Strip steel cooling control method, electronic device and storage medium

Abstract

The application relates to the technical field of metal strip processing, and particularly provides a strip steel cooling control method, an electronic device and a storage medium. The method comprises the following steps: acquiring a thermal radiation signal of cooled strip steel in a preset infrared wave band, and converting the thermal radiation signal into a two-dimensional temperature field; extracting actual coordinates of blackened points from the two-dimensional temperature field; collecting a stripe image modulated by the surface morphology of the cooled strip steel; acquiring three-dimensional morphology data of the surface of the cooled strip steel according to the stripe image, and extracting a plate shape deformation characteristic quantity based on the three-dimensional morphology data; calculating a coordinate deviation according to the actual coordinates of the blackened points and preset blackened point coordinates; and generating cooling intensity adjustment instructions for different cooling areas through preset coupling control logic according to the coordinate deviation and the plate shape deformation characteristic quantity, so as to drive corresponding execution mechanisms of the different cooling areas to adjust the flow of cooling medium. The method can effectively solve the problems of large manual observation error, serious steam interference and inaccurate plate shape control.

Description

Technical Field

[0001] This application relates to the field of metal strip processing technology, and more specifically, to a strip cooling control method, electronic equipment, and storage medium. Background Technology

[0002] In the normalizing process of grain-oriented silicon steel, after the strip is heated to a high temperature, it must enter a quenching section for rapid cooling. During this process, the accurate determination of the location of the blackening points is a core indicator for measuring the cooling effect, directly reflecting the physical location of the strip at a specific temperature. Traditional processes mainly rely on operators visually observing the blackening points and manually adjusting the cooling water spray flow rate accordingly.

[0003] However, this traditional operating mode has significant technical drawbacks. First, the contact between the high-temperature steel plate and the cooling water in the quenching section generates a large amount of steam mist. This mist not only obstructs vision but also makes it difficult for optical monitoring equipment to obtain clear images, making it highly susceptible to errors in determining the location of blackening points. Second, the production environment suffers from severe light interference; fluctuations in workshop lighting and natural light intensity can cause significant shifts in the apparent location of blackening points, leading to misjudgments by operators. Furthermore, due to the lack of real-time coupled monitoring of the cooling process and changes in plate shape, the manual adjustment process is highly subjective, making it difficult to achieve precise control of the cooling rate.

[0004] Cooling rate is a key process parameter affecting the magnetic properties and shape of silicon steel. Improper control can lead to uncontrolled precipitate morphology and abnormal microstructure, resulting in performance defects. At the same time, uneven distribution of cooling rate in the strip running direction can easily induce shape defects such as edge waviness, center waviness, or warping, and in severe cases, it can even lead to deviation or steel piling accidents.

[0005] There is currently no effective technical solution to the above problems. Summary of the Invention

[0006] The purpose of this application is to provide a strip cooling control method, electronic equipment and storage medium that can effectively solve the problems of large errors in traditional manual observation, serious steam interference and inaccurate strip shape control.

[0007] In a first aspect, this application provides a method for controlling the cooling of strip steel, which includes the following steps: S1. Acquire the thermal radiation signal of the cooled strip steel in the preset infrared band and convert the thermal radiation signal into a two-dimensional temperature field; the preset infrared band is 3.7-4.1μm; S2. Extract the spatial locations that meet the preset temperature threshold from the two-dimensional temperature field as the actual coordinates of the blackening points; S3. Project laser stripes onto the surface of the cooling strip steel and acquire stripe images modulated by the surface morphology of the cooling strip steel. S4. Obtain the three-dimensional morphology data of the surface of the cooling strip steel based on the stripe image, and extract the plate deformation feature quantity based on the three-dimensional morphology data; S5. Calculate the coordinate deviation based on the actual coordinates of the blackened points and the preset coordinates of the blackened points; S6. Based on the coordinate deviation and plate deformation characteristics, generate cooling intensity adjustment commands for different cooling zones through preset coupling control logic to drive the actuators corresponding to different cooling zones to adjust the cooling medium flow rate.

[0008] Optionally, the thermal radiation signal is acquired by an infrared thermal imager, and the strip cooling control method further includes steps performed before step S1: A1. Use the two-point calibration method to obtain the calibration gain and calibration bias coefficient of each pixel in the infrared thermal imager; Step S1 includes: S11. Obtain the thermal radiation signal of the cooled strip steel in a preset infrared band; S12. Correct the thermal radiation signal based on all calibration gains and all calibration bias coefficients; S13. Using a pre-built prior model of steam concentration spatial distribution, the estimated steam concentration distribution is obtained based on the current cooling medium flow rate corresponding to different cooling zones. S14. For each pixel in the infrared thermal imager, obtain the steam transmittance of the pixel based on the geometric optical path from the strip position corresponding to the pixel to the infrared thermal imager and the estimated steam concentration distribution. S15. Using the Biber-Lambert law, attenuation compensation of thermal radiation signal is performed based on vapor transmittance. S16. Obtain the current background radiation value of the reference area and remove the current background radiation value from the thermal radiation signal; the reference area is a region located in the steam-free zone of the quenching section where the temperature is known. S17. Convert the compensated thermal radiation signal into a two-dimensional temperature field.

[0009] Optionally, step S17 includes: S171. Perform time-domain median filtering on the compensated thermal radiation signal of multiple consecutive frames. S172. Convert the current frame thermal radiation signal, after completing the time-domain median filtering, into a two-dimensional temperature field.

[0010] Optionally, step S2 includes: S21. Extract the median value of the two-dimensional temperature field along the running direction of the cooling strip steel to obtain a one-dimensional temperature curve along the running direction of the cooling strip steel. S22. The spatial position in the one-dimensional temperature curve where the temperature value is equal to the preset temperature threshold is taken as the actual coordinate of the blackening point.

[0011] Optionally, the laser stripe is a monochromatic laser stripe projected by a high-power laser, and the stripe image is acquired by an image acquisition component, which is equipped with a narrow-band pass filter corresponding to the center fluctuation of the monochromatic laser stripe.

[0012] Optionally, step S4 includes: S401. Using morphological opening operations, isolated micro-point scattered noise around the laser stripes in the stripe image is removed. S402. Using the Steger algorithm or the gray-scale centroid method, locate the gray-scale centroid along the normal direction of the laser stripe to obtain the continuous skeleton line of the laser stripe. S403. Using a preset continuous length threshold and a directional filter consistent with the main direction of the laser stripe, stray signals that do not conform to the projection pattern in the continuous skeleton line are removed. S404. For each pixel on the continuous skeleton line, construct a spatial ray that starts from the pixel and passes through the pixel's coordinates on the image plane. Then, use the coordinates of the intersection of the spatial ray and the known spatial laser plane as the three-dimensional coordinates of the pixel. S405. Integrate all three-dimensional coordinates into a three-dimensional coordinate sequence, and use this three-dimensional coordinate sequence as the cross-sectional profile of the current frame; S406. The continuous frame cross-sectional profiles are spatially superimposed along the running direction of the cooling strip steel to obtain three-dimensional point cloud data, and the three-dimensional point cloud data is used as the three-dimensional topography data of the surface of the cooling strip steel. S407. Denoise the 3D point cloud data; S408. Using the least squares method or a robust fitting algorithm, perform plane fitting on the denoised 3D point cloud data to obtain the reference plane. S409. Calculate the normal deviation of each spatial point in the three-dimensional point cloud data relative to the reference plane point by point, and integrate all normal deviations into a two-dimensional height deviation matrix. S410. Extract the height deviation sequences of the edge region, the middle region, and the local region from the two-dimensional height deviation matrix. Then, calculate the root mean square deviation or peak-valley value corresponding to the edge region, the middle region, and the local region according to the height deviation sequences of the edge region, the middle region, and the local region, and use all root mean square deviations or all peak-valley values ​​as plate deformation characteristic quantities.

[0013] Optionally, step S6 includes: S61. Analyze whether there are plate shape defects based on the plate shape deformation characteristic quantity. If yes, proceed to step S62; otherwise, proceed to step S63. S62. Using a pre-built mapping table of plate shape defect types and correction directions, the plate shape defect types are mapped to the cooling intensity correction direction of each cooling area, and the lateral difference correction amount of each cooling area is generated based on the plate shape deformation characteristic amount. S63. Generate longitudinal difference correction amount for each cooling zone based on coordinate deviation; S64. When only the longitudinal difference correction amount exists, generate the cooling intensity adjustment command corresponding to each cooling area based on the longitudinal difference correction amount and the feedforward reference amount corresponding to each cooling area. S65. When both longitudinal and lateral differential corrections exist, generate cooling intensity adjustment commands for each cooling zone based on the longitudinal, lateral, and feedforward reference values ​​corresponding to each cooling zone. S66. Send all cooling intensity adjustment commands to the corresponding actuators to drive the actuators corresponding to different cooling zones to adjust the cooling medium flow rate.

