contrast enhancement, and grayscale conversion to improve image quality, reduce interference factors, and enhance the accuracy of image recognition algorithms.

[0070] Specifically, image data of the intermediate billet within the steel transfer area is acquired using image acquisition equipment, and the image data undergoes image transformation and preprocessing to obtain a standard image, including:

[0071] Video image data of intermediate billets in the steel transfer area is acquired by image acquisition equipment, and video frames are extracted as image data to be processed.

[0072] Image transformation is performed on the steel transfer area in the image data to be processed, so that the pixel size of the steel transfer area matches the physical size of the intermediate billet;

[0073] The image data to be processed is scaled, denoised, and contrast optimized to obtain a standard image.

[0074] Among them, the image acquisition equipment can be industrial-grade image acquisition equipment, focusing on the steel transfer area where the steel transfer roller is located, and continuously acquiring dynamic video image data of the intermediate billet. Considering that the video data volume is large and the steel transfer control requires clear images at critical moments, the video frames with clear images that can completely present the outline and position of the intermediate billet are selected through frame extraction algorithm, and the video frames are determined as the image data to be processed.

[0075] Because there is a spatial mapping deviation between the pixel size of the original image and the actual physical size of the intermediate billet, the image portion of the image data corresponding to the steel transfer area needs to be transformed using perspective transformation and coordinate calibration algorithms. By using preset physical size calibration parameters, such as the pixel distance corresponding to the known roller spacing, the image pixel size of the steel transfer area after transformation can accurately correspond to the actual physical size of the intermediate billet, ensuring that subsequent size and position calculations based on the image can directly reflect the physical state.

[0076] Finally, to address issues such as inconsistent pixel scaling, noise caused by ambient light interference, and uneven brightness in the image data to be processed, an image scaling algorithm was used to adjust the image data to the size specifications required by the target detection model. Then, Gaussian filtering was used to eliminate image noise, and a contrast enhancement algorithm was used to optimize the image brightness levels and highlight the contour differences between the central blank and the background, ultimately resulting in a standard image with high clarity, matching size, and prominent features.

[0077] S104. Input the standard image into the target detection model to obtain the feature data output by the target detection model, including the intermediate billet's posture information, position information, size information, and angle information.

[0078] When a standard image is input into the target detection model, the model can parse key information from the image, including but not limited to the pose, position, size, and angle information of the intermediate blank. The feature data is output in a structured format for easy reading and use by subsequent processing modules.

[0079] Specifically, the steps for the object detection model to output feature data include:

[0080] The intermediate blank in the standard image is identified, and the identification result is obtained;

[0081] Based on the recognition results, the spatial coordinates of the intermediate billet within the steel transfer area are calculated to obtain the position information;

[0082] Calculate the length and width of the intermediate blank to obtain dimensional information;

[0083] The angle information is obtained by determining the angle between the placement direction of the intermediate billet and the preset reference direction through the angle calculation method.

[0084] By combining position and angle information, the tilt and offset state of the intermediate billet is determined, and the attitude information is obtained.

[0085] The location, size, angle, and attitude information are integrated to form feature data.

[0086] Specifically, the preprocessed standard image is input into the pre-trained target detection model. The target detection model uses deep learning to identify the contour, edge, and texture features of the intermediate blank in the standard image, locates the region of the intermediate blank in the image, and obtains the recognition result including the pixel range of the intermediate blank and key feature points.

[0087] Next, based on the recognition results and the pixel size-object size mapping relationship established in the image transformation stage, the three-dimensional spatial coordinates of the intermediate billet within the steel transfer area are calculated to obtain positional information. For example, the X and Y axis coordinates are taken as the origin of a fixed point on the steel transfer roller conveyor.

[0088] Simultaneously, based on the pixel length and width of the intermediate blank in the recognition results, and combined with the size mapping ratio, the actual physical length and width of the intermediate blank are calculated to generate size information.

[0089] Subsequently, the angle between the current placement direction of the intermediate billet and the preset reference direction is determined by the angle calculation algorithm built into the target detection model, thus obtaining the angle information.

[0090] By combining position and angle information, the degree of inclination and offset direction of the intermediate billet are comprehensively judged to form attitude information.

[0091] Finally, the calculated position, size, angle, and attitude information are integrated according to a preset data format to form feature data with complete structure and accurate parameters.

