Method and device for controlling elongation of thin strip steel

Abstract

This application discloses a method and apparatus for controlling the elongation rate in thin strip steel rolling, relating to the field of steel strip manufacturing technology. The method includes: acquiring real-time processing data during the thin strip steel rolling process, and extracting temperature distribution data, real-time front and rear tension, and real-time rolling force from the real-time processing data; calculating the rheological stress weight along the thickness direction of the thin strip steel based on the temperature distribution data; decoupling the real-time rolling force based on the rheological stress weight and the real-time front and rear tension to obtain a dynamic offset coefficient; inputting the rheological stress weight, the dynamic offset coefficient, and the real-time processing data into a preset mapping model to estimate the neutral point position of the current deformation zone; deriving the true elongation rate of the thin strip steel based on the neutral point position; and dynamically adjusting the elongation rate of the thin strip steel based on the deviation between the preset target elongation rate and the true elongation rate. This application improves the control accuracy of the elongation rate.

Description

Technical Field

[0001] This application relates to the field of steel strip manufacturing technology, specifically to a method and apparatus for controlling the elongation rate of thin strip steel rolling. Background Technology

[0002] In the production of hot-rolled strip steel, finished product strength is one of the core indicators for measuring product quality. Strip steel strength depends not only on its chemical composition but also directly on the combined effects of the degree of plastic deformation, temperature path, and deformation distribution during rolling. Among these factors, elongation is a crucial process characteristic reflecting the actual degree of material deformation, and its control precision directly determines the final microstructure refinement and strength level.

[0003] In the continuous rolling process of hot-rolled thin strip steel, the control of strip elongation depends on the accurate prediction of metal flow behavior in the deformation zone. However, under actual production conditions, the rolled piece is simultaneously subjected to the coupled effects of temperature field evolution, friction state changes, and material rheological property changes during continuous deformation over multiple stands. This causes a significant mismatch in traditional elongation models based on the assumptions of a fixed friction coefficient and constant rheological stress. Related technologies often employ static calibration models or empirical correction methods based on historical data, which struggle to reflect the aforementioned multi-field coupled physical processes in real time, leading to a decrease in the accuracy of elongation control. Summary of the Invention

[0004] To address the technical problem that many related technologies employ static calibration models or empirical correction methods based on historical data, which are insufficient to reflect the aforementioned multi-field coupled physical processes in real time, leading to a decrease in elongation control accuracy, this application provides an elongation control method and apparatus for thin strip steel rolling.

[0005] The specific technical solution adopted is as follows: Acquire real-time processing data during the thin strip steel rolling process, and extract temperature distribution data, real-time front and rear tension, and real-time rolling force from the real-time processing data; Based on temperature distribution data, the rheological stress weight along the thickness direction of the thin strip is calculated. Based on the rheological stress weight and the real-time front and rear tension, the real-time rolling force is decoupled to obtain the dynamic offset coefficient. The dynamic offset coefficient is used to characterize the degree of offset between the friction force and the historical steady-state reference. Input the rheological stress weight, dynamic offset coefficient and real-time processing data into the preset mapping model to estimate the neutral point position of the current deformation zone; Based on the neutral point position, the true elongation of the thin strip steel is derived, and the elongation of the thin strip steel is dynamically adjusted based on the deviation between the preset target elongation and the true elongation.

[0006] In one possible embodiment of this application, the flow stress weight along the thickness direction of the thin strip steel is calculated based on temperature distribution data, including: Based on the outlet and inlet of the deformation zone, an analysis location interval is constructed; Based on the temperature distribution data in the analysis location interval, the real-time temperature value of each layer along the thickness direction of the thin strip steel is determined; The rheological stress weight along the thickness direction of the thin strip steel is calculated by the deviation between each real-time temperature value and the reference temperature value.

[0007] In one possible embodiment of this application, the flow stress weight along the thickness direction of the thin strip steel is calculated by the deviation between each real-time temperature value and a reference temperature value, including: Extract the full-layer temperature of the strip steel from multiple control cycles within a historical time period; The reference temperature value is determined based on the average temperature of each strip across the entire layer. For any given layer, calculate the relative deviation between the real-time temperature value and the reference temperature value; By linearly mapping the relative deviation values ​​of each layer, the rheological stress weights along the thickness direction of the thin strip are obtained.

[0008] In one possible embodiment of this application, a linear mapping is performed on the relative deviation values ​​of each layer to obtain the flow stress weight along the thickness direction of the thin strip steel, including: By performing an inverse linear mapping on the relative deviation values ​​of each layer, the relative stress weights of each layer are obtained. The relative stress weights are normalized to generate a relative stress weight sequence. By weighting and summing the relative stress weights using the values ​​in the relative stress weight sequence, the rheological stress weights along the thickness direction of the thin strip steel are obtained.

[0009] In one possible implementation of this application, the real-time rolling force is decoupled based on the rheological stress weight and the real-time front and rear tension to obtain a dynamic offset coefficient, including: Determine the steady-state reference rolling force for multiple control cycles within a historical time period, as well as the tension values ​​before and after the steady-state reference. Based on the rheological stress weight, the steady-state reference rolling force is corrected for fluctuations to obtain the theoretical reference rolling force; By decoupling the real-time rolling force using the steady-state reference tension value and the real-time tension value, the effective value of the rolling force is obtained. The dynamic offset coefficient is calculated based on the relative fluctuation between the theoretical reference rolling force and the effective value of the rolling force.

