A steel plate rolling deviation correction method, system, device and medium

Abstract

This application discloses a method, system, equipment, and medium for correcting steel plate rolling deviations, relating to the field of rolling production processes. The method includes: acquiring steel plate echo signal data from a rolling production line based on millimeter-wave radar, and filtering, analyzing, and sampling the steel plate echo signal data to obtain the detection distance and antenna reflection angle of multiple detection points; performing coordinate transformation calibration on the multiple detection points based on a preset three-dimensional rectangular coordinate system, obtaining the three-dimensional coordinates of the multiple detection points through the detection distance and the antenna reflection angle; calculating a first offset corresponding to a first offset point and a second offset corresponding to a second offset point based on the three-dimensional coordinates of each detection point; calculating a steel plate offset rate based on the first offset and the second offset, combined with a preset steel plate width, and correcting the movement direction of the steel plate during rolling based on the steel plate offset rate. This application can maintain high-precision steel plate rolling deviation correction in complex environments.

Description

Technical Field

[0001] This application relates to the field of rolling production processes, and in particular to a method, system, equipment and medium for correcting deviations during steel plate rolling. Background Technology

[0002] In the rolling and transport process of steel plate rolling production, factors such as incompatible equipment precision, uneven rolling and transport tension, or external disturbances can cause steel plate deviation. High-speed threading and prolonged rolling during rolling and transport increase the probability of deviation. Mild deviation can cause edge damage to the steel plate, requiring shutdown to replace the work rolls on the back stand and rework the lower rolls, impacting production efficiency. Severe deviation directly leads to scrap steel accidents, potentially damaging the mill guide devices, roller conveyors, and other equipment, requiring prolonged shutdowns for scrap steel cleanup, significantly reducing production efficiency. Furthermore, deviation easily leads to edge defects, uneven thickness, and other quality problems, affecting the yield and surface quality of the steel plate products. To achieve high-quality and stable rolling of steel plates, extend the service life of rolls, reduce production downtime, and ensure continuous and efficient production, it is necessary to correct deviation.

[0003] Currently, deviation correction in the rolling process generally relies on high-temperature resistant, vibration-proof, and fog-proof industrial inspection cameras to collect and identify the amount of deviation in images of the steel plate rolling process. Predictive models are then used to predict the deviation trend for high-precision correction. However, during the steel plate rolling process, to ensure quality and fully utilize equipment capacity, the steel plate is cooled by water spraying. The evaporation of cooling water on the hot steel plate generates a large amount of water vapor (water mist) and iron oxide scale, which severely affects the accuracy of camera-based visual recognition correction schemes. Therefore, how to maintain high-precision deviation correction in complex environments such as water vapor (water mist) to ensure the accuracy and efficiency of steel plate rolling remains a pressing problem to be solved in existing technologies. Summary of the Invention

[0004] This application provides a method, system, equipment, and medium for correcting deviations during steel plate rolling, in order to solve the technical problem that existing technologies cannot maintain high-precision deviation correction during steel plate rolling in complex environments.

[0005] According to a first aspect of the embodiments of this application, a method for correcting deviations during steel plate rolling is provided, comprising:

[0006] Based on millimeter-wave radar, the echo signal data of steel plates from the rolling production line is acquired, and the echo signal data of steel plates from the rolling production line is filtered, analyzed and sampled to obtain the detection distance and corresponding antenna reflection angle of multiple detection points;

[0007] Based on a preset three-dimensional rectangular coordinate system, coordinate transformation calibration is performed on multiple detection points, and the three-dimensional coordinates of multiple detection points are obtained through the detection distance and the antenna reflection angle;

[0008] Based on the three-dimensional coordinates of each detection point, calculate the first offset corresponding to the first offset point and the second offset corresponding to the second offset point; wherein, the first offset point and the second offset point are the two detection points that are farthest apart in the vertical direction of the three-dimensional rectangular coordinate system.

[0009] Based on the first offset and the second offset, and combined with the preset steel plate width, the steel plate offset rate is calculated, and the movement direction of the steel plate during the rolling process is corrected according to the steel plate offset rate.

[0010] This application first acquires echo signal data of steel plates from the rolling production line using millimeter-wave radar. This avoids interference from water mist and iron oxide scale generated by cooling water evaporation, as is the case with existing camera-based visual recognition and correction schemes, thus improving the accuracy of steel plate rolling correction under complex conditions. Then, multiple detection points are acquired through filtering, analysis, and sampling. These detection points are then converted and calibrated to a three-dimensional Cartesian coordinate system. This allows for accurate acquisition of multiple detection points and coordinate system conversion and calibration, reducing measurement errors and providing a precise data foundation for subsequent calculations of steel plate offset, thereby ensuring the accuracy of steel plate rolling correction. Next, the first and second offsets of the first and second offset points, which are furthest apart along the vertical axis in the three-dimensional Cartesian coordinate system, are calculated. Based on the physical meaning of the three-dimensional coordinates, accurate offset detection can be achieved through simple calculations. This ensures the accuracy of steel plate rolling correction when the steel plate offset rate is calculated in conjunction with the set width of the steel plate and the movement direction of the steel plate during rolling is corrected.

[0011] In some embodiments of this application, the step of filtering, analyzing, and sampling the echo signal data of the steel plate from the rolling production line to obtain the detection distance and corresponding antenna reflection angle of multiple detection points specifically includes:

[0012] According to the preset detection distance range, the echo signal data of the steel plate from the rolling production line is filtered and focused. During the filtering process, signal data within the detection distance range is extracted from the echo signal data of the steel plate from the rolling production line, and the extracted signal data is smoothed to obtain a first time-domain signal; wherein, the detection distance range is determined based on the water mist diffusion distance.

[0013] Estimate the first echo component in the first time domain signal, and obtain the second time domain signal based on the difference between the first time domain signal and the first echo component; wherein the first echo component is generated by static scattering of water mist and reflection from a stationary object;

[0014] In the slow time dimension of the second time domain signal, a Fourier transform is performed on the second time domain signal to obtain a first frequency domain signal. The second echo component in the first frequency domain signal is estimated and filtered out by a preset filter to obtain a second frequency domain signal. The second echo component is generated by dynamic scattering of water mist.