[0092] S105. Input standard data and feature data into the speed prediction generation model to obtain the speed control curve of the steel conveyor output by the speed prediction generation model.

[0093] The speed control curve is calculated based on standard and characteristic data and is used to guide the speed adjustment of the steel transfer rollers. The speed control curve ensures the stability and accuracy of the intermediate billet during the steel transfer process, avoiding problems such as tilting or shifting caused by improper speed, thereby improving the production efficiency and product quality of the thick plate production line.

[0094] Specifically, the steps for generating the speed control curve of the steel conveyor from the speed prediction model include:

[0095] Based on the steel transfer data stored in the historical steel transfer database, the force balance and motion law of the intermediate billet during rotation are analyzed based on standard data and feature data, and the speed values ​​of the steel transfer roller matching at different times during the steel transfer process are generated.

[0096] Based on the speed value, a speed control curve for the steel transfer roller conveyor is generated.

[0097] The rotation speed prediction generation model first uses historical data on steel transfer, including different steel grades, intermediate billet sizes, and rolling conditions, stored in the historical steel transfer database as a benchmark. It also imports standard and feature data, and uses built-in algorithms to analyze the force balance and motion patterns of the intermediate billet during rotation. For example, it analyzes the matching relationship between roller friction and intermediate billet inertial force, the linear velocity requirements for intermediate billet rotation at different angles, and combines historical and real-time data correlation analysis to calculate the required speed values ​​for the steel transfer rollers at different times during the steel transfer process. This ensures that the speed values ​​meet the requirements for stable intermediate billet rotation while also adapting to the characteristics of the current operating conditions.

[0098] The speed prediction generation model integrates the calculated speed values ​​for each time period into a time series, and eliminates abrupt changes in speed values ​​at different time periods through a smoothing interpolation algorithm, forming a continuous speed change curve of the steel transfer roller. The speed change curve of the steel transfer roller clearly shows the dynamic adjustment law of the roller speed over time during the entire steel transfer process, and is finally output as the speed control curve.

[0099] S106. Extract the steel transfer request signal from the rolling process status data, and generate the first control command based on the speed control curve.

[0100] The first control command is used to control the steel transfer roller conveyor to drive the intermediate billet to rotate according to the speed control curve.

[0101] Upon receiving a steel transfer request signal, the steel transfer execution phase begins. Based on the speed control curve, the speed value that the steel transfer roller conveyor should reach at each time point is calculated. Subsequently, the speed value is converted into specific control commands and speed setpoint instructions, which are sent to the drive unit of the steel transfer roller conveyor. After receiving the instructions, the drive unit immediately adjusts the speed of the roller conveyor to ensure that the roller conveyor can drive the intermediate billet to rotate strictly according to the speed control curve.

[0102] Specifically, the process involves extracting the steel transfer request signal from the rolling process status data, and generating control commands and speed setpoint instructions for the steel transfer roller conveyor based on the speed control curve. This control process then rotates the intermediate billet according to the speed control curve, including:

[0103] Based on the speed values ​​and speed change trends at each time point in the speed control curve, the steering control command for the steel conveyor is generated.

[0104] The speed control curve is split according to the time sequence to generate speed setpoint commands for each time period; the speed setpoint commands and steering control commands are the first control commands.

[0105] The first control command is sent to the drive device of the steel transfer roller conveyor, which drives the intermediate billet to rotate in the preset direction.

[0106] Specifically, after confirming the validity of the steel transfer request signal in the rolling process status data, the speed value and speed change trend corresponding to each time node are analyzed in combination with the speed control curve. If the speed value in the speed control curve is positive or changes according to the preset positive law, it is determined that the steel transfer roller should be driven to rotate in the forward direction. If the speed value is negative or changes according to the preset reverse law, it is determined that the steel transfer roller needs to rotate in the reverse direction. Based on this, a clear steel transfer roller direction control command is generated to ensure that the roller direction is consistent with the intermediate billet rotation requirements.

[0107] The speed control curve is segmented according to the time sequence. Based on the time nodes of the entire steel turning process, the rotation start period, the uniform speed rotation period, and the angle fine adjustment period are divided. The corresponding average speed or dynamic speed change value in each period is extracted and converted into the speed command for each period, so that the speed command for each period can accurately match the speed requirements in the curve.