[0010] In one possible implementation of this application, a dynamic offset coefficient is calculated based on the relative fluctuation between the theoretical reference rolling force and the effective value of the rolling force, including: The rolling force fluctuation value is calculated based on the deviation ratio between the theoretical reference rolling force and the effective value of the rolling force. By combining the friction coefficient corresponding to the preset steady-state benchmark with the rolling force fluctuation value, the dynamic offset coefficient is obtained.

[0011] In one possible implementation of this application, the rheological stress weight, dynamic offset coefficient, and real-time processing data are input into a preset mapping model to estimate the neutral point position of the current deformation zone, including: Obtain historical steady-state rolling data during the thin strip steel rolling process; Historical steady-state rolling data were processed using linear regression fitting to construct a fixed mapping relationship between input features and neutral point positions in a pre-defined mapping model. Based on a fixed mapping relationship, the benchmark weight coefficients corresponding to the rheological stress weight, dynamic offset coefficient, and real-time processing data are determined respectively. The neutral point position of the current deformation zone is obtained by summing the rheological stress weight, dynamic offset coefficient, and real-time processing data using the benchmark weight coefficient.

[0012] In one possible implementation of this application, the true elongation of the thin strip steel is derived based on the neutral point location, including: Obtain the linear speed of the rolls during the rolling process of thin strip steel; By combining the neutral point position with the roll linear velocity, the true inlet velocity and the true outlet velocity in the deformation zone can be obtained. The true elongation of the thin strip steel is derived based on the difference between the actual inlet velocity and the actual outlet velocity.

[0013] In one possible implementation of this application, the elongation of the thin strip steel is dynamically adjusted based on the deviation between a preset target elongation and the actual elongation, including: Based on the deviation between the preset target elongation rate and the actual elongation rate, the elongation rate control error is determined; The optimal adjustment amount is obtained by dynamically adapting the elongation control error, rheological stress weight, and dynamic offset coefficient through a preset controller. The elongation of thin strip steel is dynamically adjusted by real-time correction of the roll speed and tension between the stands using the optimal adjustment amount.

[0014] To achieve the above objectives, an elongation control device for thin strip rolling is also provided. The device includes a memory, a processor, and an elongation control program for thin strip rolling stored in the memory and executable on the processor. The elongation control program for thin strip rolling is configured to implement the steps of any of the above-described elongation control methods for thin strip rolling.

[0015] This application has, but is not limited to, the following technical effects: In this application, real-time processing data during the thin strip steel rolling process is acquired, and temperature distribution data, real-time front and rear tension, and real-time rolling force are extracted from the real-time processing data. The rheological stress weight along the thickness direction of the thin strip steel is calculated using the temperature distribution data. Based on the rheological stress weight and real-time front and rear tension, the real-time rolling force is decoupled to obtain a dynamic offset coefficient. The dynamic offset coefficient characterizes the degree of deviation of the frictional force relative to the historical steady-state benchmark. The rheological stress weight, dynamic offset coefficient, and real-time processing data are input into a preset mapping model to estimate the neutral point position of the current deformation zone. Based on the neutral point position, the true elongation of the thin strip steel is derived. Based on the deviation between the preset target elongation and the true elongation, the elongation of the thin strip steel is dynamically adjusted. Combining the rheological stress weight, dynamic offset coefficient, and real-time processing data, the above multi-field coupled physical processes are reflected in real time. The neutral point position of the deformation zone is then determined using these data to dynamically adjust the elongation of the thin strip steel, effectively suppressing tension fluctuations between stands and strip shape instability, and improving the control accuracy of the elongation. Attached Figure Description

[0016] Figure 1 This is a schematic flowchart of the first embodiment of the elongation control method for thin strip steel rolling according to this application; Figure 2 This is a schematic diagram of the device structure of the hardware operating environment involved in the embodiments of this application. Detailed Implementation

[0017] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0018] This application provides a method for controlling the elongation rate in thin strip steel rolling. In the first embodiment of this method for controlling the elongation rate in thin strip steel rolling, referring to... Figure 1 The methods include: Step S10: Obtain real-time processing data during the thin strip steel rolling process, and extract temperature distribution data, real-time front and rear tension, and real-time rolling force from the real-time processing data.

[0019] As an example, the elongation control method for thin strip rolling can be applied to elongation control devices for thin strip rolling.

[0020] As an example, the finishing process of hot-rolled thin strip steel is characterized by high-speed rolling, where the strip can complete a long distance within a very short control cycle. However, the various sensors required for elongation calculation and deformation zone state analysis are located at different positions before and after the mill stand, with a fixed physical installation distance from the rolling deformation zone. Furthermore, the sampling frequencies and data transmission links of different sensors inherently differ. Simultaneously, the specific surface area of ​​thin strip steel is much larger than that of conventional medium-thick strip steel, resulting in an extremely rapid temperature drop during rolling, easily leading to a sustained and significant temperature difference between the strip's surface and core.

[0021] First, improvements will be made to the existing hot strip rolling mill: 1. To meet the requirement of acquiring full temperature field data, an array of temperature sensors arranged along the length of the strip can be added at the inlet and outlet sides of the rack to supplement the continuous temperature distribution data of the strip surface; 2. To address the need for spatiotemporal synchronization of multi-source data, a unified clock synchronization calibration is performed on the data acquisition links of all sensors to eliminate the inherent delay difference in data transmission.

[0022] Collect all data from the standard sensors on existing hot rolling production lines, specifically including: Inherent parameters, such as work roll radius, roll body length, roll elastic modulus, frame stiffness coefficient, etc. Incoming static parameters: strip steel grade composition, incoming material specifications, and initial temperature field data at the roughing mill exit; Real-time dynamic process parameters: Process data that can be collected in real time during the rolling process, such as tension before and after the stand, rolling force, real-time speed of the rolls, thickness of the strip at the inlet and outlet, continuous temperature measurement data of the strip surface, and real-time running speed of the strip between stands, can be used to obtain real-time processing data.