[0015] The second frequency domain signal is subjected to interference filtering and target detection based on the constant false alarm rate algorithm to obtain a potential target point cloud. The potential target point cloud is then sampled for effective points to determine multiple detection points, and the detection distance and antenna reflection angle of the multiple detection points are obtained.

[0016] This application first filters signal data within the detection range and smooths it to obtain a first time-domain signal. The detection range is determined based on the water mist diffusion distance, which can reduce echo signals severely interfered with by the environment and initially focus on the effective signal range. Then, it estimates the first echo component generated by static scattering of water mist and reflection from stationary objects, and obtains a second time-domain signal by difference, which can filter out the interference echo signal generated by static scattering of water mist and reflection from stationary objects. Then, it performs a Fourier transform in the slow time dimension and filters out the second echo component generated by dynamic scattering of water mist to obtain a second frequency-domain signal, which can filter out the interference clutter signal generated by low-speed micro-motion of water mist. Finally, it uses a constant false alarm rate algorithm for interference filtering and target detection, which can adaptively filter out residual micro-motion interference on the noise substrate, so that when sampling effective points, multiple accurate detection points and their detection distances and antenna reflection angles are obtained.

[0017] In some embodiments of this application, the step of performing coordinate transformation calibration on multiple detection points based on a preset three-dimensional Cartesian coordinate system, and obtaining the three-dimensional coordinates of the multiple detection points through the detection distance and the antenna reflection angle, specifically includes:

[0018] Based on the three-dimensional rectangular coordinate system, a coordinate system transformation model is constructed; wherein, the three-dimensional rectangular coordinate system takes the vertical projection point of the central radar on the roller as the origin, the direction of movement of the steel plate on the roller as the positive direction of the horizontal axis, the counterclockwise vertical direction of the direction of movement of the steel plate on the roller as the positive direction of the vertical axis, and the direction from the plane where the roller is located to the plane where the central radar is located as the positive direction of the vertical axis.

[0019] According to the coordinate system transformation model, coordinate transformation and spatial position calibration are performed on multiple detection points. The detection distance and antenna reflection angle of each detection point are transformed and calculated into the three-dimensional coordinates of the corresponding detection point in the three-dimensional rectangular coordinate system, thereby obtaining the three-dimensional coordinates of multiple detection points.

[0020] This application first constructs a three-dimensional rectangular coordinate system based on the positional relationship between the radar, steel plate, and roller conveyor, then builds a coordinate system transformation model, and performs coordinate transformation and spatial position calibration on multiple detection points according to the coordinate system transformation model. This can accurately transform the coordinate representation of multiple detection points from the original coordinate system to the three-dimensional coordinate representation in this three-dimensional rectangular coordinate system, reduce measurement errors, and provide an accurate data basis for subsequent calculation of steel plate offset.

[0021] In some embodiments of this application, the coordinate system transformation model is specifically as follows:

[0022] ;

[0023] in, pitch angle around the longitudinal axis Roll angle around the horizontal axis and yaw angle around the vertical axis The size of the structure is The rotation matrix, The original three-dimensional coordinates of the detection point are calculated based on the detection distance and antenna reflection angle. The three-dimensional coordinates of the probe point in a three-dimensional rectangular coordinate system. This is the translation calibration vector.

[0024] This application calculates the original three-dimensional coordinates by using the detection distance of the detection point and the antenna reflection angle, and combines them with the rotation matrix and translation calibration vector to obtain the three-dimensional coordinates in the three-dimensional rectangular coordinate system. This enables accurate coordinate transformation and spatial position calibration, reduces measurement errors, and provides a precise data basis for subsequent calculation of steel plate offset.

[0025] In some embodiments of this application, calculating the first offset corresponding to the first offset point and the second offset corresponding to the second offset point based on the three-dimensional coordinates of each of the detection points specifically includes:

[0026] Based on the three-dimensional coordinates of each detection point, calculate the absolute value of the difference between the ordinates of each pair of detection points, and take the two detection points corresponding to the largest absolute value as the first offset point and the second offset point.

[0027] Using the origin of the three-dimensional rectangular coordinate system as the reference point, the Euclidean distances between the first offset point and the second offset point and the reference point are calculated respectively to obtain the corresponding first offset and second offset.

[0028] This application first determines the two detection points with the farthest vertical coordinates as the first offset point and the second offset point based on the three-dimensional coordinates of each detection point. Then, the first offset and the second offset are obtained by the Euclidean distance with the reference point. Based on the physical meaning of the three-dimensional coordinates, the offset can be accurately detected by simple calculation. While simplifying the offset calculation, the accuracy of the offset calculation is guaranteed.

[0029] In some embodiments of this application, the step of calculating the steel plate offset rate based on the first offset and the second offset, combined with a preset steel plate width, and correcting the movement direction of the steel plate during rolling based on the steel plate offset rate specifically includes:

[0030] Calculate the offset difference between the first offset and the second offset, and take the absolute value of the ratio of the offset difference to the set width of the steel plate as the steel plate offset rate;

[0031] Based on the relationship between the steel plate offset rate and the preset warning threshold and shutdown threshold, a steel plate correction operation is determined, and the movement direction of the steel plate during the rolling process is corrected according to the steel plate correction operation; wherein, the warning threshold is less than the shutdown threshold; the warning threshold and the shutdown threshold are determined based on the maximum allowable deviation amount.

[0032] This application first calculates the offset difference between the first offset and the second offset, then obtains the steel plate offset rate by combining the set width of the steel plate, and then determines the steel plate correction operation and performs correction by using the warning threshold and the shutdown threshold determined according to the maximum allowable deviation amount. This allows for accurate definition and calculation of the steel plate offset rate, thereby accurately determining the steel plate correction operation and performing high-precision correction.

[0033] In some embodiments of this application, determining the steel plate correction operation based on the numerical relationship between the steel plate offset rate and preset warning thresholds and shutdown thresholds, and correcting the movement direction of the steel plate during the rolling process according to the steel plate correction operation, specifically includes:

[0034] If the steel plate offset rate is greater than the warning threshold but not greater than the shutdown threshold, the steel plate offset rate is uploaded to the frame adjustment system so that the frame adjustment system selects the roll gap with the corresponding concavity according to the steel plate offset rate for steel plate rolling, and completes the correction of the steel plate rolling according to the roll gap selected by the frame adjustment system.