[0014] Optionally, the cooling medium is water, and the formula for calculating the feedforward reference quantity is: V_i_base=f(Q_i_feedforward); Where Vi_i_base represents the feedforward reference value corresponding to the i-th cooling zone, and Qi_i_feedforward represents the feedforward flow setting value corresponding to the i-th cooling zone; The formula for calculating the feedforward flow setpoint is: Q_i_feedforward=k_i×P_total / (c_water×ΔT_water); Where k_i represents the length ratio coefficient of the actuator corresponding to the i-th cooling zone, P_total represents the total cooling capacity, c_water represents the specific heat capacity of water, and ΔT_water represents the allowable temperature rise of water. The formula for calculating the total cooling capacity is: P_total=ρ×Cp×h×w×v×(T_in-T_target) / 60; Where ρ represents the density of the cooling strip, Cp represents the specific heat capacity of the cooling strip, h represents the thickness of the cooling strip, w represents the width of the cooling strip, v represents the running speed of the cooling strip, T_in represents the actual temperature of the cooling strip at the inlet of the quench section, and T_target represents the target temperature of the cooling strip at the outlet of the quench section.

[0015] Secondly, this application provides an electronic device including a processor and a memory, the memory storing computer-readable instructions, which, when executed by the processor, perform the steps of the method provided in the first aspect above.

[0016] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the steps of the method provided in the first aspect above.

[0017] As can be seen from the above, the strip cooling control method, electronic equipment, and storage medium provided in this application acquire infrared thermal radiation signals and perform steam compensation and background removal, combined with laser stripe scanning to obtain three-dimensional morphology, thereby achieving precise positioning of blackening points and real-time monitoring of strip shape characteristics. By dynamically adjusting the cooling intensity through coupled control logic, it effectively solves the problems of large errors in traditional manual observation, severe steam interference, and inaccurate strip shape control. It has the advantages of effectively overcoming steam interference and ambient light fluctuations, achieving real-time, precise, and automated control of the strip cooling process, effectively suppressing strip shape defects, and improving product quality. Attached Figure Description

[0018] Figure 1 A flowchart of a strip cooling control method provided in an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of an electronic device structure provided in an embodiment of this application.

[0020] Reference numerals: 101, processor; 102, memory; 103, communication bus. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0022] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0023] Firstly, such as Figure 1As shown, this application provides a method for controlling the cooling of strip steel, which includes the following steps: S1. Acquire the thermal radiation signal of the cooled strip steel in the preset infrared band and convert the thermal radiation signal into a two-dimensional temperature field; the preset infrared band is 3.7-4.1μm; S2. Extract the spatial locations that meet the preset temperature threshold from the two-dimensional temperature field as the actual coordinates of the blackening points; S3. Project laser stripes onto the surface of the cooling strip steel and acquire stripe images modulated by the surface morphology of the cooling strip steel. S4. Obtain the three-dimensional morphology data of the surface of the cooling strip steel based on the stripe image, and extract the plate deformation feature quantity based on the three-dimensional morphology data; S5. Calculate the coordinate deviation based on the actual coordinates of the blackened points and the preset coordinates of the blackened points; S6. Based on the coordinate deviation and plate deformation characteristics, generate cooling intensity adjustment commands for different cooling zones through preset coupling control logic to drive the actuators corresponding to different cooling zones to adjust the cooling medium flow rate.

[0024] For ease of understanding, some key terms in this embodiment are explained below. During the strip cooling process, the blackening point refers to the location where the strip surface temperature drops to a specific threshold (e.g., 550°C). This location is a key indicator for judging the cooling rate and uniformity. The two-dimensional temperature field in this embodiment is generated by acquiring the thermal radiation signal of the strip surface using an infrared thermal imager, processing and converting it to form a two-dimensional image representing the temperature distribution at various points on the strip surface. This image can intuitively reflect the temperature gradient and distribution of the strip surface. The three-dimensional topography data in this embodiment is obtained through laser stripe projection and image acquisition technology to acquire the elevation information of the strip surface, forming a dataset containing the three-dimensional geometry of the strip surface. This data can accurately describe the flatness, waviness, and other topographic features of the strip surface. The strip shape deformation characteristics in this embodiment are parameters extracted from the three-dimensional topography data to quantify strip shape defects, such as root mean square deviation or peak-to-valley values. These characteristics can reflect the degree of deformation at the edges, center, or local areas of the strip. The coordinate deviation in this embodiment is the difference between the actual coordinates of the blackening point and the preset coordinates of the blackening point. This deviation quantifies the longitudinal temperature control error in the cooling process and is an important basis for adjusting the cooling medium flow rate. The cooling intensity adjustment command in this embodiment is generated by a preset coupled control logic based on the coordinate deviation and plate deformation characteristics. This command is used to adjust the cooling medium flow rate in different cooling zones, enabling refined and differentiated control of the cooling process. The actuator in this embodiment is a device that receives the cooling intensity adjustment command and actually adjusts the cooling medium flow rate, such as a water valve or spray head.

[0025] This application proposes a strip cooling control method, which aims to solve the problems of apparent drift of blackening point position caused by ambient light fluctuations during the normalization process of grain-oriented silicon steel, and the inability of manual observation to accurately control the cooling rate and strip shape in real time.

[0026] This method first acquires the thermal radiation signal of the cooled strip steel in a preset infrared band and converts the thermal radiation signal into a two-dimensional temperature field. In actual operation, an infrared thermal imager can be placed above the cooled strip steel to collect the thermal radiation signal of the strip steel surface in real time. The infrared thermal imager can be configured to operate in the preset infrared band of 3.7-4.1μm. Since the energy of ambient light (natural light or conventional industrial lighting) is mainly concentrated in the visible light and near-infrared bands (0.4-2.5μm), the radiation energy of ambient light itself in the mid-wave infrared band of 3.7-4.1μm is extremely low and can be almost ignored. Moreover, at high temperatures above 1100℃, the strip steel has extremely strong thermal radiation output in the mid-wave infrared band, with a signal intensity several times that of any possible ambient stray light, resulting in an extremely high signal-to-noise ratio. Therefore, this embodiment can effectively avoid ambient light interference by setting the preset infrared band to 3.7-4.1μm. The collected thermal radiation signal can be the original grayscale image data, which is then converted into a two-dimensional temperature field by an image processing algorithm, where the grayscale value of each pixel corresponds to the temperature of the strip surface at that location.

[0027] Next, the spatial locations that meet a preset temperature threshold are extracted from the two-dimensional temperature field as the actual coordinates of the blackening points. For example, a specific temperature threshold (such as 550℃) can be set, and then an image processing algorithm can be used to identify all pixels in the two-dimensional temperature field whose temperature is equal to or close to that threshold. The spatial coordinates of these pixels will be recorded and used as the actual coordinates of the blackening points.

[0028] Subsequently, laser stripes are projected onto the surface of the cooled strip, and images of the stripes modulated by the surface topography are acquired. This embodiment can use a high-power laser to project a series of parallel or intersecting laser stripes onto the strip surface. Simultaneously, an image acquisition component is positioned above or to the side of the strip to capture, in real-time, images of the laser stripes modulated by the strip surface topography. These images contain geometric information about the three-dimensional topography of the strip surface.

[0029] Then, three-dimensional topographic data of the cooled strip surface is obtained from the stripe images, and strip deformation features are extracted based on the three-dimensional topographic data. The obtained stripe images can be analyzed using image processing and computer vision algorithms. For example, using the principle of triangulation, the deformation of the stripes is converted into three-dimensional coordinates of each point on the strip surface. These three-dimensional coordinates constitute the three-dimensional topographic data of the strip surface. Based on this three-dimensional topographic data, strip deformation features can be further calculated. For example, by fitting a reference plane and calculating the normal deviation of each point from the reference plane, defects such as flatness, waviness, or warping of the strip can be quantified.

[0030] Next, the coordinate deviation is calculated based on the actual coordinates of the blackened point and the preset blackened point coordinates. The preset blackened point coordinates are the ideal blackened point positions set according to process requirements. By comparing the actually detected blackened point coordinates with the preset coordinates, the deviation between the two can be calculated. This deviation reflects the difference between the current cooling rate and the target cooling rate.

[0031] Finally, based on the coordinate deviation and strip deformation characteristics, a preset coupling control logic generates cooling intensity adjustment commands for different cooling zones, driving the actuators corresponding to those zones to adjust the cooling medium flow rate. The preset coupling control logic can be a pre-trained temperature-strip shape coupling control model that comprehensively considers the longitudinal deviation of the blackening point position and the lateral characteristics of the strip deformation to generate refined cooling intensity adjustment commands. For example, if the blackening point position deviates from the preset position, the overall cooling medium flow rate needs to be adjusted; if there are local strip defects, the cooling medium flow rate corresponding to the defect area needs to be adjusted differentially. These commands are then sent to the corresponding actuators, such as water valves, to adjust the cooling medium (e.g., water) flow rate in real time, thereby achieving closed-loop control of the strip cooling process.

[0032] The following example provides a more detailed explanation of the above technical solution: Assume a rapid cooling section in the normalizing process of grain-oriented silicon steel, where the cooling strip is running at high speed. Traditional cooling control methods rely on operators visually observing the location of blackening points and manually adjusting the cooling water flow. However, due to the large amount of steam mist generated when the high-temperature steel plate encounters water, and fluctuations in ambient light intensity, the determination of the blackening point location is extremely inaccurate, thus affecting the cooling rate and strip shape control.