[0023] As an example, after data collection is completed, the acquired data is preprocessed: Based on the measured running speed of the strip in different sections between the sensor and the frame, the precise spatial coordinates of the strip are calculated by integrating the detection data of each sensor. Sensor data from different acquisition times and different installation positions are matched to the same spatial position of the strip, realizing synchronous processing with spatial position alignment as the main focus and timestamp correction as the secondary focus, eliminating data misalignment caused by physical installation distance and transmission delay. As an example, temperature distribution data can be the full temperature distribution data along the thickness direction of the strip within the deformation zone. The thickness direction of the strip refers to the physical depth direction perpendicular to the strip rolling surface, penetrating from the upper surface through the core to the lower surface. The acquisition method can be: based on continuous temperature measurement data of the strip surface, the thermal property parameters of the corresponding steel grade, and the temperature drop law during the rolling process, a one-dimensional unsteady heat conduction equation is used to reconstruct the continuous temperature field of the strip along the thickness direction from the surface to the core, thus obtaining the temperature distribution data.

[0024] Furthermore, a variable window adaptive filtering algorithm is adopted to dynamically adjust the size of the filtering window based on the real-time running speed of the strip: the filtering window is reduced under high-speed rolling conditions to retain the signal characteristics of transient conditions; the filtering window is expanded under low-speed steady-state rolling conditions to suppress random noise; and abnormal data is identified and removed simultaneously, taking into account both the smoothness and real-time performance of the data.

[0025] The formula for variable window adaptive filtering can be:

[0026] in, C represents the real-time running speed of the strip steel, and C is a preset spatial characteristic distance constant. and The upper and lower limits of the number of frames in the limit window are set manually. After preprocessing, standardized processing data that accurately matches the spatial position of the strip, has been denoised, and contains the full temperature field in the thickness direction is obtained for subsequent analysis and calculation.

[0027] Step S20: Based on temperature distribution data, the rheological stress weight along the thickness direction of the thin strip is calculated. Based on the rheological stress weight and real-time front and rear tension, the real-time rolling force is decoupled to obtain the dynamic offset coefficient. The dynamic offset coefficient is used to characterize the degree of offset between the friction force and the historical steady-state reference.

[0028] As an example, the rheological stress weight represents the relative characteristic value of the overall rheological stress in the deformation zone, characterizing its bias and degree relative to the average level. Based on the fundamental physical logic of hot rolling that "the higher the temperature, the lower the material's rheological stress," a reverse linear mapping is performed on the temperature distribution data to obtain the rheological stress weight. Through hierarchical linear mapping at the data level, the relative weight of the rheological stress at different locations in the thickness direction relative to the average level is quantified, providing dynamic stress boundary data for neutral point location estimation.

[0029] As an example, hot-rolled thin strip steel inherently possesses a large specific surface area and a rapid temperature drop rate during the rolling process. The continuous temperature difference along the thickness direction of the strip steel triggers significant fluctuations in rheological stress, which in turn causes substantial changes in rolling force. This effect is completely coupled with the influence of changes in interfacial friction state on rolling force. However, traditional methods do not decouple these two types of effects, directly attributing the overall fluctuation of rolling force entirely to changes in friction state, which can lead to significant calculation errors in the thin strip steel rolling scenario.

[0030] Meanwhile, the thin strip steel's flat deformation zone is much more sensitive to changes in the interfacial friction coefficient than conventional medium-thick strip steel. Even small fluctuations in the friction coefficient can cause significant changes in the rolling force. Therefore, it is necessary to first eliminate the interference components caused by the fluctuation of rheological stress in order to accurately capture the true change law of the interfacial friction state. Otherwise, it will directly lead to the complete failure of the dynamic tracking ability of the subsequent neutral point position estimation. Based on this, a dynamic offset coefficient is calculated to reflect the true change law of the interfacial friction state.

[0031] Step S20 further includes steps S21 to S23: Step S21: Based on the outlet and inlet of the deformation zone, construct the analysis location interval; As an example, traditional methods use a fixed time window to match data, which leads to a misalignment between the data used in the calculation and the actual spatial location of the strip in the deformation zone, resulting in a distorted neutral point estimation benchmark. To avoid data misalignment errors, it is necessary to match the physical extent of the deformation zone currently undergoing plastic deformation. Taking the exit of the rolling deformation zone as the reference zero point, and based on the real-time running speed of the strip, an analysis position interval that perfectly matches the physical range of the current deformation zone is defined: the strip position corresponding to the entrance of the deformation zone is the starting point of the interval, and the strip position corresponding to the exit of the deformation zone is the ending point of the interval. The length of the interval is completely consistent with the physical length of the contact arc of the deformation zone. Based on the analysis location range, from the preprocessed real-time processing data, using a sliding step size synchronized with the current hot rolling equipment basic automation control cycle, all valid data that have completed spatial synchronization matching within the range are extracted, and irrelevant data outside the range are removed. Each control cycle synchronously updates the valid data within the sliding window to ensure that the data within the window corresponds in real time to the physical position of the current deformation zone. The extracted data within the window is classified into features to construct a feature dataset specific to the deformation region, which is divided into three non-overlapping feature groups: The first category is the basic boundary feature group, which includes fixed or gradually changing parameters such as strip inlet and outlet thickness, work roll radius, and frame stiffness; The second category is dynamic process feature group, which includes real-time rapidly changing parameters such as rolling force, front and rear tension, roll speed, and strip running speed; The third category is the temperature field feature group, which includes layered temperature data in the thickness direction of the strip and continuous surface temperature data.