[0108] Finally, the generated steering control command and the speed setting instructions for each time period are sent synchronously to the drive device of the steel transfer roller conveyor. After receiving the instructions, the drive device determines the rotation direction of the roller conveyor according to the steering control command, and adjusts the running speed of the roller conveyor according to the speed setting instructions for each time period, ultimately driving the intermediate billet along the preset rotation direction.

[0109] In some embodiments, the intelligent steel transfer control method for thick plate production lines provided in this application further includes:

[0110] In the speed prediction generation model, a safe steel-turning zone for the steel-turning roller is pre-defined. Combining the pusher position in the standard data, the intermediate billet size data, and the position information in the feature data, it is determined whether the intermediate billet is in the safe steel-turning zone.

[0111] If the intermediate billet is not in the safe steel transfer zone, the position adjustment parameters are generated by the speed prediction generation model to determine the adjustment direction and adjustment range of the steel transfer roller table.

[0112] Specifically, the rotation speed prediction model pre-defines a safe rotation zone based on the physical dimensions of the slab conveyor, the spatial requirements during slab rotation, and the layout of surrounding equipment. This ensures the slab does not interfere with equipment and maintains a stable rotation trajectory. Then, it calls upon standard data reflecting the positions of mill auxiliary equipment (push bed position data), slab size data characterizing the physical specifications of the slab, and position information reflecting the real-time spatial coordinates of the slab from feature data. Spatial correlation analysis is performed on the push bed position data, slab size data, and position information to determine whether the current center position and edge contour of the slab completely fall within the pre-defined safe rotation zone. If the slab is determined not to be within the safe rotation zone—for example, if the slab is biased towards one side of the conveyor and its edge exceeds the boundary of the safe zone—the rotation speed prediction model automatically calculates and generates position adjustment parameters based on the deviation distance between the current position of the slab and the center of the safe rotation zone, the slab size, and the push bed position constraints. The position adjustment parameters determine the direction of the steel transfer roller conveyor to rotate forward or backward by the positive or negative direction of the deviation distance. The adjustment range of the roller conveyor to be operated is determined by the magnitude of the deviation distance combined with the size ratio of the intermediate billet, so as to ensure that the position adjustment parameters can accurately drive the steel transfer roller conveyor to move the intermediate billet to the safe steel transfer area.

[0113] In some embodiments, the intelligent steel transfer control method for thick plate production lines provided in this application further includes:

[0114] If the intermediate billet deviates from the safe steel transfer area and / or the rotation angle of the intermediate billet is not within the preset range, the speed command will be dynamically adjusted.

[0115] Specifically, during the process of rotating the intermediate billet by the steel transfer rollers, image data of the intermediate billet is continuously acquired by the image acquisition equipment. Combined with the target detection model, the spatial position and rotation angle of the intermediate billet are monitored simultaneously to determine whether the intermediate billet deviates from the preset safe steel transfer area due to abnormal working conditions such as roller slippage or uneven force.

[0116] In addition, the system checks whether the current rotation angle of the intermediate billet is within the preset angle range that matches the steel transfer progress. If the intermediate billet is detected to deviate from the safe steel transfer area, or if the rotation angle does not reach the preset range, such as rotating too slowly and causing the angle to lag behind the planned progress, the speed command is dynamically adjusted based on real-time deviation data. For example, if the intermediate billet is biased towards the safe area, the speed of the corresponding side roller can be finely adjusted to correct its position; if the rotation angle is lagging, the roller speed is appropriately increased to speed up the rotation progress. Through real-time adjustments, the system ensures that the intermediate billet always rotates within the safe range at the expected pace, guaranteeing the accurate completion of the steel transfer operation.

[0117] In some embodiments, the intelligent steel transfer control method for thick plate production lines provided in this application further includes:

[0118] If the rotation angle of the intermediate billet reaches the preset steel transfer completion angle range, a steel transfer completion signal is generated to control the drive device to stop the operation of the steel transfer roller conveyor.

[0119] Specifically, during the steel transfer process, image data of the intermediate billet is continuously acquired through image acquisition equipment. Combined with the target detection model, the rotation angle of the intermediate billet is calculated and monitored in real time. The rotation angle is compared with the preset steel transfer completion angle range in real time. When the rotation angle of the intermediate billet falls into the preset completion angle range and the position is stable without deviation, a steel transfer completion signal is generated. The steel transfer completion signal will be sent to the drive device of the steel transfer roller conveyor.