[0033] To address this issue, this application proposes a strip cooling control method. First, an infrared thermal imager is installed above the quench section, configured to operate in a preset infrared band of 3.7-4.1 μm. As the cooling strip passes through the quench section, the infrared thermal imager acquires the thermal radiation signal from the strip surface in real time. Because this band is insensitive to ambient light and can penetrate a certain concentration of vapor, the acquired thermal radiation signal accurately reflects the surface temperature of the strip. These thermal radiation signals are then transmitted to an industrial control computer and converted into a two-dimensional temperature field. For example, the industrial control computer can convert the radiation intensity of each pixel into a corresponding temperature value according to Planck's radiation law, thereby generating a two-dimensional matrix containing the temperature distribution of the strip surface.

[0034] Next, the industrial control computer extracts the actual coordinates of the blackening points from the two-dimensional temperature field. Specifically, a preset temperature threshold, such as 550℃, is considered the critical temperature for phase transformation in the strip. The industrial control computer searches for all pixels in the two-dimensional temperature field with a temperature value equal to 550℃ and uses the spatial coordinates of these points as the actual coordinates of the blackening points. For example, the median value of the two-dimensional temperature field can be extracted along the strip's running direction to obtain a one-dimensional temperature curve, and then the spatial locations with temperature values ​​equal to the preset temperature threshold can be found from this curve.

[0035] Meanwhile, on the other side of the cooled strip, a high-power laser projects a series of monochromatic laser stripes onto the strip surface. An image acquisition unit (e.g., an industrial camera equipped with a narrow-bandpass filter corresponding to the laser wavelength) acquires images of the stripes modulated by the strip surface topography in real time.

[0036] Subsequently, the industrial control computer acquires the three-dimensional topographic data of the cooled strip surface based on the collected stripe images. For example, the continuous skeleton lines of the laser stripes are extracted using the Steger algorithm or the gray-scale centroid method, and each pixel on the skeleton line is converted into corresponding three-dimensional coordinates using the principle of triangulation. These three-dimensional coordinates are integrated into three-dimensional point cloud data, representing the three-dimensional topography of the strip surface. Based on this three-dimensional topographic data, the industrial control computer further extracts plate deformation features. For example, a reference plane is obtained by performing plane fitting on the three-dimensional point cloud data, and then the normal deviation of each point relative to the reference plane is calculated. The height deviation sequences of the edge region, the middle region, and the local region are extracted from the data, and then the root mean square deviation or peak-valley value of each region is calculated as the plate deformation features.

[0037] After obtaining the actual coordinates of the blackening point and the plate deformation characteristics, the industrial control computer compares the actual coordinates of the blackening point with the preset blackening point coordinates and calculates the coordinate deviation. For example, if the actual blackening point is located further back than the preset position, it indicates that the cooling rate is slower, and the coordinate deviation is positive; otherwise, it is negative.

[0038] Finally, the industrial control computer generates cooling intensity adjustment commands for different cooling zones based on coordinate deviation and strip deformation characteristics through preset coupled control logic. This coupled control logic comprehensively considers the deviation of the longitudinal cooling rate and transverse strip shape defects. For example, if the coordinate deviation indicates a slow cooling rate, it generates a command to increase the overall cooling medium flow rate; if the strip deformation characteristics indicate the presence of left-side waviness, it generates a command to reduce the cooling intensity of the left-side cooling zone. These cooling intensity adjustment commands are then sent to the actuators corresponding to different cooling zones in the quench section, such as water valves, to adjust the flow rate of the cooling medium (such as water) in real time. In this way, the system can achieve closed-loop, refined control of the strip cooling process, not only accurately controlling the blackening point at the target location but also effectively suppressing and correcting strip shape defects.

[0039] The overall technical concept of this application makes a significant contribution to the prior art. Traditional methods for controlling strip cooling primarily rely on operator visual observation. This method is highly susceptible to fluctuations in vapor mist and ambient light, leading to inaccurate determination of the location of blackening points, which in turn results in imprecise cooling rate control and strip shape defects. For example, when ambient light intensity fluctuates, the boundary of the perceived "blackened" area shifts significantly. This shift does not correspond to the actual temperature distribution, making it impossible for operators to accurately judge the cooling status.

[0040] In contrast, this application introduces mid-wave infrared thermal imaging technology to directly acquire the thermal radiation signal emitted by the strip itself, constructing a realistic temperature field distribution. The preset infrared band is 3.7-4.1 μm, which is insensitive to ambient light and exhibits extremely strong thermal radiation output from the strip in this band, resulting in a very high signal-to-noise ratio. This fundamentally eliminates the interference of ambient light fluctuations on the determination of the blackening point location, ensuring that the location of the blackening point is entirely determined by the physical temperature of the strip itself, providing a true and reliable measurement benchmark.

[0041] Furthermore, this application incorporates laser three-dimensional shape detection technology. By projecting laser stripes onto the strip surface and acquiring modulated images, three-dimensional shape data of the strip surface is obtained, and strip deformation characteristics are extracted. For example, by installing a narrow-bandpass filter corresponding to the fluctuation at the center of the monochromatic laser stripes on the image acquisition component, broadband ambient light scattered by vapor and self-emission from the high-temperature strip can be effectively blocked, ensuring the accuracy of the three-dimensional shape data. This enables the system to capture minute deformations of the strip during the cooling process in real time, providing a direct geometric basis for strip shape control.

[0042] Ultimately, this application, through a pre-defined coupled control logic, integrates the coordinate deviation of the blackening point and the plate deformation characteristics to generate cooling intensity adjustment commands for different cooling zones. This coupled control logic enables refined and differentiated adjustment of the cooling medium flow rate, not only correcting the longitudinal cooling rate but also improving the plate shape through lateral differentiated adjustments. For example, when a left-side edge wave is detected, the system can automatically generate a command to reduce the cooling intensity of the left-side cooling zone, thereby effectively suppressing the generation of edge waves. This closed-loop automated cooling control system represents a leap from manual visual observation to physical quantitative monitoring, significantly improving the accuracy of cooling rate control and the stability of plate shape control. It overcomes the problems of blind adjustment and significant susceptibility to environmental factors in traditional methods, and has significant technological value.

[0043] In some preferred embodiments, the thermal radiation signal is acquired by an infrared thermal imager, and the strip cooling control method further includes steps performed before step S1: A1. Use the two-point calibration method to obtain the calibration gain and calibration bias coefficient of each pixel in the infrared thermal imager; Step S1 includes: S11. Obtain the thermal radiation signal of the cooled strip steel in a preset infrared band; S12. Correct the thermal radiation signal based on all calibration gains and all calibration bias coefficients; S13. Using a pre-built prior model of steam concentration spatial distribution, the estimated steam concentration distribution is obtained based on the current cooling medium flow rate corresponding to different cooling zones. S14. For each pixel in the infrared thermal imager, obtain the steam transmittance of the pixel based on the geometric optical path from the strip position corresponding to the pixel to the infrared thermal imager and the estimated steam concentration distribution. S15. Using the Biber-Lambert law, attenuation compensation of thermal radiation signal is performed based on vapor transmittance. S16. Obtain the current background radiation value of the reference area and remove the current background radiation value from the thermal radiation signal; the reference area is a region located in the steam-free zone of the quenching section where the temperature is known. S17. Convert the compensated thermal radiation signal into a two-dimensional temperature field.

[0044] An infrared thermal imager is a device that detects the infrared radiation emitted by an object and converts it into a visible image. It can employ uncooled focal plane array detectors (such as microbolometer arrays) or cooled detectors (such as mercury cadmium telluride (MCT) or indium antimonide (InSb) detectors). The calibration gain and bias coefficient of each pixel in the infrared thermal imager are obtained using a two-point calibration method. This method involves pointing the infrared thermal imager at two uniform blackbody radiation sources of known temperature, acquiring two images respectively, and calculating the gain and bias coefficients for each pixel based on these two images to eliminate inconsistencies in the response of pixels in the detector array. Acquiring the thermal radiation signal of cooled strip steel in a preset infrared band refers to the real-time reception and digitization of the thermal radiation energy emitted by the strip steel surface in a specific infrared band (3.7-4.1 μm) by the infrared thermal imager. This signal is usually represented in the form of raw grayscale values ​​or digital counts. Step S12 aims to eliminate the non-uniformity error inherent in the infrared thermal imager. By correcting the original signal pixel by pixel, a signal that is non-uniformly corrected and closer to the true radiation intensity is obtained. Using a pre-built a priori model of steam concentration spatial distribution, the estimated steam concentration distribution is obtained based on the current cooling medium flow rate corresponding to different cooling zones. This model can be a complex three-dimensional model based on fluid dynamics simulation or a simplified empirical model. By inputting the current cooling medium flow rate of each cooling zone, the model can output the steam concentration distribution at different spatial locations within the quench section in real time. Step S14 is equivalent to establishing a geometric optical path from a point on the strip surface to the corresponding pixel, and integrating the steam concentration along this optical path to obtain the integrated optical thickness on the optical path, thereby calculating the steam transmittance. The Bieber-Lambert law describes the attenuation law of light propagating in a medium. Therefore, this embodiment can use the Bieber-Lambert law to inversely calculate the intensity of the original thermal radiation signal without steam attenuation based on the steam transmittance (attenuation compensation for the thermal radiation signal). The compensation method can be by dividing the original signal by the transmittance. Step S16 involves acquiring the radiation signal from the reference area to obtain the background radiation value under the current environment, and subtracting it from the thermal radiation signal after steam attenuation compensation. The compensated thermal radiation signal is then converted into a two-dimensional temperature field. This conversion is typically based on Planck's radiation law and takes into account the emissivity parameters of the strip steel material. The true temperature value corresponding to each pixel is calculated pixel by pixel from the corrected and compensated radiation signal value.