[0032] Step S22: Based on the temperature distribution data in the analysis location interval, determine the real-time temperature value of each layer along the thickness direction of the thin strip steel.

[0033] As an example, hot-rolled thin strip steel has a larger specific surface area than conventional medium-thick strip steel, and the radiation and convection temperature drop rates are higher during the rolling process. The surface and core of the strip steel are prone to forming a large and continuous temperature difference. The rheological stress increases nonlinearly with decreasing temperature. Traditional technology uses the average temperature of the cross section to calculate the single average rheological stress, which completely ignores the non-uniformity of stress in the thickness direction, resulting in a relatively large calculation deviation.

[0034] In addition, thin strip steel has the characteristic of synchronous plastic deformation of the entire thickness of the thin deformation zone: the length-to-thickness ratio of the deformation zone is less than or equal to 1, and the metal of the entire thickness of the strip steel participates in plastic deformation synchronously without difference. There is no deformation gradient of "large deformation on the surface and small deformation in the core" in medium and thick strip steel. Therefore, the difference in rheological stress between the surface and the core will directly affect the force balance of the entire deformation zone, rather than only affecting the surface. The traditional assumption of average stress cannot be adapted to this feature at all.

[0035] As an example, layered temperature data along the thickness direction of the strip is extracted from the temperature field feature set (i.e., the temperature distribution data in the analysis location interval). The thickness direction of the strip is divided into N independent data layers at equal intervals (in this embodiment, N is an odd number between 5 and 11, where the upper surface layer, lower surface layer, and core of the strip are all set as independent data layers). The real-time temperature value corresponding to each data layer is extracted and recorded as the real-time temperature value of the i-th layer. .

[0036] Step S23: Calculate the rheological stress weight along the thickness direction of the thin strip steel by using the deviation between each real-time temperature value and the reference temperature value.

[0037] As an example, the relative deviation between the temperature of each layer and the reference temperature is calculated by the deviation between each real-time temperature value and the reference temperature value. Then, by summing the relative deviation values ​​of each layer, the overall rheological stress weight along the thickness direction of the thin strip steel is obtained.

[0038] Step S23 includes: Extract the full-layer temperature of the strip steel from multiple control cycles within a historical time period.

[0039] The reference temperature value is determined based on the average temperature of each strip across all layers.

[0040] As an example, the historical time period can be two hours, three hours, etc., and the control period can be five minutes, etc., without any specific limitation.

[0041] As an example, the strip full-layer temperature represents the strip full-layer temperature at steady-state temperature. The reference temperature value is obtained by averaging the strip full-layer temperature of each control cycle.

[0042] For any given layer, calculate the relative deviation between the real-time temperature value and the reference temperature value; As an example, for the i-th stratum, the relative deviation value The calculation method can be:

[0043] in, It is a very small positive number to prevent the denominator from being 0. In this embodiment... , This represents the real-time temperature value of the i-th layer. This indicates the reference temperature value.

[0044] By linearly mapping the relative deviation values ​​of each layer, the rheological stress weights along the thickness direction of the thin strip are obtained.

[0045] The step of linearly mapping the relative deviation values ​​of each layer to obtain the rheological stress weight along the thickness direction of the thin strip steel includes: By performing an inverse linear mapping on the relative deviation values ​​of each layer, the relative stress weights of each layer are obtained.

[0046] As an example, based on the fundamental physical logic of hot rolling that "the higher the temperature, the lower the material's rheological stress," an inverse linear mapping is performed on the relative temperature deviation of each layer to obtain the relative stress weights of each layer:

[0047] In the formula, This represents the relative stress weight of the rheological stress in the i-th layer. The inverse linear mapping proportionality coefficient is a dimensionless fixed constant, pre-calibrated based on historical production data of the corresponding steel grade. It is only used to quantify the linear amplitude of the relative temperature deviation and the relative change of rheological stress. It is not adjusted in real time during single-coil rolling. In this embodiment, the value range of k is set to (0.01, 0.05) according to historical data.

[0048] It should be noted that the relative deviation value of the stratification when the temperature is higher than the reference value The rheological stress weight is lower than the baseline level by 1; the temperature is lower than the baseline value for the relative deviation of the stratification. The rheological stress weight is higher than the baseline level by 1.

[0049] The relative stress weights are normalized to generate a relative stress weight sequence.

[0050] As an example, the relative stress weights of all layers are normalized proportionally to obtain the relative stress weight sequence of the entire layer in the thickness direction. By weighting and summing the relative stress weights using the values ​​in the relative stress weight sequence, the rheological stress weights along the thickness direction of the thin strip steel are obtained.

[0051] As an example, rheological stress weighting The calculation method can be:

[0052] in, This represents the i-th sequence value in the relative stress weight sequence. Represents relative stress weights. Using 1 as the base value, if This indicates that the overall rheological stress in the deformation zone is higher than the average level across the entire thickness; if This indicates a level below average and is ultimately used to characterize the relative high or low bias and degree of overall rheological stress.

[0053] Among them, step S20, which involves decoupling the real-time rolling force based on the rheological stress weight and the real-time front and rear tensions to obtain the dynamic offset coefficient, also includes: Determine the steady-state reference rolling force for multiple control cycles within a historical time period, as well as the tension values ​​before and after the steady-state reference.