[0035] When the steel plate offset rate is between the warning threshold and the shutdown threshold, this application uploads the steel plate offset rate to the frame adjustment system so that the frame adjustment system can perform steel plate correction operation, thereby achieving accurate and high-precision steel plate correction.

[0036] According to a second aspect of the embodiments of this application, a steel plate rolling correction system is provided, including an echo filtering sampling module, a coordinate transformation calibration module, a steel plate offset calculation module, and a steel plate rolling correction module;

[0037] The echo filtering and sampling module is used to acquire the echo signal data of the steel plate in the rolling production line based on millimeter-wave radar, and to filter, analyze and sample the echo signal data of the steel plate in the rolling production line to obtain the detection distance and corresponding antenna reflection angle of multiple detection points.

[0038] The coordinate transformation calibration module is used to perform coordinate transformation calibration on multiple detection points based on a preset three-dimensional rectangular coordinate system, and obtain the three-dimensional coordinates of multiple detection points through the detection distance and the antenna reflection angle;

[0039] The steel plate offset calculation module is used to calculate the first offset corresponding to the first offset point and the second offset corresponding to the second offset point based on the three-dimensional coordinates of each detection point; wherein, the first offset point and the second offset point are the two detection points that are farthest apart in the vertical axis direction of the three-dimensional rectangular coordinate system.

[0040] The steel plate rolling correction module is used to calculate the steel plate offset rate based on the first offset and the second offset, combined with the preset steel plate width, and to correct the movement direction of the steel plate during the rolling process based on the steel plate offset rate.

[0041] In some embodiments of this application, the echo filtering sampling module includes a signal screening and smoothing unit, a signal estimation and filtering unit, a signal transformation and filtering unit, and a signal detection and sampling unit;

[0042] The signal filtering and smoothing unit is used to filter and focus the echo signal data of the steel plate from the rolling production line according to a preset detection distance range. During the filtering process, signal data within the detection distance range in the echo signal data of the steel plate from the rolling production line is extracted, and the extracted signal data is smoothed to obtain a first time-domain signal. The detection distance range is determined based on the water mist diffusion distance.

[0043] The signal estimation and filtering unit is used to estimate the first echo component in the first time domain signal and obtain the second time domain signal based on the difference between the first time domain signal and the first echo component; wherein the first echo component is generated by static scattering of water mist and reflection from a stationary object;

[0044] The signal transformation and filtering unit is used to perform a Fourier transform on the second time domain signal in the slow time dimension to obtain a first frequency domain signal, and to estimate and filter out the second echo component in the first frequency domain signal through a preset filter to obtain a second frequency domain signal; wherein, the second echo component is generated by dynamic scattering of water mist;

[0045] The signal detection and sampling unit is used to perform interference filtering and target detection on the second frequency domain signal based on the constant false alarm rate algorithm to obtain a potential target point cloud, and to perform effective point sampling on the potential target point cloud to determine multiple detection points, and to obtain the detection distance and antenna reflection angle of the multiple detection points.

[0046] In some embodiments of this application, the coordinate transformation calibration module includes a model building unit and a transformation calibration unit;

[0047] The model building unit is used to build a coordinate system transformation model based on the three-dimensional rectangular coordinate system; wherein, the three-dimensional rectangular coordinate system takes the vertical projection point of the central radar on the roller as the origin, the direction of movement of the steel plate on the roller as the positive direction of the horizontal axis, the counterclockwise vertical direction of the direction of movement of the steel plate on the roller as the positive direction of the vertical axis, and the direction from the plane where the roller is located to the plane where the central radar is located as the positive direction of the vertical axis.

[0048] The transformation and calibration unit is used to perform coordinate transformation and spatial position calibration on multiple detection points according to the coordinate system transformation model, and to calculate the detection distance and antenna reflection angle of each detection point into the three-dimensional coordinates of the corresponding detection point in the three-dimensional rectangular coordinate system, thereby obtaining the three-dimensional coordinates of multiple detection points.

[0049] In some embodiments of this application, the coordinate system transformation model is specifically as follows:

[0050] ;

[0051] in, pitch angle around the longitudinal axis Roll angle around the horizontal axis and yaw angle around the vertical axis The size of the structure is The rotation matrix, The original three-dimensional coordinates of the detection point are calculated based on the detection distance and antenna reflection angle. The three-dimensional coordinates of the probe point in a three-dimensional rectangular coordinate system. This is the translation calibration vector.

[0052] In some embodiments of this application, the steel plate offset calculation module includes an offset point determination unit and an offset calculation unit;

[0053] The offset point determination unit is used to calculate the absolute value of the difference between the ordinates of each pair of the detection points based on the three-dimensional coordinates of each detection point, and to take the two detection points corresponding to the largest absolute value as the first offset point and the second offset point.

[0054] The offset calculation unit is used to calculate the Euclidean distance between the first offset point and the second offset point and the reference point, with the origin of the three-dimensional rectangular coordinate system as the reference point, to obtain the corresponding first offset and second offset.

[0055] In some embodiments of this application, the steel plate rolling correction module includes an offset rate determination unit and a steel plate rolling correction unit;

[0056] The offset rate determination unit is used to calculate the offset difference between the first offset and the second offset, and to take the absolute value of the ratio of the offset difference to the set width of the steel plate as the steel plate offset rate.

[0057] The steel plate rolling correction unit is used to determine the steel plate correction operation based on the numerical relationship between the steel plate offset rate and preset warning thresholds and shutdown thresholds, and to correct the movement direction of the steel plate during the rolling process according to the steel plate correction operation; wherein, the warning threshold is less than the shutdown threshold; the warning threshold and the shutdown threshold are determined based on the maximum allowable deviation amount.

[0058] In some embodiments of this application, the steel plate rolling correction unit includes a first correction subunit; the first correction subunit is used to upload the steel plate offset rate to the frame adjustment system if the steel plate offset rate is greater than the warning threshold and not greater than the shutdown threshold, so that the frame adjustment system selects a roll gap with corresponding concavity for steel plate rolling according to the steel plate offset rate, and completes the correction of steel plate rolling according to the roll gap selected by the frame adjustment system.