[0045] This scheme introduces a multi-dimensional signal correction and compensation mechanism to construct a complete data processing link from hardware calibration to environmental interference elimination, aiming to improve the accuracy and robustness of temperature field measurement. First, before step S1, the method uses a two-point calibration method to calibrate the infrared thermal imager, obtaining the calibration gain and calibration bias coefficient of each pixel. This preprocessing step is fundamental to ensuring the accuracy of subsequent temperature measurements. It eliminates the pixel response inconsistencies caused by differences in the manufacturing process of the infrared thermal imager, making the original thermal radiation signal spatially uniform and accurate. Subsequently, after acquiring the thermal radiation signal of the cooled strip steel in the preset infrared band, the system immediately corrects the thermal radiation signal based on all calibration gains and all calibration bias coefficients. This correction process directly affects the original data, correcting the non-ideal response of the detector itself and providing a more reliable input for subsequent environmental interference compensation. To cope with the complex and variable steam environment in the quenching section, this method further introduces a dynamic compensation mechanism for steam interference. Specifically, the system uses a pre-constructed prior model of the spatial distribution of steam concentration, combined with the current cooling medium flow rate, to obtain the estimated steam concentration distribution in real time. For each pixel in the infrared thermal imager, the steam transmittance of that pixel is calculated based on the geometric optical path from its corresponding strip position to the imager and the estimated steam concentration distribution. Subsequently, the Biber-Lambert law is used to compensate for the attenuation of the thermal radiation signal based on the steam transmittance. This series of steps forms a closed-loop steam interference compensation process, which can dynamically capture and quantify the absorption and scattering effects of steam on infrared radiation, effectively restoring the true radiation intensity after attenuation by steam, thus solving the temperature measurement deviation caused by steam concentration fluctuations. Simultaneously, to eliminate background thermal noise generated by the high-temperature environment in the quenching section, this method also includes acquiring the current background radiation value of a reference area and removing it from the thermal radiation signal. This reference area is set as a region located in the steam-free zone of the quenching section with a known temperature, ensuring the accuracy of the background radiation value. By stripping away the background radiation, the final thermal radiation signal can more purely reflect the radiation characteristics of the strip itself. Finally, the thermal radiation signal, after all the above corrections and compensations, is converted into a two-dimensional temperature field. Through the progressive steps described above, this scheme achieves in-depth cleaning and reconstruction of infrared thermal radiation signals, enabling the final converted two-dimensional temperature field to accurately reflect the actual temperature distribution of the strip steel. Compared to the basic scheme that directly converts thermal radiation signals into a two-dimensional temperature field, this significantly improves the accuracy and robustness of temperature field measurement, especially in complex industrial environments with inherent non-uniformity of infrared thermal imagers, steam interference, and background radiation.Therefore, the actual coordinates of the blackening point and the plate deformation characteristics extracted based on this precise temperature field will be more realistic and reliable, providing a solid data foundation for strip cooling control. This effectively overcomes the impact of environmental interference on measurement accuracy in complex industrial environments, making the determination of the blackening point location no longer subject to environmental factors, and providing a key guarantee for achieving precise and uniform cooling control.

[0046] In some preferred embodiments, step S17 includes: S171. Perform time-domain median filtering on the compensated thermal radiation signal of multiple consecutive frames. S172. Convert the current frame thermal radiation signal, after completing the time-domain median filtering, into a two-dimensional temperature field.

[0047] The thermal radiation signal after continuous multi-frame compensation refers to multiple thermal radiation image data acquired continuously in time after a series of preprocessing steps (such as non-uniformity correction (using calibration gain and calibration bias coefficient to correct the thermal radiation signal), attenuation compensation, and background radiation removal). These signals are essentially radiation values ​​captured by an infrared thermal imager in a preset infrared band (e.g., 3.7-4.1 μm), reflecting the radiation intensity of the strip surface and having undergone preliminary correction. Their function is to provide raw input data for subsequent filtering. Although these data have undergone preliminary compensation, they may still contain transient noise and random interference. Temporal median filtering is a nonlinear digital filtering technique. Its principle is to sort multiple data points (in this case, the radiation values ​​at the same pixel location in multiple consecutive frames) within a sliding time window and select the median value as the output value for that pixel location. The main function of this filter is to effectively suppress impulse noise and random noise, while also preserving the edge information of the signal and avoiding blurring. A possible implementation of temporal median filtering is sliding window median filtering, which involves maintaining a buffer containing N frames of image data. For each pixel in each frame, the median of the N historical values ​​at the corresponding position in the buffer is calculated as the output of that pixel in the current frame. The thermal radiation signal of the current frame after temporal median filtering refers to the thermal radiation signal at the current moment after processing. Compared to the original compensated signal, it has higher stability and lower noise levels, providing a cleaner radiation intensity distribution that more closely approximates the real physical state, laying the foundation for accurate conversion into a two-dimensional temperature field.

[0048] This solution effectively enhances the anti-interference capability of the temperature field construction by introducing temporal median filtering technology. After acquiring the compensated thermal radiation signal (non-uniformity correction, steam attenuation compensation, and background radiation removal), the system does not directly convert the single-frame signal into a temperature field. Instead, it utilizes redundant information in the time dimension to smooth the random noise in the single-frame signal. Specifically, the system continuously acquires multiple frames of compensated thermal radiation signals and performs temporal median filtering on the radiation values ​​at the same pixel location in these consecutive frames. The core of this process is that the median filter can effectively identify and eliminate abnormal pulse signals or outliers caused by instantaneous changes in steam concentration, sensor circuit noise, or environmental electromagnetic interference. These outliers usually appear as isolated peaks or valleys in the time series, and median calculation can ignore these extreme values, thus obtaining a more stable and realistic radiation intensity distribution. Based on this, the current frame thermal radiation signal, after temporal smoothing, is converted into a two-dimensional temperature field. This processing ensures that the constructed two-dimensional temperature field has higher continuity and accuracy, avoiding local jumps or noise points caused by single-frame noise. In this way, this application further optimizes the conversion process from thermal radiation signal to two-dimensional temperature field based on the aforementioned basic scheme. The basic scheme has significantly improved the quality and anti-interference capability of the thermal radiation signal from both physical and signal processing perspectives through methods such as selection of specific infrared bands, non-uniformity correction, vapor attenuation compensation, and background radiation removal. However, even after these refined compensations, transient and random noise may still exist. The scheme in this application further smooths the signal in the time dimension after these preprocessing steps, effectively adding temporal anti-interference capability on top of spatial and spectral compensation. This multi-dimensional, multi-level signal optimization strategy makes the final two-dimensional temperature field more stable and reliable, providing a high-quality data foundation for the subsequent accurate extraction of blackening point coordinates. This significantly improves the accuracy and stability of the entire strip cooling control method, avoiding misjudgments and inaccurate control caused by measurement errors.

[0049] In some preferred embodiments, step S2 includes: S21. Extract the median value of the two-dimensional temperature field along the running direction of the cooling strip steel to obtain a one-dimensional temperature curve along the running direction of the cooling strip steel. S22. The spatial position in the one-dimensional temperature curve where the temperature value is equal to the preset temperature threshold is taken as the actual coordinate of the blackening point.

[0050] Median extraction of the two-dimensional temperature field along the running direction of the cooled strip involves statistically processing the two-dimensional temperature field data to eliminate local outliers and noise interference. Median extraction is a nonlinear filtering method that sorts all data within a sliding window and selects the median value as the output. This method effectively suppresses impulse noise while preserving image edge information. The one-dimensional temperature curve is a simplified representation of the temperature distribution of the strip along its running direction. This curve more clearly shows the temperature decrease process of the strip from the inlet to the outlet of the quench section (running direction), facilitating subsequent blackening point location. The preset temperature threshold is a specific temperature value pre-set according to process requirements and material characteristics to define key state points in the strip cooling process. For example, it can be set as the target temperature for the strip to undergo a phase change or achieve a specific cooling effect. This threshold is the benchmark for determining the blackening point. The actual coordinates of the blackening point refer to the precise spatial location when the surface temperature of the strip exactly reaches the preset temperature threshold. This coordinate is an important parameter for feedback adjustment of the cooling control system, and its accuracy directly affects the cooling effect and product quality.

[0051] This scheme extracts the median value of the two-dimensional temperature field along the running direction of the cooling strip, thus obtaining a one-dimensional temperature curve along the same direction. This processing method effectively reduces the complexity of two-dimensional temperature information into more easily analyzable one-dimensional data. Furthermore, the smoothing characteristic of median extraction significantly suppresses outliers caused by local noise, steam interference, or differences in surface oxide scale that may exist in the two-dimensional temperature field. Based on this, by searching for spatial locations on the one-dimensional temperature curve where the temperature value equals a preset temperature threshold, the actual coordinates of the blackening point can be determined more stably and robustly. This method avoids the uncertainty caused by directly judging from a noisy two-dimensional temperature field, making the determination of the blackening point location more accurate and reliable. The accurate actual coordinates of the blackening point are then used to calculate the coordinate deviation with the preset blackening point coordinates. Combined with the plate deformation characteristics, a cooling intensity adjustment command is generated through coupled control logic, providing a more accurate feedback signal for the entire strip cooling control system. This ensures precise adjustment of the cooling medium flow rate, thereby improving the stability and uniformity of the cooling rate control.