[0054] Based on the rheological stress weight, the steady-state reference rolling force is corrected for fluctuations to obtain the theoretical reference rolling force.

[0055] As an example, from the dynamic process feature group mentioned above, the real-time rolling force and tension data of the first 10 steady-state rolling control cycles are extracted, and their arithmetic mean is calculated to obtain the steady-state reference rolling force. , tension values ​​before and after steady-state reference; As an example, based on rheological stress weighting By correcting for rheological stress fluctuations in the steady-state reference rolling force, the theoretical reference rolling force under the current rheological stress level is obtained. .

[0056] By decoupling the real-time rolling force using the steady-state reference tension value and the real-time tension value, the effective value of the rolling force is obtained.

[0057] As an example, the real-time tension before and after the current control cycle and the real-time rolling force are extracted. Based on the difference between the current real-time tension before and after the current tension data and the steady-state reference tension value before and after the current tension, the rolling force change component caused by tension fluctuation is calculated and removed, resulting in the effective value of the rolling force that is only affected by the internal state of the deformation zone. Specifically, the effective value of rolling force The calculation method can be:

[0058] in, To measure the rolling force, and These are the real-time front and rear tensions, and Corresponding to the tension before and after the steady-state reference, and The front and rear tension interference compensation coefficients are calibrated based on historical production data. This represents the component of rolling force change caused by tension fluctuations.

[0059] The dynamic offset coefficient is calculated based on the relative fluctuation between the theoretical reference rolling force and the effective value of the rolling force.

[0060] The step of calculating the dynamic offset coefficient based on the relative fluctuation between the theoretical reference rolling force and the effective value of the rolling force includes: The rolling force fluctuation value is calculated based on the deviation ratio between the theoretical reference rolling force and the effective value of the rolling force.

[0061] As an example, the rolling force fluctuation value is calculated relative to the corrected theoretical reference rolling force. This fluctuation value is entirely caused by changes in the interfacial friction state, completely eliminating the interference of rheological stress and tension fluctuations. The calculation method can be:

[0062] In the formula, This represents the rolling force fluctuation caused solely by changes in frictional state. A positive value indicates that the measured rolling force is higher than the theoretical reference, while a negative value indicates that it is lower than the theoretical reference. A very small positive number is used to prevent the denominator from being 0, in this embodiment. .

[0063] By combining the friction coefficient corresponding to the preset steady-state benchmark with the rolling force fluctuation value, the dynamic offset coefficient is obtained.

[0064] As an example, based on the fundamental physical logic of hot rolling that "the greater the interfacial friction coefficient, the greater the contact friction in the deformation zone, and the higher the rolling force," the friction coefficient corresponding to the steady-state benchmark is used as the normalized benchmark value 1. The rolling force fluctuation value caused by pure friction is linearly mapped to the dynamic offset coefficient of the friction coefficient. :

[0065] In the formula, This indicates the rolling force fluctuation value. The friction-rolling force mapping ratio coefficient is pre-calibrated as a fixed constant based on the historical production data of the corresponding production line. According to historical experience, the value range of this ratio coefficient is set to (0.1, 0.5).

[0066] Finally, After performing amplitude limiting and normalization, with the amplitude limiting range set to [0.5, 1.5], and removing abnormal jump data, the dynamic normalized characterization value of the interface friction coefficient under the current control cycle is obtained: This indicates that the frictional state is stronger than the steady-state reference. This indicates that the frictional state is weaker than the steady-state reference, and characterizes the direction and extent of its offset relative to the steady-state reference.

[0067] Step S30: Input the rheological stress weight, dynamic offset coefficient and real-time processing data into the preset mapping model to estimate the neutral point position of the current deformation zone.

[0068] As an example, the neutral point location represents the physical location point in the deformation zone where the strip running speed is exactly equal to the roll linear speed. The preset mapping model represents a linear model with a fixed mapping relationship between the input data and the neutral point location, which is used to estimate the neutral point location using the data from the input model.

[0069] As an example, the inherent characteristics of ultra-high-speed rolling of hot-rolled thin strip steel determine that its basic automated control has extremely high real-time requirements, and the neutral point position estimation must be calculated and output within a single control cycle.

[0070] Traditional techniques employ implicit equations for rolling force balance to iteratively solve for the neutral point, which suffers from high computational complexity, numerous iteration steps, and long calculation time per step, making them completely unsuitable for the control cycle requirements of thin strip steel rolling. Furthermore, the nonlinear characteristics of the flat and thin deformation zone of thin strip steel are more prominent, and iterative solutions are prone to non-convergence problems, which can cause interruptions in control command output in high-speed rolling scenarios.

[0071] In addition, the transient fluctuation amplitude of rolling parameters during the rolling process of thin strip steel is significantly higher than that of conventional medium and heavy strip steel. Traditional solution methods are highly sensitive to fluctuations in input parameters. Even small parameter fluctuations can cause large jumps in the solution results, making it impossible to output a stable neutral point estimate, which in turn leads to continuous oscillations in subsequent control commands.

[0072] Based on this, a multi-feature data mapping matrix for the neutral point position is constructed. This matrix includes the rheological stress weight, dynamic offset coefficient, and the relative ratio of front and rear tensions and the ratio of strip inlet and outlet thicknesses in the real-time processing data. Then, the neutral point position of the current deformation zone is obtained by linearly combining the multi-feature data mapping matrix through a preset mapping model.

[0073] Step S30 includes: Obtain historical steady-state rolling data during the thin strip steel rolling process.