[0059] This application first acquires echo signal data of steel plates from the rolling production line using millimeter-wave radar. This avoids interference from water mist and iron oxide scale generated by cooling water evaporation, as is the case with existing camera-based visual recognition and correction schemes, thus improving the accuracy of steel plate rolling correction under complex conditions. Then, multiple detection points are acquired through filtering, analysis, and sampling. These detection points are then converted and calibrated to a three-dimensional Cartesian coordinate system. This allows for accurate acquisition of multiple detection points and coordinate system conversion and calibration, reducing measurement errors and providing a precise data foundation for subsequent calculations of steel plate offset, thereby ensuring the accuracy of steel plate rolling correction. Next, the first and second offsets of the first and second offset points, which are furthest apart along the vertical axis in the three-dimensional Cartesian coordinate system, are calculated. Based on the physical meaning of the three-dimensional coordinates, accurate offset detection can be achieved through simple calculations. This ensures the accuracy of steel plate rolling correction when the steel plate offset rate is calculated in conjunction with the set width of the steel plate and the movement direction of the steel plate during rolling is corrected.

[0060] According to a third aspect of the embodiments of this application, a computer device is provided, comprising: a processor; a memory; and a computer program stored in the memory and configured to be executed by the processor; wherein the processor executes the computer program to implement a steel plate rolling correction method according to this application.

[0061] According to a fourth aspect of the embodiments of this application, a computer-readable storage medium is provided, the computer-readable storage medium storing a plurality of instructions adapted for loading by a processor to execute a steel plate rolling correction method according to this application. Attached Figure Description

[0062] Figure 1 This is a schematic flowchart illustrating a steel plate rolling correction method according to certain embodiments of this application.

[0063] Figure 2 This is a modular structure diagram of a steel plate rolling correction system shown in certain embodiments of this application. Detailed Implementation

[0064] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below in conjunction with the accompanying drawings are exemplary and are only used to explain some embodiments of this application, and should not be construed as limiting the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments shown in this application without inventive effort are within the protection scope of this application.

[0065] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, unless otherwise explicitly specified, "a plurality of" or "several" means two or more.

[0066] Currently, the main solution for deviation correction in the rolling process involves using high-temperature resistant, vibration-proof, and fog-proof industrial inspection cameras to collect and identify the amount of deviation in images of the steel plate rolling process. Predictive models are then used to predict the deviation trend for high-precision correction. However, this solution does not consider the interference from water mist and iron oxide scale generated during water cooling during rolling, which severely affects the accuracy of camera-based visual recognition correction schemes. Therefore, maintaining high-precision deviation correction in complex environments such as water vapor (water mist) to ensure the accuracy and efficiency of steel plate rolling remains a pressing problem to be solved in existing technologies.

[0067] Those skilled in the art will readily understand that the technical solution provided in this application is applied to the correction of deviation in the rolling production process of steel plates, especially to the correction of the movement direction of the steel plates during the rolling process; generally, the steel plates mentioned in this application are any one of strip steel and medium-thick plates.

[0068] Based on the above technical background, please refer to Figure 1 This application provides a method for correcting deviations during steel plate rolling, including steps S101 to S104, each step of which is as follows:

[0069] Step S101: Based on millimeter-wave radar, acquire the echo signal data of the steel plate from the rolling production line, and filter, analyze, and sample the echo signal data of the steel plate from the rolling production line to obtain the detection distance of multiple detection points and the corresponding antenna reflection angle.

[0070] Specifically, the millimeter-wave radar is installed on the top of the steel plate rolling mill frame, and the central radar of the millimeter-wave radar is located directly above the center point of the roller conveyor.

[0071] In some embodiments of this application, the step of filtering, analyzing, and sampling the echo signal data of the steel plate from the rolling production line to obtain the detection distance and corresponding antenna reflection angle of multiple detection points specifically includes:

[0072] According to the preset detection distance range, the echo signal data of the steel plate from the rolling production line is filtered and focused. During the filtering process, signal data within the detection distance range is extracted from the echo signal data of the steel plate from the rolling production line, and the extracted signal data is smoothed to obtain a first time-domain signal; wherein, the detection distance range is determined based on the water mist diffusion distance.

[0073] Estimate the first echo component in the first time domain signal, and obtain the second time domain signal based on the difference between the first time domain signal and the first echo component; wherein the first echo component is generated by static scattering of water mist and reflection from a stationary object;

[0074] In the slow time dimension of the second time domain signal, a Fourier transform is performed on the second time domain signal to obtain a first frequency domain signal. The second echo component in the first frequency domain signal is estimated and filtered out by a preset filter to obtain a second frequency domain signal. The second echo component is generated by dynamic scattering of water mist.

[0075] The second frequency domain signal is subjected to interference filtering and target detection based on the constant false alarm rate algorithm to obtain a potential target point cloud. The potential target point cloud is then sampled for effective points to determine multiple detection points, and the detection distance and antenna reflection angle of the multiple detection points are obtained.

[0076] Specifically, the detection distance range is Among them, less than the shortest detection distance The echo signal was severely interfered with by water mist and iron oxide scale, resulting in large data errors and rendering it unreliable. This exceeded the maximum effective detection range. The echo signal exceeds the effective range, and the data also has a large error, making it unreliable. The echo generated by static scattering of water mist and reflection from stationary objects has a relatively constant amplitude and no Doppler frequency shift (velocity approximately equal to 0). Therefore, the interference of the echo can be eliminated by estimating the echo and subtracting it. By performing a Fourier transform in the slow time dimension, based on the characteristic that dynamic targets have radial velocities relative to the radar that produce Doppler frequency shifts, dynamic particles with different velocities (including real detection targets and moving water mist particles) can be divided into different frequency points in the spectrum. Then, based on the physical characteristic that water mist micro-motion clutter is usually in a specific low-speed range, a filter is constructed for adaptive filtering to eliminate the interference clutter generated by dynamic scattering of water mist. Finally, based on the constant false alarm rate algorithm, residual micro-motion interference is further filtered out, and the potential target point cloud is output, thereby obtaining more accurate multiple effective detection points.