[0052] In some preferred embodiments, the laser stripe is a monochromatic laser stripe projected by a high-power laser, and the stripe image is acquired by an image acquisition component, which is equipped with a narrow-band pass filter corresponding to the center fluctuation of the monochromatic laser stripe.

[0053] A high-power laser is a device capable of outputting a high-intensity, high-energy laser beam. Its function is to ensure that the laser stripes projected onto the surface of a cooled steel strip have sufficient brightness to remain clearly visible against the background of the intense thermal radiation from the high-temperature steel strip itself, thus providing a high-quality raw signal for subsequent image acquisition and processing. High-power lasers can be implemented using various technologies. For example, semiconductor laser arrays can be used, integrating multiple laser diodes to achieve high power output; alternatively, fiber lasers can be used, which offer excellent beam quality and high conversion efficiency. Monochromatic laser stripes refer to light stripes formed by a single wavelength or an extremely narrow wavelength range of laser light. Their characteristics include high spectral purity and energy concentrated at a specific wavelength. This characteristic makes them easy to separate from complex backgrounds using optical filtering, thereby improving the signal-to-noise ratio. A narrow-bandpass filter is an optical element designed to allow only light within a specific wavelength range to pass through while blocking other wavelengths. Its center wavelength is strictly matched to the wavelength of the laser stripe to effectively suppress non-target wavelength light. Narrow-bandpass filters can be in the form of dielectric film interference filters or absorption filters. Their selection should take into account factors such as transmittance, cutoff depth, bandwidth, and environmental adaptability.

[0054] This solution constructs a high signal-to-noise ratio laser stripe imaging system by combining a high-power laser with a narrow-bandpass filter, aiming to solve image interference problems under complex working conditions from an optical physics perspective. The high-power laser, as the light source, ensures that the laser stripes projected onto the surface of the cooled steel strip have sufficient energy intensity, maintaining a clear outline even against the backdrop of intense thermal radiation from the high-temperature steel strip itself, providing high-quality raw images for subsequent morphological feature extraction. The core anti-interference measure is the addition of a narrow-bandpass filter to the image acquisition component, corresponding to the center wavelength of the monochromatic laser stripe. Specifically, this filter effectively blocks ambient light other than the laser wavelength, the thermal radiation from the cooled steel strip itself, and water mist scattering light, allowing only specific wavelengths of light from the laser stripe to pass through. This wavelength-selective filtering physically isolates the laser stripe signal, which is originally mixed with strong background noise, ensuring that the acquired stripe image has extremely high contrast, thus laying a solid foundation for subsequent accurate calculation of three-dimensional morphological data. This design enables the acquisition of clear and accurate stripe images even under harsh conditions such as high temperatures, water mist, and stray light when projecting laser stripes onto the surface of the cooled steel strip and acquiring stripe images modulated by the surface topography. This lays a solid foundation for the subsequent accurate acquisition of three-dimensional topographic data of the cooled steel strip surface. By enhancing signal strength and suppressing noise, this scheme significantly improves the quality of the stripe images, thereby ensuring the accuracy of the three-dimensional topographic data reconstruction and making the subsequent extraction of plate deformation features more reliable.

[0055] In some preferred embodiments, step S4 includes: S401. Using morphological opening operations, isolated micro-point scattered noise around the laser stripes in the stripe image is removed. S402. Using the Steger algorithm or the gray-scale centroid method, locate the gray-scale centroid along the normal direction of the laser stripe to obtain the continuous skeleton line of the laser stripe. S403. Using a preset continuous length threshold and a directional filter consistent with the main direction of the laser stripe, stray signals that do not conform to the projection pattern in the continuous skeleton line are removed. S404. For each pixel on the continuous skeleton line, construct a spatial ray that starts from the pixel and passes through the pixel's coordinates on the image plane. Then, use the coordinates of the intersection of the spatial ray and the known spatial laser plane as the three-dimensional coordinates of the pixel. S405. Integrate all three-dimensional coordinates into a three-dimensional coordinate sequence, and use this three-dimensional coordinate sequence as the cross-sectional profile of the current frame; S406. The continuous frame cross-sectional profiles are spatially superimposed along the running direction of the cooling strip steel to obtain three-dimensional point cloud data, and the three-dimensional point cloud data is used as the three-dimensional topography data of the surface of the cooling strip steel. S407. Denoise the 3D point cloud data; S408. Using the least squares method or a robust fitting algorithm, perform plane fitting on the denoised 3D point cloud data to obtain the reference plane. S409. Calculate the normal deviation of each spatial point in the three-dimensional point cloud data relative to the reference plane point by point, and integrate all normal deviations into a two-dimensional height deviation matrix. S410. Extract the height deviation sequences of the edge region, the middle region, and the local region from the two-dimensional height deviation matrix. Then, calculate the root mean square deviation or peak-valley value corresponding to the edge region, the middle region, and the local region according to the height deviation sequences of the edge region, the middle region, and the local region, and use all root mean square deviations or all peak-valley values ​​as plate deformation characteristic quantities.

[0056] First, morphological opening operations are used to remove isolated micro-bright spot scattering noise scattered around the laser stripes in the fringe image. Morphological opening is an image processing technique that, through a combination of erosion and dilation operations, can effectively remove small-sized, isolated bright spot noise from the image while preserving the structure of larger-sized areas. Its role is to preprocess the fringe image to eliminate micro-bright spots caused by vapor scattering, ambient stray light reflection, or sensor noise, ensuring the accuracy of subsequent fringe extraction.

[0057] Subsequently, the Steger algorithm or the gray-level centroid method is used to locate the gray-level centroid along the normal direction of the laser stripes to obtain the continuous skeleton lines of the laser stripes. Obtaining the continuous skeleton lines of the laser stripes is a crucial step in 3D reconstruction, as it represents the precise center position of the laser stripes on the image. The Steger algorithm is a sub-pixel edge detection algorithm based on Hessian matrix eigenvalue analysis, capable of accurately locating the center of high-contrast linear structures in an image, especially suitable for stripes with uneven width. The gray-level centroid method determines the center position by calculating the centroid of the gray-level distribution across the cross-section of the laser stripes. This method has relatively low computational cost and is robust to noise.

[0058] Next, using a preset continuous length threshold and a directional filter aligned with the principal direction of the laser stripes, stray signals that do not conform to the projection rules in the continuous skeleton lines are removed. Removing stray signals ensures that only genuine laser stripe skeleton lines are used for 3D reconstruction, avoiding artifact interference caused by vapor scattering, etc. The preset continuous length threshold sets a minimum length; only skeleton line segments exceeding this threshold are considered valid laser stripes, thus filtering out short, discontinuous noise segments. The directional filter aligned with the principal direction of the laser stripes, such as a Gabor filter or a directional gradient filter, enhances image features parallel to the principal direction of the laser stripes while suppressing noise and artifacts in other directions, further filtering out skeleton lines that conform to the physical projection rules.

[0059] Step S404 is the core of the transformation from 2D image coordinates to 3D spatial coordinates. After precise calibration, the image acquisition component's internal and external parameters are known. For each pixel on the skeleton line, a spatial ray can be constructed from the camera's optical center and passing through that pixel, based on the camera's intrinsic parameters and the pixel's coordinates on the image plane. Simultaneously, the equation of the laser plane projected by the laser in the world coordinate system is also known. By calculating the intersection of this spatial ray and the known laser plane, the pixel's coordinates in the real 3D space can be obtained. This method, based on the principle of triangulation, is a commonly used technique in structured light 3D reconstruction.

[0060] All three-dimensional coordinates are integrated into a three-dimensional coordinate sequence, which is then used as the cross-sectional profile of the current frame. After these steps, each pixel on the laser stripe skeleton line is assigned its corresponding three-dimensional spatial coordinates. Arranging these discrete three-dimensional coordinate points according to their order on the skeleton line forms a three-dimensional coordinate sequence. This sequence precisely describes the cross-sectional shape formed by the laser stripe on the strip surface, i.e., the cross-sectional profile of the strip at the current moment.

[0061] Continuous frame cross-sectional profiles are spatially superimposed along the running direction of the cooling strip to obtain three-dimensional point cloud data, which is then used as the three-dimensional topographic data of the cooling strip surface. The strip moves continuously during the cooling process. By continuously acquiring laser stripe images and obtaining the cross-sectional profile of each frame, combined with the strip's running speed information, these continuous two-dimensional cross-sectional profiles can be precisely superimposed in three-dimensional space. This superposition operation stitches together the cross-sectional data acquired at different times, thereby constructing complete three-dimensional point cloud data for a specified area of ​​the strip surface. This point cloud data contains all height information of the strip surface in both width and length directions, serving as the original data characterizing the three-dimensional topography of the strip.

[0062] The 3D point cloud data undergoes denoising processing. Since the original acquired 3D point cloud data may contain outliers or noise points due to measurement errors, environmental interference, or sensor limitations, this embodiment can eliminate these inaccurate points through denoising processing, improving the quality of the point cloud data and the accuracy of subsequent processing. Common denoising methods include statistical filtering, which identifies and removes outliers by analyzing the statistical characteristics of each point's neighborhood; or radius filtering, which removes points whose number of neighboring points within a specified radius is less than a certain threshold. A reference plane is used to evaluate deviations in the surface topography of the strip. Least squares is a commonly used data fitting method that determines the plane equation by minimizing the sum of squared distances from all points in the point cloud to the fitting plane, thus obtaining an optimal fitting plane. Robust fitting algorithms (such as RANSAC) are more robust fitting methods that can estimate model parameters from data containing a large number of noise or outliers, fit the model by iteratively selecting a subset of data, and evaluate the number of interior points in the model, thus obtaining a reference plane that is insensitive to outliers.