[0120] After receiving the signal that the steel transfer is complete, the drive unit immediately executes a stop command to control the steel transfer roller conveyor to stop running, thereby ending the steel transfer operation and preparing for the intermediate billet to be fed into the rolling mill or enter the next process. At the same time, it avoids the intermediate billet from over-rotating or shifting position due to the continuous operation of the roller conveyor, ensuring the quality of steel transfer and the smoothness of the production line process.

[0121] In some embodiments, such as Figure 2 As shown, this application provides an intelligent steel transfer control system for a thick plate production line, applied to the intelligent steel transfer control method for a thick plate production line provided in the above embodiments, including:

[0122] The data acquisition module 210 is used to acquire rolling data of the roughing stage of the thick plate production line; the rolling data includes intermediate billet data and rolling process status data;

[0123] Data processing module 220 is used to perform correlation analysis and preprocessing on rolling data to obtain standard data;

[0124] Image data of intermediate billets in the steel transfer area are acquired using image acquisition equipment, and image transformation and preprocessing are performed on the image data to obtain a standard image;

[0125] The target detection module 230 is used to input a standard image into the target detection model and output feature data including the pose information, position information, size information and angle information of the intermediate blank;

[0126] The speed prediction generation module 240 is used to input standard data and feature data into the speed prediction generation model and output the speed control curve of the steel conveyor.

[0127] The execution drive module 250 is used to extract the steel transfer request signal from the rolling process status data, and generate control commands and speed setting instructions for the steel transfer roller table according to the speed control curve, so as to control the steel transfer roller table to drive the intermediate billet to rotate according to the speed control curve.

[0128] As can be seen from the above technical solutions, this application provides an intelligent steel transfer control method and system for a thick plate production line. The method includes: acquiring rolling data of the roughing stage of the thick plate production line; the rolling data includes intermediate billet data and rolling process status data; performing correlation analysis and preprocessing on the rolling data to obtain standard data; acquiring image data of the intermediate billet in the steel transfer area through an image acquisition device, and performing image transformation and preprocessing on the image data to obtain a standard image; inputting the standard image to a target detection model, and outputting feature data including the attitude information, position information, size information, and angle information of the intermediate billet; inputting the standard data and feature data to a speed prediction generation model, and outputting the speed control curve of the steel transfer roller; extracting the steel transfer request signal from the rolling process status data, and generating control commands and speed setpoint instructions for the steel transfer roller according to the speed control curve, controlling the steel transfer roller to drive the intermediate billet to rotate according to the speed control curve. This application solves the problems of low accuracy, slow response, and low efficiency of manual steel transfer by extracting the steel transfer request signal from the rolling process status data and judging the start-up conditions by combining the intermediate billet position and posture characteristics. Once the conditions are met, a steering command and a time-limited speed setting instruction are generated based on the speed control curve, and the roller conveyor is driven to run according to the instructions.

[0129] Similar parts between the embodiments provided in this application can be referred to mutually. The specific implementation methods provided above are only a few examples under the overall concept of this application and do not constitute a limitation on the scope of protection of this application. For those skilled in the art, any other implementation methods extended from the solution of this application without creative effort shall fall within the scope of protection of this application.

Claims

1. A method for intelligent steel transfer control in a thick plate production line, characterized in that, include: Obtain rolling data from the roughing stage of the thick plate production line; the rolling data includes intermediate billet data and rolling process status data; The rolling data is subjected to correlation analysis and preprocessing to obtain standard data; Image data of intermediate billets in the steel transfer area are acquired using an image acquisition device, and the image data is transformed and preprocessed to obtain a standard image. The standard image is input into the target detection model to obtain feature data output by the target detection model, including the pose information, position information, size information and angle information of the intermediate blank; Input the standard data and the feature data into the speed prediction generation model to obtain the speed control curve of the steel conveyor output by the speed prediction generation model; Extract the steel transfer request signal from the rolling process status data, and generate a first control command based on the speed control curve. The first control command is used to control the steel transfer roller table to drive the intermediate billet to rotate according to the speed control curve.

2. The intelligent steel transfer control method for thick plate production lines according to claim 1, characterized in that, The rolling data is subjected to correlation analysis and preprocessing to obtain standard data, including: The acquired intermediate billet data and the rolling process status data are associated and matched according to billet number and pass information to establish initial standard data that identifies the correspondence between the intermediate billet data and the rolling process status data. The initial standard data is subjected to noise identification and anomaly removal, and the format is standardized and the accuracy is calibrated to obtain standard data.