[0077] Specifically, the constant false alarm rate algorithm (CFAR algorithm) includes, but is not limited to, the CA-CFAR algorithm, the OS-CFAR algorithm and their corresponding variants, with the CA-CFAR algorithm being preferred.

[0078] This application first filters signal data within the detection range and smooths it to obtain a first time-domain signal. The detection range is determined based on the water mist diffusion distance, which can reduce echo signals severely interfered with by the environment and initially focus on the effective signal range. Then, it estimates the first echo component generated by static scattering of water mist and reflection from stationary objects, and obtains a second time-domain signal by difference, which can filter out the interference echo signal generated by static scattering of water mist and reflection from stationary objects. Then, it performs a Fourier transform in the slow time dimension and filters out the second echo component generated by dynamic scattering of water mist to obtain a second frequency-domain signal, which can filter out the interference clutter signal generated by low-speed micro-motion of water mist. Finally, it uses a constant false alarm rate algorithm for interference filtering and target detection, which can adaptively filter out residual micro-motion interference on the noise substrate, so that when sampling effective points, multiple accurate detection points and their detection distances and antenna reflection angles are obtained.

[0079] Step S102: Based on a preset three-dimensional rectangular coordinate system, perform coordinate transformation calibration on multiple detection points, and obtain the three-dimensional coordinates of multiple detection points through the detection distance and the antenna reflection angle.

[0080] In some embodiments of this application, the step of performing coordinate transformation calibration on multiple detection points based on a preset three-dimensional Cartesian coordinate system, and obtaining the three-dimensional coordinates of the multiple detection points through the detection distance and the antenna reflection angle, specifically includes:

[0081] Based on the three-dimensional rectangular coordinate system, a coordinate system transformation model is constructed; wherein, the three-dimensional rectangular coordinate system takes the vertical projection point of the central radar on the roller as the origin, the direction of movement of the steel plate on the roller as the positive direction of the horizontal axis, the counterclockwise vertical direction of the direction of movement of the steel plate on the roller as the positive direction of the vertical axis, and the direction from the plane where the roller is located to the plane where the central radar is located as the positive direction of the vertical axis.

[0082] According to the coordinate system transformation model, coordinate transformation and spatial position calibration are performed on multiple detection points. The detection distance and antenna reflection angle of each detection point are transformed and calculated into the three-dimensional coordinates of the corresponding detection point in the three-dimensional rectangular coordinate system, thereby obtaining the three-dimensional coordinates of multiple detection points.

[0083] This application first constructs a three-dimensional rectangular coordinate system based on the positional relationship between the radar, steel plate, and roller conveyor, then builds a coordinate system transformation model, and performs coordinate transformation and spatial position calibration on multiple detection points according to the coordinate system transformation model. This can accurately transform the coordinate representation of multiple detection points from the original coordinate system to the three-dimensional coordinate representation in this three-dimensional rectangular coordinate system, reduce measurement errors, and provide an accurate data basis for subsequent calculation of steel plate offset.

[0084] In some embodiments of this application, the coordinate system transformation model is specifically as follows:

[0085] ;

[0086] in, pitch angle around the longitudinal axis Roll angle around the horizontal axis and yaw angle around the vertical axis The size of the structure is The rotation matrix, The original three-dimensional coordinates of the detection point are calculated based on the detection distance and antenna reflection angle. The three-dimensional coordinates of the probe point in a three-dimensional rectangular coordinate system. This is the translation calibration vector.

[0087] This application calculates the original three-dimensional coordinates of the detection point and the antenna reflection angle, and combines them with the rotation matrix and translation calibration vector to obtain the three-dimensional coordinates in the three-dimensional rectangular coordinate system. This enables accurate coordinate transformation and spatial position calibration, reduces measurement errors, and provides a precise data basis for subsequent calculation of steel plate offset.

[0088] Step S103: Calculate the first offset corresponding to the first offset point and the second offset corresponding to the second offset point based on the three-dimensional coordinates of each detection point; wherein the first offset point and the second offset point are the two detection points that are farthest apart in the vertical direction of the three-dimensional rectangular coordinate system.

[0089] In some embodiments of this application, calculating the first offset corresponding to the first offset point and the second offset corresponding to the second offset point based on the three-dimensional coordinates of each of the detection points specifically includes:

[0090] Based on the three-dimensional coordinates of each detection point, calculate the absolute value of the difference between the ordinates of each pair of detection points, and take the two detection points corresponding to the largest absolute value as the first offset point and the second offset point.

[0091] Using the origin of the three-dimensional rectangular coordinate system as a reference point, the Euclidean distances between the first offset point and the second offset point and the reference point are calculated respectively to obtain the corresponding first offset and second offset.

[0092] Specifically, the Euclidean distance is as follows:

[0093] ;

[0094] ;

[0095] in, These are the first offset point, the second offset point, and the reference point, respectively. These are the first offset and the second offset, respectively.

[0096] This application first determines the two detection points with the farthest vertical coordinates as the first offset point and the second offset point based on the three-dimensional coordinates of each detection point. Then, the first offset and the second offset are obtained by the Euclidean distance with the reference point. Based on the physical meaning of the three-dimensional coordinates, the offset can be accurately detected by simple calculation. While simplifying the offset calculation, the accuracy of the offset calculation is guaranteed.

[0097] Step S104: Based on the first offset and the second offset, and combined with the preset steel plate width, calculate the steel plate offset rate, and correct the movement direction of the steel plate during the rolling process based on the steel plate offset rate.

[0098] In some embodiments of this application, the step of calculating the steel plate offset rate based on the first offset and the second offset, combined with a preset steel plate width, and correcting the movement direction of the steel plate during rolling based on the steel plate offset rate specifically includes:

[0099] Calculate the offset difference between the first offset and the second offset, and take the absolute value of the ratio of the offset difference to the set width of the steel plate as the steel plate offset rate;

[0100] Based on the relationship between the steel plate offset rate and the preset warning threshold and shutdown threshold, a steel plate correction operation is determined, and the movement direction of the steel plate during the rolling process is corrected according to the steel plate correction operation; wherein, the warning threshold is less than the shutdown threshold; the warning threshold and the shutdown threshold are determined based on the maximum allowable deviation amount.