[0063] After determining the reference plane, it is necessary to quantify the degree to which each point on the strip surface deviates from the ideal flatness. Normal deviation refers to the vertical distance from each 3D point in the point cloud to the fitted reference plane; this distance can be positive or negative. Arranging all these normal deviation values ​​according to their spatial relationship on the strip surface forms a two-dimensional matrix, where each element of the matrix represents the height deviation at the corresponding position on the strip surface. This two-dimensional height deviation matrix visually reflects the undulation of the strip surface.

[0064] To quantify different types of strip shape defects, regional analysis of the height deviation matrix is ​​necessary. Edge regions typically refer to specific width ranges on either side of the strip's width direction, central regions refer to the central area of ​​the strip's width direction, and local regions can be small areas with drastic height deviation variations. The corresponding height deviation values ​​are extracted from these regions to form their respective sequences. The root mean square deviation (RMSD) is a statistical measure used to assess the dispersion of the data sequence, reflecting the average level of height fluctuations within a region. Peak-valley values ​​represent the difference between the maximum and minimum height deviations within a region, reflecting the extreme degree of shape variation within that region. By calculating the RMS or peak-valley values ​​for these regions, quantified strip shape deformation characteristics can be obtained, providing precise input for subsequent cooling control.

[0065] This solution achieves the transformation from raw stripe images to high-precision plate-shaped feature quantities through a complete image processing and 3D reconstruction workflow. In the image acquisition stage, the use of a high-power laser to project monochromatic laser stripes, coupled with a narrow-bandpass filter corresponding to the central fluctuation of the monochromatic laser stripes on the image acquisition component, significantly reduces stray light from the broadband ambient light scattered by vapor and the self-emission of high-temperature steel strips entering the camera. This significantly improves the signal-to-noise ratio and clarity of the laser stripe image, laying the foundation for subsequent precise processing. Based on this, the acquired stripe image is first processed using morphological opening operations to remove isolated micro-bright spot-like scattering noise scattered around the laser stripes, ensuring the purity of the image in subsequent processing. Then, high-precision gray-scale centroid positioning is performed along the normal direction of the laser stripes using the Steger algorithm or the gray-scale centroid method, accurately extracting the continuous skeleton lines of the laser stripes and overcoming the influence of stripe width variations on positioning accuracy. To further eliminate interference from non-laser projection, a preset continuous length threshold and a directional filter aligned with the main direction of the laser stripes are used to remove stray signals from the continuous skeleton lines that do not conform to the projection pattern, retaining only the laser skeleton lines with true geometric continuity. In the 3D reconstruction stage, for each pixel on the continuous skeleton line, a spatial ray is constructed from that pixel and passing through its coordinates on the image plane. The coordinates of the intersection of this spatial ray and the known spatial laser plane are then used as the corresponding 3D coordinates of that pixel, achieving a precise conversion from 2D image coordinates to true 3D spatial coordinates. All these 3D coordinates are integrated into a 3D coordinate sequence, serving as the cross-sectional profile of the current frame. Since the cooling strip is continuously moving, by spatially superimposing the cross-sectional profiles of consecutive frames along the running direction of the cooling strip, complete 3D point cloud data covering a specified area can be constructed, serving as the 3D topographic data of the cooling strip surface. To improve the reliability of the topographic data, the original 3D point cloud data is denoised to eliminate measurement errors and outliers. Subsequently, the denoised 3D point cloud data is fitted using the least squares method or a robust fitting algorithm to obtain a reference plane representing the ideal flat state. On this reference plane, the normal deviation of each spatial point in the 3D point cloud data relative to the reference plane is calculated point by point, and all normal deviations are integrated into a two-dimensional height deviation matrix, thus transforming complex topographic information into intuitive quantitative data. Finally, height deviation sequences of edge regions, central regions, and local regions are extracted from the two-dimensional height deviation matrix, and the corresponding root mean square deviation or peak-valley values ​​are calculated based on these sequences. These statistics are used as plate deformation feature quantities. This series of steps not only solves the problem of difficult laser stripe extraction in complex environments, but also provides a precise quantitative basis for the subsequent generation of cooling intensity adjustment commands for different cooling regions through regional feature quantity calculation, ensuring the targeting and effectiveness of plate shape control.The above scheme can provide accurate strip deformation characteristics for strip cooling control, thereby enabling the cooling intensity adjustment command to respond more precisely to the actual shape of the strip and effectively avoid production problems caused by strip shape defects.

[0066] In some preferred embodiments, step S6 includes: S61. Analyze whether there are plate shape defects based on the plate shape deformation characteristic quantity. If yes, proceed to step S62; otherwise, proceed to step S63. S62. Using a pre-built mapping table of plate shape defect types and correction directions, the plate shape defect types are mapped to the cooling intensity correction direction of each cooling area, and the lateral difference correction amount of each cooling area is generated based on the plate shape deformation characteristic amount. S63. Generate longitudinal difference correction amount for each cooling zone based on coordinate deviation; S64. When only the longitudinal difference correction amount exists, generate the cooling intensity adjustment command corresponding to each cooling area based on the longitudinal difference correction amount and the feedforward reference amount corresponding to each cooling area. S65. When both longitudinal and lateral differential corrections exist, generate cooling intensity adjustment commands for each cooling zone based on the longitudinal, lateral, and feedforward reference values ​​corresponding to each cooling zone. S66. Send all cooling intensity adjustment commands to the corresponding actuators to drive the actuators corresponding to different cooling zones to adjust the cooling medium flow rate.

[0067] Step S61 aims to evaluate the strip deformation features extracted from the three-dimensional morphology data of the strip surface to determine whether the current strip has shape defects such as edge waviness, center waviness, or wavy defects. This analysis can be based on a preset threshold comparison. For example, if the root mean square deviation or peak-to-valley value corresponding to the left edge region of the strip is greater than the preset threshold and other root mean square deviations or peak-to-valley values ​​are less than the preset threshold, then a left edge waviness defect is considered to exist; if the root mean square deviation or peak-to-valley value corresponding to the right edge region of the strip is greater than the preset threshold and other root mean square deviations or peak-to-valley values ​​are less than the preset threshold, then a right edge waviness defect is considered to exist; if the root mean square deviation or peak-to-valley value corresponding to the center region of the strip is greater than the preset threshold and other root mean square deviations or peak-to-valley values ​​are less than the preset threshold, then a center waviness defect is considered to exist; if the root mean square deviation or peak-to-valley value corresponding to a local region is greater than the preset threshold and other root mean square deviations or peak-to-valley values ​​are less than the preset threshold, then a wavy defect is considered to exist.

[0068] Using a pre-built mapping table of plate shape defect types and correction directions, the types of plate shape defects are mapped to the cooling intensity correction directions of each cooling region, and the lateral difference correction amount for each cooling region is generated based on the plate shape deformation characteristics. This mapping table can be a lookup table, which predefines the association between different plate shape defect types (e.g., left-side wave, right-side wave, middle wave, swivel, etc.) and the corresponding cooling intensity correction directions (e.g., increasing or decreasing cooling intensity) of the cooling region. For example, when a left-side wave is detected, the mapping table indicates that the cooling intensity of the left cooling region needs to be reduced. The lateral difference correction amount is calculated based on the specific values ​​of the plate shape deformation characteristics (e.g., the degree of deformation) to determine the magnitude of the correction. This correction amount can be converted from the plate deformation characteristic into a specific cooling intensity adjustment amount through a preset functional relationship. Specifically, the calculation formulas for the lateral difference correction amount corresponding to the left edge region, the right edge region, and the center region are similar. The lateral difference correction amount corresponding to the left edge region is: δ_edge_left=K_p_D×ε_edge_left+K_i_D×∫ε_edge_leftdt+K_d_D×dε_edge_left / dt; where δ_edge_left represents the lateral difference correction amount corresponding to the left edge region, K_p_D, Ki_D, and K_d_D are all PID control parameters for plate shape deviation, and ε_edge_left represents the normalized plate shape deviation coefficient corresponding to the left edge region.

[0069] The longitudinal difference correction amount for each cooling zone is generated based on the coordinate deviation. The coordinate deviation reflects the difference between the actual coordinates of the blackening point and the preset blackening point coordinates, directly indicating the speed of the cooling process in the longitudinal direction. For example, when the actual coordinates of the blackening point deviate from the preset coordinates, it indicates that the cooling speed is too fast or too slow. The longitudinal difference correction amount can be calculated based on this coordinate deviation using a feedback control algorithm (e.g., a proportional-integral-derivative (PID) control algorithm) to adjust the overall cooling intensity and bring the blackening point back to the target position.

[0070] When only the longitudinal difference correction exists, the cooling intensity adjustment command for each cooling zone is generated based on the longitudinal difference correction and the feedforward reference value. In this scenario, the system primarily focuses on the longitudinal control of the blackening point location. By superimposing the longitudinal difference correction and the preset feedforward reference value, a cooling intensity adjustment command for each cooling zone is formed. The feedforward reference value can be understood as the initial cooling intensity setting required to maintain the target cooling effect under ideal operating conditions.