3. The intelligent steel transfer control method for thick plate production lines according to claim 1, characterized in that, Image data of the intermediate billet within the steel transfer area is acquired using an image acquisition device, and the image data is then transformed and preprocessed to obtain a standard image, including: Video image data of intermediate billets in the steel transfer area is acquired by image acquisition equipment, and video frames are extracted as image data to be processed. The image transformation is performed on the steel transfer region in the image data to be processed so that the pixel size of the steel transfer region matches the physical size of the intermediate billet; The image data to be processed is scaled, denoised, and contrast optimized to obtain a standard image.

4. The intelligent steel transfer control method for thick plate production lines according to claim 1, characterized in that, The steps for the target detection model to output feature data include: The intermediate blank in the standard image is identified to obtain the identification result; Based on the identification results, the spatial coordinates of the intermediate billet in the steel transfer area are calculated to obtain the position information; Calculate the length and width of the intermediate blank to obtain dimensional information; The angle information is obtained by determining the angle between the placement direction of the intermediate blank and the preset reference direction through the angle calculation method. By combining the position information and the angle information, the tilt and offset state of the intermediate billet is determined to obtain the attitude information; The location information, size information, angle information, and attitude information are integrated to form feature data and then output.

5. The intelligent steel transfer control method for thick plate production lines according to claim 1, characterized in that, The steps for generating the speed control curve of the steel conveyor output by the speed prediction model include: Based on the steel transfer data stored in the historical steel transfer database, the force balance and motion law of the intermediate billet during rotation are analyzed based on the standard data and the feature data, and the speed values ​​of the steel transfer roller table matched at different times during the steel transfer process are generated. Based on the speed value, a speed control curve for the steel transfer roller conveyor is generated.

6. The intelligent steel transfer control method for thick plate production lines according to claim 1, characterized in that, Also includes: In the speed prediction generation model, a safe steel-turning zone for the steel-turning roller conveyor is pre-defined. Combining the pusher position, intermediate billet size data in the standard data, and the position information in the feature data, it is determined whether the intermediate billet is in the safe steel-turning zone. If the intermediate billet is not in the safe steel transfer zone, the position adjustment parameters are generated by the rotation speed prediction generation model to determine the adjustment direction and adjustment range of the steel transfer roller table.

7. The intelligent steel transfer control method for thick plate production lines according to claim 6, characterized in that, Extract the steel transfer request signal from the rolling process status data, and generate a first control command based on the speed control curve, including: Based on the speed values ​​and speed change trends at each time point in the speed control curve, a steering control command for the steel conveyor is generated. The speed control curve is split according to a time series to generate speed command corresponding to each time period; the speed command and the steering control command are the first control commands. The first control command is sent to the drive device of the steel transfer roller conveyor, which drives the intermediate billet to rotate in a preset direction.

8. The intelligent steel transfer control method for thick plate production lines according to claim 7, characterized in that, Also includes: If the intermediate billet deviates from the safe steel transfer area and / or the rotation angle of the intermediate billet is not within the preset range, the rotation speed command will be dynamically adjusted.

9. The intelligent steel transfer control method for thick plate production lines according to claim 7, characterized in that, Also includes: If the rotation angle of the intermediate billet reaches the preset steel transfer completion angle range, a steel transfer completion signal is generated to control the drive device to stop the operation of the steel transfer roller conveyor.

10. A smart steel transfer control system for a thick plate production line, applied to the smart steel transfer control method for a thick plate production line as described in any one of claims 1-9, characterized in that, include: The data acquisition module is used to acquire rolling data from the roughing stage of the thick plate production line; the rolling data includes intermediate billet data and rolling process status data. The data processing module is used to perform correlation analysis and preprocessing on the rolling data to obtain standard data; Image data of intermediate billets in the steel transfer area are acquired using an image acquisition device, and the image data is transformed and preprocessed to obtain a standard image. The target detection module is used to input the standard image into the target detection model to obtain feature data output by the target detection model, including the pose information, position information, size information and angle information of the intermediate blank; The speed prediction generation module is used to input the standard data and the feature data into the speed prediction generation model to obtain the speed control curve of the steel conveyor output by the speed prediction generation model. The execution drive module is used to extract the steel transfer request signal from the rolling process status data and generate a first control command according to the speed control curve. The first control command is used to control the steel transfer roller table to drive the intermediate billet to rotate according to the speed control curve.

Images