[0101] Specifically, the steel plate offset rate is as follows:

[0102] ;

[0103] in, For steel plate offset rate, Set the width for the steel plate.

[0104] Specifically, the maximum permissible deviation is denoted as... The preferred warning threshold is The preferred shutdown threshold is .

[0105] This application first calculates the offset difference between the first offset and the second offset, then obtains the steel plate offset rate by combining the set width of the steel plate, and then determines the steel plate correction operation and performs correction by using the warning threshold and the shutdown threshold determined according to the maximum allowable deviation amount. This allows for accurate definition and calculation of the steel plate offset rate, thereby accurately determining the steel plate correction operation and performing high-precision correction.

[0106] In some embodiments of this application, determining the steel plate correction operation based on the numerical relationship between the steel plate offset rate and preset warning thresholds and shutdown thresholds, and correcting the movement direction of the steel plate during the rolling process according to the steel plate correction operation, specifically includes:

[0107] If the steel plate offset rate is greater than the warning threshold but not greater than the shutdown threshold, the steel plate offset rate is uploaded to the frame adjustment system so that the frame adjustment system selects the roll gap with the corresponding concavity according to the steel plate offset rate for steel plate rolling, and completes the correction of the steel plate rolling according to the roll gap selected by the frame adjustment system.

[0108] In some embodiments of this application, if the steel plate offset rate is not greater than the warning threshold, the steel plate offset rate is uploaded to the rack adjustment system so that the rack adjustment system can continuously monitor the steel plate offset rate. Specifically, if the rack adjustment system detects that the steel plate offset rate is greater than the warning threshold for N (N≥3) upload cycles, preferably greater than 5% of the warning threshold, then a warning command is issued to execute the steel plate deviation warning; if the rack adjustment system detects that the steel plate offset rate is greater than the shutdown threshold for N upload cycles, preferably greater than 10% of the shutdown threshold, then a shutdown command is issued to execute the steel plate deviation shutdown.

[0109] In some embodiments of this application, if the steel plate offset rate is greater than the shutdown threshold, the steel plate offset rate is uploaded to the stand adjustment system so that the stand adjustment system stops steel plate rolling and performs a shutdown operation based on the steel plate offset rate.

[0110] When the steel plate offset rate is between the warning threshold and the shutdown threshold, this application uploads the steel plate offset rate to the frame adjustment system so that the frame adjustment system can perform steel plate correction operation, thereby achieving accurate and high-precision steel plate correction.

[0111] Compared to existing technologies, this application first acquires the echo signal data of steel plates from the rolling production line using millimeter-wave radar. This avoids interference from water mist and iron oxide scale generated by cooling water evaporation, as is the case with existing camera-based visual recognition and correction schemes, thus improving the accuracy of steel plate rolling correction under complex conditions. Then, multiple detection points are acquired through filtering, analysis, and sampling. These detection points are then converted and calibrated to a three-dimensional Cartesian coordinate system. This allows for accurate acquisition of multiple detection points and coordinate system conversion and calibration, reducing measurement errors and providing a precise data foundation for subsequent calculations of steel plate offset, thereby ensuring the accuracy of steel plate rolling correction. Next, the first and second offsets of the first and second offset points, which are furthest apart along the vertical axis in the three-dimensional Cartesian coordinate system, are calculated. Based on the physical meaning of the three-dimensional coordinates, accurate offset detection can be achieved through simple calculations. This ensures the accuracy of steel plate rolling correction when the steel plate offset rate is calculated in conjunction with the set width of the steel plate and the movement direction of the steel plate during rolling is corrected.

[0112] For a method corresponding to the one described above, please refer to [link to relevant documentation]. Figure 2 This application provides a steel plate rolling correction system, including an echo filtering sampling module 210, a coordinate transformation calibration module 220, a steel plate offset calculation module 230, and a steel plate rolling correction module 240.

[0113] The echo filtering and sampling module 210 is used to acquire the echo signal data of the steel plate in the rolling production line based on millimeter-wave radar, and to filter, analyze and sample the echo signal data of the steel plate in the rolling production line to obtain the detection distance and corresponding antenna reflection angle of multiple detection points.

[0114] The coordinate transformation calibration module 220 is used to perform coordinate transformation calibration on multiple detection points based on a preset three-dimensional rectangular coordinate system, and obtain the three-dimensional coordinates of multiple detection points through the detection distance and the antenna reflection angle;

[0115] The steel plate offset calculation module 230 is used to calculate the first offset corresponding to the first offset point and the second offset corresponding to the second offset point based on the three-dimensional coordinates of each detection point; wherein, the first offset point and the second offset point are the two detection points that are farthest apart in the vertical axis direction of the three-dimensional rectangular coordinate system.

[0116] The steel plate rolling correction module 240 is used to calculate the steel plate offset rate based on the first offset and the second offset, combined with the preset steel plate width, and to correct the movement direction of the steel plate during the rolling process based on the steel plate offset rate.

[0117] In some embodiments of this application, the echo filtering sampling module 210 includes a signal screening and smoothing unit, a signal estimation and filtering unit, a signal transformation and filtering unit, and a signal detection and sampling unit;

[0118] The signal filtering and smoothing unit is used to filter and focus the echo signal data of the steel plate from the rolling production line according to a preset detection distance range. During the filtering process, signal data within the detection distance range in the echo signal data of the steel plate from the rolling production line is extracted, and the extracted signal data is smoothed to obtain a first time-domain signal. The detection distance range is determined based on the water mist diffusion distance.

[0119] The signal estimation and filtering unit is used to estimate the first echo component in the first time domain signal and obtain the second time domain signal based on the difference between the first time domain signal and the first echo component; wherein the first echo component is generated by static scattering of water mist and reflection from a stationary object;

[0120] The signal transformation and filtering unit is used to perform a Fourier transform on the second time domain signal in the slow time dimension to obtain a first frequency domain signal, and to estimate and filter out the second echo component in the first frequency domain signal through a preset filter to obtain a second frequency domain signal; wherein, the second echo component is generated by dynamic scattering of water mist;

[0121] The signal detection and sampling unit is used to perform interference filtering and target detection on the second frequency domain signal based on the constant false alarm rate algorithm to obtain a potential target point cloud, and to perform effective point sampling on the potential target point cloud to determine multiple detection points, and to obtain the detection distance and antenna reflection angle of the multiple detection points.