[0071] When both the longitudinal deviation correction amount and the transverse deviation correction amount exist simultaneously, a cooling intensity adjustment instruction corresponding to each cooling zone is generated according to the longitudinal deviation correction amount, the transverse deviation correction amount and the feedforward reference amount corresponding to each cooling zone. In this scenario, the system needs to take into account both longitudinal temperature control and transverse strip shape correction. At this time, the final cooling intensity adjustment instruction is the coupling superposition of the feedforward reference amount, the longitudinal deviation correction amount and the transverse deviation correction amount, so as to realize fine and differentiated distribution of the cooling medium flow. All cooling intensity adjustment instructions are sent to the corresponding actuators, so as to drive the actuators corresponding to different cooling zones to adjust the cooling medium flow. These adjustment instructions may be analog quantity signals (for example, 4-20mA current signals) or digital signals, and are sent via an industrial communication network (for example, Modbus, Profibus) to the actuators corresponding to each cooling zone (for example, electric control valves, variable-frequency water pumps). After receiving the instructions, the actuators will adjust the flow of the cooling medium (for example, water) correspondingly, so as to change the cooling intensity of the corresponding zone. Preferably, the cooling intensity adjustment instruction is the valve opening corresponding to each nozzle quantity, and before sending the cooling intensity adjustment instruction, this embodiment may first perform engineering constraint processing on the cooling intensity adjustment instruction: 1. Amplitude limiting processing: if the valve opening > 100%, the valve opening is set to 100%; if the valve opening < V_min (minimum safe opening, ensuring no dry burning of the nozzles), the valve opening is set to V_min; 2. Change rate limitation: the variation of the valve position opening in adjacent control periods does not exceed the preset maximum change rate, so as to prevent thermal impact on the strip steel caused by sudden change of cooling intensity; 3. Fail-safe state: when the system detects that the sensor signal is lost or communication is interrupted, each valve automatically maintains the current opening or switches to the preset safe opening, and an alarm is issued to the operator at the same time, This solution achieves closed-loop, refined regulation of the cooling process by introducing coupled control logic of plate shape defect analysis and cooling intensity correction. First, by analyzing plate shape deformation characteristics, the existence of plate shape defects can be determined in real time, thus deciding whether lateral difference correction is needed. This judgment mechanism ensures the targeted nature of the control strategy. When plate shape defects exist, a pre-built mapping table is used to convert the defect type into a specific cooling intensity correction direction. The lateral difference correction amount is then calculated based on the plate shape deformation characteristics. This process allows for differentiated adjustment of the flow rate in different cooling zones according to the specific shape fluctuations, effectively suppressing plate shape defects such as edge and center waves. Simultaneously, a longitudinal difference correction amount is generated through coordinate deviation, compensating for longitudinal cooling intensity offsets at blackening point locations, ensuring the temperature control accuracy of the cooling process. When generating cooling intensity adjustment commands, the system executes corresponding command generation logic based on different combinations of longitudinal difference correction, lateral difference correction, and feedforward reference values. This hierarchical and superimposed control method enables the system to simultaneously handle the coupling problem of longitudinal temperature distribution and lateral plate shape distribution, ensuring the scientific and rational allocation of cooling medium flow rate. Ultimately, the system achieves dynamic optimization of the cooling process by driving the actuator. This scheme, combined with techniques such as acquiring the thermal radiation signal of the cooled strip steel in a preset infrared band and converting it into a two-dimensional temperature field, extracting the actual coordinates of the blackening points from the two-dimensional temperature field, projecting laser stripes onto the surface of the cooled strip steel and acquiring stripe images, obtaining three-dimensional morphology data from the stripe images and extracting plate shape deformation characteristics, and calculating coordinate deviations based on the actual coordinates and preset coordinates of the blackening points, makes the measurement of the blackening point position and plate shape deformation characteristics more accurate and reliable. This provides a real and reliable measurement benchmark for subsequent coupled control, overcoming the misjudgment problem caused by changes in ambient light during traditional manual observation, and laying the foundation for achieving refined and differentiated allocation of cooling medium flow rate.

[0072] In some preferred embodiments, the cooling medium is water, and the formula for calculating the feedforward reference quantity is: V_i_base=f(Q_i_feedforward); Where Vi_i_base represents the feedforward reference value corresponding to the i-th cooling zone, and Qi_i_feedforward represents the feedforward flow setting value corresponding to the i-th cooling zone; The formula for calculating the feedforward flow setpoint is: Q_i_feedforward=k_i×P_total / (c_water×ΔT_water); Where k_i represents the length ratio coefficient of the actuator corresponding to the i-th cooling zone, P_total represents the total cooling capacity, c_water represents the specific heat capacity of water, and ΔT_water represents the allowable temperature rise of water. The formula for calculating the total cooling capacity is: P_total=ρ×Cp×h×w×v×(T_in-T_target) / 60; Where ρ represents the density of the cooling strip, Cp represents the specific heat capacity of the cooling strip, h represents the thickness of the cooling strip, w represents the width of the cooling strip, v represents the running speed of the cooling strip, T_in represents the actual temperature of the cooling strip at the inlet of the quench section, and T_target represents the target temperature of the cooling strip at the outlet of the quench section.

[0073] In some preferred embodiments, the infrared thermal imager of this embodiment is equipped with a purging device. The infrared thermal imager is mounted on top of the quench section, and its viewing window (typically a protective window made of infrared-transmitting materials such as germanium, silicon, or sapphire) is directly exposed to the steam environment generated by the high-temperature steel strip and water spray below. Although the selected mid-wave infrared band can effectively penetrate the water vapor film, if contaminants such as liquid water film, scale, dust, or oil adhere to the outer surface of the viewing window, it may cause problems such as infrared radiation attenuation and unevenness, temperature reference drift, and long-term reliability damage. Therefore, the purpose of equipping the infrared thermal imager with a purging device in this embodiment is to form an air curtain barrier on the outer surface of the viewing window by continuously or as needed spraying clean, dry gas, preventing water mist and dust from contacting the viewing window surface, while simultaneously blowing away any adhering trace contaminants, keeping the viewing window clean, and ensuring the long-term stability and accuracy of the infrared temperature measurement optical path. The control logic of the purging device is automatically determined and executed by the industrial control computer based on the operating conditions, and is divided into the following four modes: 1. Normal protective purging (triggering conditions: production line is running (passing through steel belt and spray is on), and the window contamination index is lower than the light contamination threshold, and the spray flow rate has not reached the high flow range): It operates at the first preset power, generating a stable air curtain to maintain the cleanliness of the window, providing basic protection. 2. Enhanced cleaning purging (triggering conditions: window contamination index reaches the heavy contamination threshold; or the total spray flow rate enters the high flow range and the steam generation increases significantly; or the production line switches from shutdown to the initial stage after operation): It operates at the second preset power, which is greater than the first preset power, providing a higher pressure / flow airflow. While maintaining the air curtain, it generates sufficient shear force to forcibly blow away the attached contaminants. After the triggering conditions disappear, the enhanced power is maintained for a preset time (e.g., 30 seconds), and then automatically drops back to the first preset power. 3. Intermittent Purge During Production Shutdown (Triggering Conditions: Production line shutdown, waiting for materials, or maintenance, and spray system turned off): Turns off purging or performs intermittent purging at extremely low power (e.g., purging for 10 seconds every 5 minutes) to conserve air and prevent dust accumulation during periods of inactivity. 4. Fault Alarm (Triggering Conditions: Detection of purging air pressure below the safety threshold): Outputs an alarm signal to the operator interface, prompting manual inspection of the air source. Simultaneously, the infrared thermal imager marks a decrease in the confidence level of the current temperature data to prevent erroneous control decisions due to sudden contamination of the viewing window.

[0074] Please refer to Figure 2 , Figure 2This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device includes a processor 101 and a memory 102. The processor 101 and the memory 102 are interconnected and communicate with each other via a communication bus 103 and / or other forms of connection mechanisms (not shown). The memory 102 stores computer-readable instructions executable by the processor 101. When the electronic device is running, the processor 101 executes these computer-readable instructions to perform the method in any optional implementation of the above embodiments, thereby achieving the following function: Step S1: Acquire the thermal radiation signal of the cooled strip steel in a preset infrared band and convert the thermal radiation signal into a two-dimensional temperature field; the preset infrared band is 3... 0.7-4.1μm; Step S2: Extract the spatial position that meets the preset temperature threshold from the two-dimensional temperature field as the actual coordinates of the blackening point; Step S3: Project laser stripes onto the surface of the cooling strip steel and acquire the stripe image after the surface morphology of the cooling strip steel is modulated; Step S4: Obtain the three-dimensional morphology data of the surface of the cooling strip steel according to the stripe image, and extract the plate deformation feature quantity based on the three-dimensional morphology data; Step S5: Calculate the coordinate deviation according to the actual coordinates of the blackening point and the preset blackening point coordinates; Step S6: Based on the coordinate deviation and the plate deformation feature quantity, generate the cooling intensity adjustment command for different cooling areas through the preset coupling control logic, so as to drive the actuator corresponding to different cooling areas to adjust the cooling medium flow rate.