[0074] Historical steady-state rolling data were processed using linear regression fitting to construct a fixed mapping relationship between input features and neutral point positions in a pre-defined mapping model.

[0075] As an example, based on historical steady-state rolling data from the production line, a fixed mapping relationship between input features and the neutral point position is constructed in a pre-defined mapping model through linear regression fitting, providing a benchmark for online estimation.

[0076] Based on a fixed mapping relationship, the baseline weight coefficients corresponding to the rheological stress weight, dynamic offset coefficient, and real-time processing data are determined respectively.

[0077] As an example, the rheological stress weight, dynamic offset coefficient, and real-time processing data are input into a preset mapping model. The measured neutral point position under historical steady-state rolling is used as the label, and the benchmark weight coefficient corresponding to each input feature is obtained through linear regression fitting.

[0078] It should be noted that when the friction coefficient increases, the neutral point moves towards the inlet of the deformation zone, and the corresponding weight is positive; when the rheological stress increases, the neutral point moves towards the outlet of the deformation zone, and the corresponding weight is negative; after the benchmark weight coefficient is pre-calibrated, it is not modified in real time during the online calculation, but is only used as a fixed mapping benchmark.

[0079] The neutral point position of the current deformation zone is obtained by summing the rheological stress weight, dynamic offset coefficient, and real-time processing data using the benchmark weight coefficient.

[0080] As an example, the real-time processing data includes two types of data: the relative ratio of front and rear tensions and the ratio of strip thickness at the inlet and outlet. A total of four types of data are input, specifically: By linearly combining the four types of input feature values ​​of the current control cycle with the corresponding benchmark weight coefficients, the neutral point position of the current deformation zone can be quickly obtained: The four types of input feature values ​​for the current control cycle are multiplied by their corresponding baseline weighting coefficients and then summed.

[0081] in, This is the initial estimate of the neutral point location. The baseline weight coefficients are the values ​​corresponding to the m-th type of feature. The measured or calculated value of the m-th feature after standardization in the current period. The reference bias constant is a pre-calibrated reference bias constant based on historical steady-state data. For example, after a cold start or roll change, the first N historical steady-state rolled steel coil data can be extracted, and the initial reference bias constant can be obtained offline by using conventional regression algorithms such as the least squares method.

[0082] It should be noted that the neutral point position uses a dimensionless relative value throughout, with the deformation zone exit as point 0 and the deformation zone entrance as point 1. The value ranges from 0 to 1; a larger value indicates that the neutral point is closer to the entrance, and a smaller value indicates that it is closer to the exit. Finally, considering the physical characteristics of the thin strip steel's flat deformation zone, the initial estimate is subjected to boundary limiting processing to arrive at the final estimate of the relative position of the neutral point. With the amplitude limited to (0.5, 1), abnormal estimates that exceed the physically reasonable range are eliminated to obtain the final estimate of the neutral point position under the current control cycle.

[0083] Step S40: Based on the neutral point position, derive the true elongation of the thin strip steel, and dynamically adjust the elongation of the thin strip steel based on the deviation between the preset target elongation and the true elongation.

[0084] As an example, the true elongation of thin strip steel is derived by combining the neutral point position with the roll linear speed. The reconstructed true elongation is then compared with the preset target elongation. Based on the deviation between the two, the adjustment parameters in the thin strip steel rolling process are determined, thereby dynamically adjusting the elongation of the thin strip steel.

[0085] The steps for deriving the true elongation of thin strip steel based on the neutral point location include: Obtain the linear speed of the rolls during the rolling process of thin strip steel.

[0086] By combining the neutral point position with the roll linear velocity, the true inlet velocity and the true outlet velocity in the deformation zone can be obtained.

[0087] As an example, before calculation, a target range for the relative position of the optimal neutral point in steady-state rolling is set based on historical rolling data; the current neutral point position is then calculated. The positional deviation of the target interval, and according to The relationship between the drift direction and amplitude and the upper and lower limits of the target interval determines the drift direction and amplitude. When the value exceeds the upper limit of the target interval, the neutral point is too close to the entrance. When the value is less than the lower limit of the target interval, the neutral point is too close to the exit. Based on the direction and magnitude of the deviation, the pre-calibrated compensation coefficient is matched to output the feedforward correction: when the neutral point is near the inlet, the inter-stand tension is slightly reduced and the roll speed is slightly increased; when the neutral point is near the outlet, the inter-stand tension is slightly increased and the roll speed is slightly reduced; the feedforward correction includes speed correction and tension correction, and the specific correction methods can be: Speed ​​correction amount = *( Target interval center value); tension correction amount = * ( Target interval center value), defined and The gain coefficients for the manually set rotational speed and tension are positively adjusted, and the feedforward correction is synchronously superimposed on the control setpoint to offset the disturbance of the elongation rate caused by the neutral point drift in advance and suppress the sudden change in elongation rate.

[0088] As an example, the neutral point position is related to the roll linear speed. Combined, calculate the actual inlet velocity of the strip. Actual export speed This formula is a method for calculating the velocity representation in the control algorithm, not a strict physical velocity.

[0089] The true elongation of the thin strip steel is derived based on the difference between the actual inlet velocity and the actual outlet velocity.

[0090] As an example, true stretch rate The calculation method can be:

[0091] In the formula, Indicates the actual export speed. Indicates the actual entry speed, using It replaces the traditional approximation based on speed difference as the system feedback quantity.

[0092] The step of dynamically adjusting the elongation of the thin strip steel based on the deviation between the preset target elongation and the actual elongation includes: Based on the deviation between the preset target elongation rate and the actual elongation rate, the elongation rate control error is determined; As an example, the elongation control error e is obtained by subtracting the reconstructed actual elongation from the preset target elongation, and the error change rate ec between the current period and the previous period is calculated.