[0122] In some embodiments of this application, the coordinate transformation calibration module 220 includes a model building unit and a transformation calibration unit;

[0123] The model building unit is used to build a coordinate system transformation model based on the three-dimensional rectangular coordinate system; wherein, the three-dimensional rectangular coordinate system takes the vertical projection point of the central radar on the roller as the origin, the direction of movement of the steel plate on the roller as the positive direction of the horizontal axis, the counterclockwise vertical direction of the direction of movement of the steel plate on the roller as the positive direction of the vertical axis, and the direction from the plane where the roller is located to the plane where the central radar is located as the positive direction of the vertical axis.

[0124] The transformation and calibration unit is used to perform coordinate transformation and spatial position calibration on multiple detection points according to the coordinate system transformation model, and to calculate the detection distance and antenna reflection angle of each detection point into the three-dimensional coordinates of the corresponding detection point in the three-dimensional rectangular coordinate system, thereby obtaining the three-dimensional coordinates of multiple detection points.

[0125] In some embodiments of this application, the coordinate system transformation model is specifically as follows:

[0126] ;

[0127] in, pitch angle around the longitudinal axis Roll angle around the horizontal axis and yaw angle around the vertical axis The size of the structure is The rotation matrix, The original three-dimensional coordinates of the detection point are calculated based on the detection distance and antenna reflection angle. The three-dimensional coordinates of the probe point in a three-dimensional rectangular coordinate system. This is the translation calibration vector.

[0128] In some embodiments of this application, the steel plate offset calculation module 230 includes an offset point determination unit and an offset calculation unit;

[0129] The offset point determination unit is used to calculate the absolute value of the difference between the ordinates of each pair of the detection points based on the three-dimensional coordinates of each detection point, and to take the two detection points corresponding to the largest absolute value as the first offset point and the second offset point.

[0130] The offset calculation unit is used to calculate the Euclidean distance between the first offset point and the second offset point and the reference point, with the origin of the three-dimensional rectangular coordinate system as the reference point, to obtain the corresponding first offset and second offset.

[0131] In some embodiments of this application, the steel plate rolling correction module 240 includes an offset rate determination unit and a steel plate rolling correction unit;

[0132] The offset rate determination unit is used to calculate the offset difference between the first offset and the second offset, and to take the absolute value of the ratio of the offset difference to the set width of the steel plate as the steel plate offset rate.

[0133] The steel plate rolling correction unit is used to determine the steel plate correction operation based on the numerical relationship between the steel plate offset rate and preset warning thresholds and shutdown thresholds, and to correct the movement direction of the steel plate during the rolling process according to the steel plate correction operation; wherein, the warning threshold is less than the shutdown threshold; the warning threshold and the shutdown threshold are determined based on the maximum allowable deviation amount.

[0134] In some embodiments of this application, the steel plate rolling correction unit includes a first correction subunit; the first correction subunit is used to upload the steel plate offset rate to the frame adjustment system if the steel plate offset rate is greater than the warning threshold and not greater than the shutdown threshold, so that the frame adjustment system selects a roll gap with corresponding concavity for steel plate rolling according to the steel plate offset rate, and completes the correction of steel plate rolling according to the roll gap selected by the frame adjustment system.

[0135] This application first acquires the echo signal data of the steel plate in the rolling production line using millimeter-wave radar. This avoids interference from water mist and iron oxide scale generated by cooling water evaporation, as is the case with existing camera-based visual recognition and correction schemes, thus improving the accuracy of steel plate rolling correction under complex conditions. Then, multiple detection points are acquired through filtering, analysis, and sampling. These detection points are then converted and calibrated to a three-dimensional Cartesian coordinate system. This allows for accurate acquisition of multiple detection points and coordinate system conversion and calibration, reducing measurement errors and providing a precise data basis for subsequent calculation of steel plate offset, thereby ensuring the accuracy of steel plate rolling correction. Next, the first and second offsets of the first and second offset points, which are farthest apart along the vertical axis in the three-dimensional Cartesian coordinate system, are calculated. Based on the physical meaning of the three-dimensional coordinates, accurate offset detection can be achieved through simple calculation. This ensures the accuracy of steel plate rolling correction when the steel plate offset rate is calculated in conjunction with the set width of the steel plate and the movement direction of the steel plate during rolling is corrected.

[0136] It should be understood that the system provided in this application is corresponding to the aforementioned method. The steel plate rolling correction system provided in this application can implement the steel plate rolling correction method provided in any of the embodiments of this application.

[0137] Adaptively, embodiments of this application also provide a computer device and a computer-readable storage medium.

[0138] The computer device includes: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor;

[0139] The processor executes the computer program to implement a steel plate rolling correction method of this application.

[0140] The computer-readable storage medium stores multiple instructions, which are adapted for a processor to load and execute a steel plate rolling correction method of this application.

[0141] The above description represents some embodiments of this application, providing a further detailed explanation of the purpose, technical solution, and beneficial effects of this application. It should be understood that the above-described embodiments of this application should not be construed as limiting this application. In particular, any changes, modifications, equivalent substitutions, and variations made by those skilled in the art within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for correcting deviations during steel plate rolling, characterized in that, include: Based on millimeter-wave radar, echo signal data of steel plates from the rolling production line is acquired. This data is then filtered, analyzed, and sampled to obtain the detection distances and corresponding antenna reflection angles for multiple detection points. Specifically, the echo signal data is filtered and focused according to a preset detection distance interval. During filtering, signal data falling within the detection distance interval is extracted, and the extracted signal data is smoothed to obtain a first time-domain signal. The detection distance interval is determined based on the water mist diffusion distance. A first echo component in the first time-domain signal is estimated, and the results are compared with the first time-domain signal... The difference is used to obtain a second time-domain signal; wherein the first echo component is generated by static scattering of water mist and reflection from a stationary object; in the slow time dimension of the second time-domain signal, a Fourier transform is performed on the second time-domain signal to obtain a first frequency-domain signal, and the second echo component in the first frequency-domain signal is estimated and filtered out by a preset filter to obtain a second frequency-domain signal; wherein the second echo component is generated by dynamic scattering of water mist; based on a constant false alarm rate algorithm, interference filtering and target detection are performed on the second frequency-domain signal to obtain a potential target point cloud, and effective point sampling is performed on the potential target point cloud to determine multiple detection points, and the detection distance and antenna reflection angle of the multiple detection points are obtained; Based on a preset three-dimensional rectangular coordinate system, coordinate transformation calibration is performed on multiple detection points, and the three-dimensional coordinates of multiple detection points are obtained through the detection distance and the antenna reflection angle; Based on the three-dimensional coordinates of each detection point, calculate the first offset corresponding to the first offset point and the second offset corresponding to the second offset point; wherein, the first offset point and the second offset point are the two detection points that are farthest apart in the vertical direction of the three-dimensional rectangular coordinate system. Based on the first offset and the second offset, and combined with the preset steel plate width, the steel plate offset rate is calculated, and the movement direction of the steel plate during the rolling process is corrected according to the steel plate offset rate.