[0075] This application embodiment also provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it executes the method in any optional implementation of the above embodiments to achieve the following functions: Step S1: Acquire the thermal radiation signal of the cooled strip steel in a preset infrared band and convert the thermal radiation signal into a two-dimensional temperature field; the preset infrared band is 3.7-4.1μm; Step S2: Extract the spatial position that meets the preset temperature threshold from the two-dimensional temperature field as the actual coordinates of the blackening point; Step S3: Project laser stripes onto the surface of the cooled strip steel and acquire the stripe image after the surface morphology of the cooled strip steel is modulated; Step S4: Acquire the three-dimensional morphology data of the surface of the cooled strip steel according to the stripe image, and extract the plate deformation feature quantity based on the three-dimensional morphology data; Step S5: Calculate the coordinate deviation according to the actual coordinates of the blackening point and the preset blackening point coordinates; Step S6: Generate cooling intensity adjustment instructions for different cooling areas through preset coupling control logic according to the coordinate deviation and the plate deformation feature quantity, so as to drive the actuators corresponding to different cooling areas to adjust the cooling medium flow rate. The computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0076] As can be seen from the above, the strip cooling control method, electronic equipment, and storage medium provided in this application acquire infrared thermal radiation signals and perform steam compensation and background removal, combined with laser stripe scanning to obtain three-dimensional morphology, thereby achieving precise positioning of blackening points and real-time monitoring of strip shape characteristics. By dynamically adjusting the cooling intensity through coupled control logic, it effectively solves the problems of large errors in traditional manual observation, severe steam interference, and inaccurate strip shape control. It has the advantages of effectively overcoming steam interference and ambient light fluctuations, achieving real-time, precise, and automated control of the strip cooling process, effectively suppressing strip shape defects, and improving product quality.

[0077] In the embodiments provided in this application, it should be understood that relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0078] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for controlling the cooling of strip steel, characterized in that, The strip cooling control method includes the following steps: S1. Acquire the thermal radiation signal of the cooled strip steel in a preset infrared band, and convert the thermal radiation signal into a two-dimensional temperature field; the preset infrared band is 3.7-4.1μm; S2. Extract the spatial location that meets the preset temperature threshold from the two-dimensional temperature field as the actual coordinates of the blackening point; S3. Project laser stripes onto the surface of the cooling strip steel and acquire the stripe image modulated by the surface morphology of the cooling strip steel; S4. Obtain three-dimensional morphological data of the surface of the cooling strip steel based on the stripe image, and extract plate deformation feature quantities based on the three-dimensional morphological data; S5. Calculate the coordinate deviation based on the actual coordinates of the blackened point and the preset coordinates of the blackened point; S6. Based on the coordinate deviation and the plate deformation characteristic, generate cooling intensity adjustment commands for different cooling zones through preset coupling control logic to drive the actuators corresponding to different cooling zones to adjust the cooling medium flow rate.

2. The strip cooling control method according to claim 1, characterized in that, The thermal radiation signal is acquired by an infrared thermal imager, and the strip cooling control method further includes steps performed before step S1: A1. Obtain the calibration gain and calibration bias coefficient of each pixel in the infrared thermal imager using the two-point calibration method; Step S1 includes: S11. Obtain the thermal radiation signal of the cooled strip steel in a preset infrared band; S12. Correct the thermal radiation signal according to all the calibration gains and all the calibration bias coefficients; S13. Using a pre-built prior model of steam concentration spatial distribution, the estimated steam concentration distribution is obtained based on the current cooling medium flow rate corresponding to different cooling zones. S14. For each pixel in the infrared thermal imager, the steam transmittance of the pixel is obtained based on the geometric optical path from the strip position corresponding to the pixel to the infrared thermal imager and the estimated steam concentration distribution. S15. Using the Bierber-Lambert law, attenuation compensation is performed on the thermal radiation signal based on the vapor transmittance. S16. Obtain the current background radiation value of the reference area and remove the current background radiation value from the thermal radiation signal; the reference area is a region located in the steam-free zone of the quenching section and whose temperature is known. S17. Convert the compensated thermal radiation signal into a two-dimensional temperature field.

3. The strip cooling control method according to claim 2, characterized in that, Step S17 includes: S171. Perform time-domain median filtering on the compensated thermal radiation signal of multiple consecutive frames. S172. Convert the current frame thermal radiation signal, after completing the time-domain median filtering, into a two-dimensional temperature field.

4. The strip cooling control method according to claim 1, characterized in that, Step S2 includes: S21. Extract the median value of the two-dimensional temperature field along the running direction of the cooling strip steel to obtain a one-dimensional temperature curve along the running direction of the cooling strip steel. S22. The spatial position in the one-dimensional temperature curve where the temperature value is equal to the preset temperature threshold is taken as the actual coordinate of the blackening point.

5. The strip cooling control method according to claim 1, characterized in that, The laser stripe is a monochromatic laser stripe projected by a high-power laser. The stripe image is acquired by an image acquisition component, which is equipped with a narrow-band pass filter corresponding to the center fluctuation of the monochromatic laser stripe.

6. The strip cooling control method according to claim 5, characterized in that, Step S4 includes: S401. Using morphological opening operations, remove isolated micro-point scattered noise around the laser stripes in the stripe image; S402. Using the Steger algorithm or the gray-scale centroid method, perform gray-scale centroid positioning along the normal direction of the laser stripe to obtain the continuous skeleton line of the laser stripe. S403. Using a preset continuous length threshold and a directional filter consistent with the main direction of the laser stripe, stray signals that do not conform to the projection pattern in the continuous skeleton line are removed. S404. For each pixel on the continuous skeleton line, construct a spatial ray that starts from the pixel and passes through the coordinates of the pixel on the image plane, and then use the coordinates of the intersection of the spatial ray and the known spatial laser plane as the three-dimensional coordinates of the pixel. S405. Integrate all the three-dimensional coordinates into a three-dimensional coordinate sequence, and use the three-dimensional coordinate sequence as the cross-sectional contour of the current frame; S406. The cross-sectional contours of the consecutive frames are spatially superimposed along the running direction of the cooling strip steel to obtain three-dimensional point cloud data, and the three-dimensional point cloud data is used as the three-dimensional topography data of the surface of the cooling strip steel. S407. Denoise the three-dimensional point cloud data; S408. Using the least squares method or a robust fitting algorithm, perform plane fitting on the denoised 3D point cloud data to obtain the reference plane. S409. Calculate the normal deviation of each spatial point in the three-dimensional point cloud data relative to the reference plane point by point, and integrate all the normal deviations into a two-dimensional height deviation matrix; S410. Extract the height deviation sequences of the edge region, the middle region, and the local region from the two-dimensional height deviation matrix. Then, calculate the root mean square deviation or peak-valley value corresponding to the edge region, the middle region, and the local region according to the height deviation sequences of the edge region, the middle region, and the local region, and use all the root mean square deviations or all the peak-valley values ​​as plate deformation feature quantities.

7. The strip cooling control method according to claim 1, characterized in that, Step S6 includes: S61. Analyze whether there is a plate shape defect based on the plate shape deformation characteristic quantity. If yes, proceed to step S62; otherwise, proceed to step S63. S62. Using a pre-built mapping table of plate shape defect types and correction directions, the plate shape defect types are mapped to the cooling intensity correction directions of each cooling region, and the lateral difference correction amount of each cooling region is generated according to the plate shape deformation characteristic amount. S63. Generate longitudinal difference correction amount for each cooling zone based on the coordinate deviation; S64. When only the longitudinal difference correction amount exists, generate a cooling intensity adjustment command corresponding to each cooling region based on the longitudinal difference correction amount and feedforward reference amount corresponding to each cooling region. S65. When both the longitudinal difference correction amount and the lateral difference correction amount exist simultaneously, generate a cooling intensity adjustment command corresponding to each cooling region based on the longitudinal difference correction amount, the lateral difference correction amount, and the feedforward reference amount corresponding to each cooling region. S66. Send all the cooling intensity adjustment commands to the corresponding actuators to drive the actuators corresponding to different cooling zones to adjust the cooling medium flow rate.

8. The strip cooling control method according to claim 7, characterized in that, The cooling medium is water, and the formula for calculating the feedforward reference quantity is: V_i_base=f(Q_i_feedforward); Where Vi_i_base represents the feedforward reference value corresponding to the i-th cooling zone, and Qi_i_feedforward represents the feedforward flow setting value corresponding to the i-th cooling zone; The formula for calculating the feedforward flow setpoint is: Q_i_feedforward=k_i×P_total / (c_water×ΔT_water); Where k_i represents the length ratio coefficient of the actuator corresponding to the i-th cooling zone, P_total represents the total cooling capacity, c_water represents the specific heat capacity of water, and ΔT_water represents the allowable temperature rise of water. The formula for calculating the total cooling capacity is: P_total=ρ×Cp×h×w×v×(T_in-T_target) / 60; Where ρ represents the density of the cooling strip, Cp represents the specific heat capacity of the cooling strip, h represents the thickness of the cooling strip, w represents the width of the cooling strip, v represents the running speed of the cooling strip, T_in represents the actual temperature of the cooling strip at the inlet of the quench section, and T_target represents the target temperature of the cooling strip at the outlet of the quench section.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory storing computer-readable instructions that, when executed by the processor, perform the steps of the method as described in any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it performs the steps of the method as described in any one of claims 1-8.

Images