[0093] The optimal adjustment amount is obtained by dynamically adapting the elongation control error, rheological stress weight, and dynamic offset coefficient through a preset controller.

[0094] As an example, the preset controller can be a fuzzy PID controller, which inputs the elongation control error e, the error change rate ec, along with the rheological stress weight and the dynamic offset coefficient.

[0095] The rheological stress weight characterizes the material's deformation resistance level, while the dynamic offset coefficient characterizes the interfacial friction resistance level. The fuzzy rule base uses these as boundary conditions in inference. For example, when the rheological stress weight is greater than 1, the material hardens, and the rule automatically triggers a positive compensation value to increase the proportional gain Kp. When the dynamic offset coefficient fluctuates drastically, the integral gain Ki is reduced to prevent system oscillation, resulting in the real-time compensation amount d for the three PID parameters. Kp d Ki d Kd .

[0096] The preset baseline PID parameters of the control system are superimposed on d respectively. Kp d Ki d Kd This generates the actual Kp, Ki, and Kd for the current control cycle. The error e is then substituted into the PID positional or incremental algorithm formula to calculate the overall optimal adjustment.

[0097] The elongation of thin strip steel is dynamically adjusted by real-time correction of the roll speed and tension between the stands using the optimal adjustment amount.

[0098] As an example, the decoupling ratio of the production line equipment is preset (e.g., weighting of rotational speed). and tension allocation weight The optimal adjustment amount is decomposed into the frame speed correction value and the inter-frame tension correction value, which are sent as closed-loop control quantities to the speed controller and tension controller of the underlying PLC. They are superimposed on the feedforward control quantity and executed together, thereby dynamically adjusting the elongation of the thin strip steel.

[0099] This application provides a method for controlling the elongation of thin strip steel during rolling. In this method, real-time processing data during the thin strip steel rolling process is acquired, and temperature distribution data, real-time front and rear tension, and real-time rolling force are extracted from the real-time processing data. The rheological stress weight along the thickness direction of the thin strip steel is calculated using the temperature distribution data. Based on the rheological stress weight and real-time front and rear tension, the real-time rolling force is decoupled to obtain a dynamic offset coefficient. The dynamic offset coefficient characterizes the degree of deviation of the frictional force relative to the historical steady-state benchmark. The rheological stress weight, dynamic offset coefficient, and real-time processing data are input into a preset mapping model to estimate the neutral point position of the current deformation zone. Based on the neutral point position, the true elongation of the thin strip steel is derived. Based on the deviation between the preset target elongation and the true elongation, the elongation of the thin strip steel is dynamically adjusted. Combining the rheological stress weight, dynamic offset coefficient, and real-time processing data, the method reflects the aforementioned multi-field coupled physical process in real time. The neutral point position of the deformation zone is then determined using these data to dynamically adjust the elongation of the thin strip steel, effectively suppressing tension fluctuations between stands and plate instability, and improving the control accuracy of the elongation.

[0100] This application embodiment also provides an elongation control device for thin strip steel rolling, the device including: a memory, a processor, and an elongation control program for thin strip steel rolling stored in the memory and executable on the processor, the elongation control program for thin strip steel rolling being configured to implement the steps of any of the above-described elongation control methods for thin strip steel rolling.

[0101] Reference Figure 2 , Figure 2 This is a schematic diagram of the device structure of the hardware operating environment involved in the embodiments of this application.

[0102] like Figure 2 As shown, the elongation control device for thin strip steel rolling may include: a processor 1001, a memory 1003, and a communication bus 1002. The communication bus 1002 is used to realize the connection and communication between the processor 1001 and the memory 1003.

[0103] Optionally, the elongation control device for thin strip steel rolling may also include a user interface, a network interface, a camera, RF (Radio Frequency) circuitry, sensors, a WiFi module, etc. The user interface may include a display screen and an input submodule such as a keyboard; optional user interfaces may also include standard wired or wireless interfaces. The network interface may include standard wired or wireless interfaces (such as a Wi-Fi interface).

[0104] Those skilled in the art will understand that Figure 2 The elongation control device structure shown in the figure does not constitute a limitation on the elongation control device for thin strip rolling. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0105] like Figure 2 As shown, the memory 1003, serving as a storage medium, may include an operating system, a network communication module, and an elongation control program for thin strip steel rolling. The operating system is a program that manages and controls the hardware and software resources of the elongation control device for thin strip steel rolling, supporting the operation of the elongation control program and other software and / or programs. The network communication module is used to enable communication between the various components within the memory 1003, as well as communication with other hardware and software in the elongation control system for thin strip steel rolling.

[0106] exist Figure 2In the elongation control device for thin strip rolling shown, the processor 1001 is used to execute the elongation control program for thin strip rolling stored in the memory 1003 to implement the steps of the elongation control method for thin strip rolling described in any of the above claims.

[0107] The specific implementation of the elongation control device for thin strip steel rolling in this application is basically the same as the various embodiments of the elongation control method for thin strip steel rolling described above, and will not be repeated here.

[0108] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0109] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0110] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0111] The above are merely preferred embodiments of this application and do not limit the scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the scope of protection of this application.