2. The method for correcting deviations during steel plate rolling according to claim 1, characterized in that, The process involves performing coordinate transformation calibration on multiple detection points based on a preset three-dimensional Cartesian coordinate system. The three-dimensional coordinates of these detection points are obtained using the detection distance and the antenna reflection angle. Specifically, this includes: Based on the three-dimensional rectangular coordinate system, a coordinate system transformation model is constructed; wherein, the three-dimensional rectangular coordinate system takes the vertical projection point of the central radar on the roller as the origin, the direction of movement of the steel plate on the roller as the positive direction of the horizontal axis, the counterclockwise vertical direction of the direction of movement of the steel plate on the roller as the positive direction of the vertical axis, and the direction from the plane where the roller is located to the plane where the central radar is located as the positive direction of the vertical axis. According to the coordinate system transformation model, coordinate transformation and spatial position calibration are performed on multiple detection points. The detection distance and antenna reflection angle of each detection point are transformed and calculated into the three-dimensional coordinates of the corresponding detection point in the three-dimensional rectangular coordinate system, thereby obtaining the three-dimensional coordinates of multiple detection points.

3. The method for correcting deviations during steel plate rolling according to claim 2, characterized in that, The coordinate system transformation model is as follows: ； in, pitch angle around the longitudinal axis Roll angle around the horizontal axis and yaw angle around the vertical axis The size of the structure is The rotation matrix, The original three-dimensional coordinates of the detection point are calculated based on the detection distance and antenna reflection angle. The three-dimensional coordinates of the probe point in a three-dimensional rectangular coordinate system. This is the translation calibration vector.

4. The method for correcting deviations during steel plate rolling according to claim 1, characterized in that, The step of calculating the first offset corresponding to the first offset point and the second offset corresponding to the second offset point based on the three-dimensional coordinates of each detection point specifically includes: Based on the three-dimensional coordinates of each detection point, calculate the absolute value of the difference between the ordinates of each pair of detection points, and take the two detection points corresponding to the largest absolute value as the first offset point and the second offset point. Using the origin of the three-dimensional rectangular coordinate system as a reference point, the Euclidean distances between the first offset point and the second offset point and the reference point are calculated respectively to obtain the corresponding first offset and second offset.

5. The method for correcting deviations during steel plate rolling according to claim 1, characterized in that, The step of calculating the steel plate offset rate based on the first offset and the second offset, combined with a preset steel plate width, and correcting the movement direction of the steel plate during the rolling process based on the steel plate offset rate, specifically includes: Calculate the offset difference between the first offset and the second offset, and take the absolute value of the ratio of the offset difference to the set width of the steel plate as the steel plate offset rate; Based on the relationship between the steel plate offset rate and the preset warning threshold and shutdown threshold, a steel plate correction operation is determined, and the movement direction of the steel plate during the rolling process is corrected according to the steel plate correction operation; wherein, the warning threshold is less than the shutdown threshold; the warning threshold and the shutdown threshold are determined based on the maximum allowable deviation amount.

6. The method for correcting deviations during steel plate rolling according to claim 5, characterized in that, The process involves determining a steel plate correction operation based on the relationship between the steel plate offset rate and preset warning and shutdown thresholds, and then correcting the movement direction of the steel plate during rolling according to the steel plate correction operation. Specifically, this includes: If the steel plate offset rate is greater than the warning threshold but not greater than the shutdown threshold, the steel plate offset rate is uploaded to the frame adjustment system so that the frame adjustment system selects the roll gap with the corresponding concavity according to the steel plate offset rate for steel plate rolling, and completes the correction of the steel plate rolling according to the roll gap selected by the frame adjustment system.

7. A steel plate rolling correction system, characterized in that, The method for implementing the steel plate rolling correction method as described in any one of claims 1 to 6 includes an echo filtering sampling module, a coordinate transformation calibration module, a steel plate offset calculation module, and a steel plate rolling correction module. The echo filtering and sampling module is used to acquire the echo signal data of the steel plate in the rolling production line based on millimeter-wave radar, and to filter, analyze and sample the echo signal data of the steel plate in the rolling production line to obtain the detection distance and corresponding antenna reflection angle of multiple detection points. The coordinate transformation calibration module is used to perform coordinate transformation calibration on multiple detection points based on a preset three-dimensional rectangular coordinate system, and obtain the three-dimensional coordinates of multiple detection points through the detection distance and the antenna reflection angle; The steel plate offset calculation module is used to calculate the first offset corresponding to the first offset point and the second offset corresponding to the second offset point based on the three-dimensional coordinates of each detection point; wherein, the first offset point and the second offset point are the two detection points that are farthest apart in the vertical axis direction of the three-dimensional rectangular coordinate system. The steel plate rolling correction module is used to calculate the steel plate offset rate based on the first offset and the second offset, combined with the preset steel plate width, and to correct the movement direction of the steel plate during the rolling process based on the steel plate offset rate.

8. A computer device, characterized in that, include: processor; Memory; A computer program stored in the memory and configured to be executed by the processor; When the processor executes the computer program, it implements a steel plate rolling correction method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of instructions adapted for loading by a processor to execute a steel plate rolling correction method according to any one of claims 1 to 6.

Images