[0112] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0113] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for controlling elongation in thin strip steel rolling, characterized in that, The method includes: Real-time processing data during the thin strip steel rolling process is obtained, and temperature distribution data, real-time front and rear tension, and real-time rolling force are extracted from the real-time processing data; Based on the temperature distribution data, the rheological stress weight along the thickness direction of the thin strip is calculated. Based on the rheological stress weight and the real-time front and rear tension, the real-time rolling force is decoupled to obtain the dynamic offset coefficient. The dynamic offset coefficient is used to characterize the degree of offset of the frictional force relative to the historical steady-state reference. The rheological stress weight, the dynamic offset coefficient, and the real-time processing data are input into a preset mapping model to estimate the neutral point position of the current deformation zone. Based on the neutral point position, the true elongation of the thin strip is derived, and the elongation of the thin strip is dynamically adjusted based on the deviation between the preset target elongation and the true elongation.

2. The method for controlling elongation in thin strip steel rolling as described in claim 1, characterized in that, The calculation of the flow stress weight along the thickness direction of the thin strip steel based on the temperature distribution data includes: Based on the outlet and inlet of the deformation zone, an analysis location interval is constructed; Based on the temperature distribution data in the analysis location interval, the real-time temperature value of each layer along the thickness direction of the thin strip steel is determined; The rheological stress weight along the thickness direction of the thin strip steel is calculated by the deviation between each real-time temperature value and the reference temperature value.

3. The elongation control method for thin strip steel rolling as described in claim 2, characterized in that, The calculation of the flow stress weight along the thickness direction of the thin strip steel by means of the deviation between each of the real-time temperature values ​​and the reference temperature value includes: Extract the full-layer temperature of the strip steel from multiple control cycles within a historical time period; A reference temperature value is determined based on the average temperature of the entire strip. For any given layer, calculate the relative deviation between the real-time temperature value and the reference temperature value; By linearly mapping the relative deviation values ​​of each layer, the rheological stress weights along the thickness direction of the thin strip are obtained.

4. The elongation control method for thin strip steel rolling as described in claim 3, characterized in that, The linear mapping of the relative deviation values ​​of each layer to obtain the rheological stress weight along the thickness direction of the thin strip steel includes: By performing an inverse linear mapping on the relative deviation values ​​of each layer, the relative stress weights of each layer are obtained. The relative stress weights are normalized to generate a relative stress weight sequence. By using the sequence values ​​in the relative stress weight sequence, the relative stress weight is weighted and summed to obtain the rheological stress weight along the thickness direction of the thin strip steel.

5. The method for controlling elongation in thin strip rolling as described in claim 1, characterized in that, The decoupling process of the real-time rolling force based on the rheological stress weight and the real-time front and rear tensions to obtain the dynamic offset coefficient includes: Determine the steady-state reference rolling force for multiple control cycles within a historical time period, as well as the tension values ​​before and after the steady-state reference. Based on the rheological stress weight, the steady-state reference rolling force is corrected for fluctuations to obtain the theoretical reference rolling force; The real-time rolling force is decoupled using the steady-state reference tension value and the real-time tension value to obtain the effective value of the rolling force. The dynamic offset coefficient is calculated based on the relative fluctuation between the theoretical benchmark rolling force and the effective value of the rolling force.

6. The elongation control method for thin strip steel rolling as described in claim 5, characterized in that, The dynamic offset coefficient is calculated based on the relative fluctuation between the theoretical benchmark rolling force and the effective value of the rolling force, including: The rolling force fluctuation value is calculated based on the deviation ratio between the theoretical benchmark rolling force and the effective value of the rolling force. The dynamic offset coefficient is obtained by combining the friction coefficient corresponding to the preset steady-state benchmark with the rolling force fluctuation value.

7. The method for controlling elongation in thin strip rolling as described in claim 1, characterized in that, The step of inputting the rheological stress weight, the dynamic offset coefficient, and the real-time processing data into a preset mapping model to estimate the neutral point position of the current deformation zone includes: Obtain historical steady-state rolling data during the thin strip steel rolling process; The historical steady-state rolling data is processed by linear regression fitting to construct a fixed mapping relationship between input features and neutral point positions in a preset mapping model; Based on the fixed mapping relationship, determine the benchmark weight coefficients corresponding to the rheological stress weight, the dynamic offset coefficient, and the real-time processing data, respectively. The neutral point position of the current deformation zone is obtained by summing the rheological stress weight, the dynamic offset coefficient, and the real-time processing data using the reference weight coefficient.

8. The method for controlling elongation in thin strip rolling as described in claim 1, characterized in that, The derivation of the true elongation of the thin strip steel based on the neutral point position includes: Obtain the linear speed of the rolls during the rolling process of thin strip steel; By combining the neutral point position with the roll linear velocity, the true inlet velocity and the true outlet velocity in the deformation zone can be obtained. The true elongation of the thin strip steel is derived based on the difference between the actual inlet velocity and the actual outlet velocity.

9. The method for controlling elongation in thin strip rolling as described in claim 1, characterized in that, The method of dynamically adjusting the elongation of the thin strip steel based on the deviation between the preset target elongation and the actual elongation includes: Based on the deviation between the preset target elongation rate and the actual elongation rate, the elongation rate control error is determined; The optimal adjustment amount is obtained by dynamically adapting the elongation control error, the rheological stress weight, and the dynamic offset coefficient through a preset controller. The elongation of the thin strip steel is dynamically adjusted by real-time correction of the roll speed and tension between the stands using the optimal adjustment amount.

10. An elongation control device for thin strip steel rolling, characterized in that, The apparatus includes: a memory, a processor, and an elongation control program for thin strip rolling stored in the memory and executable on the processor, the elongation control program for thin strip rolling being configured to implement the steps of the elongation control method for thin strip rolling as claimed in any one of claims 1 to 9.

Images