Automatic control system of hot-rolled strip steel temper mill

Abstract

The invention discloses an automatic control system of a hot-rolled strip steel temper mill, relates to the technical field of steel rolling automatic control, and is used for solving the problems of inaccurate material structure control and unstable pressing force in hot-rolled strip steel production. Through multi-dimensional sensing data fusion, accurate regulation and control of tissue transformation are realized based on a phase change kinetic model. And real-time closed-loop adjustment is adopted for pressing force control, and it is guaranteed that rolling force is stable and evenly distributed. The plate shape adjusting module dynamically adjusts the roll bending pressure through the multi-point sensing and hydraulic servo technology, and efficient plate shape correction is achieved. And the elongation module calculates the deformation rate in real time, automatically adjusts rolling parameters and ensures stable strip steel quality. And the feedback optimization module integrates data of each link, realizes multi-module cooperative adjustment by applying an intelligent control strategy, and improves the self-adaptive capability and the operation reliability of the system. According to the whole scheme, precise control and dynamic optimization of the rolling process are achieved, and efficient and stable hot rolling strip steel production is supported.

Description

Technical Field

[0001] The present invention relates to the technical field of steel rolling automation control, and more particularly to an automation control system for a hot-rolled strip steel skin-pass unit. Background Art

[0002] As an important industrial basic material, hot-rolled steel strip is widely used in the automotive, home appliance, construction and other fields. Its quality is directly related to the performance and safety of downstream products. During the hot rolling process, the microstructural state of the steel strip has a significant impact on the mechanical properties, especially the cooling control at the end of the hot rolling stage plays a key role in the evolution of the microstructure. However, existing cooling control technologies mainly rely on empirical parameters and simple temperature monitoring, lack real-time perception of the austenite-ferrite phase transformation dynamics, and are unable to accurately control the microstructure, resulting in an increase in the yield point stress difference, which easily causes transverse fold defects and affects the product qualification rate.

[0003] Rolling force is a key process parameter affecting strip thickness and shape. Traditional electro-hydraulic closed-loop control systems have limited response speed and control accuracy, making them incapable of meeting the demands of high-speed, dynamic rolling. This leads to frequent thickness fluctuations and shape anomalies. Furthermore, shape adjustment often relies on manual experience to adjust the bending roll force, lacking automatic recognition and dynamic compensation capabilities, making it difficult to achieve rapid and precise shape control.

[0004] In addition, strip elongation control is usually achieved through simple speed and tension adjustment, lacking closed-loop tracking and abnormal protection mechanisms, which can easily cause uneven elongation and affect flatness and mechanical properties.

[0005] In view of the above problems, the present invention proposes a solution. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an automatic control system for a hot-rolled strip skin-pass mill to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] In a preferred embodiment, it includes: a microstructure regulation and cooling control module, a pressing force electro-hydraulic closed-loop control module, a plate shape adaptive adjustment and bending roll control module, an elongation automatic tracking control module, and a feedback optimization and collaborative integration module, and the signal connections between the modules;

[0009] The microstructure regulation and cooling control module is mainly used to extract the chemical composition and process information of steel grades, build a multi-dimensional sensor data channel, predict the microstructure evolution using the phase change kinetics model, and control the spray cooling flow and time in different zones.

[0010] The electro-hydraulic closed-loop control module for the pressing force is mainly used to calculate the target rolling force, adjust the electro-hydraulic proportional valve in a closed loop to control the action of the hydraulic cylinder, and dynamically compensate for the pressing error;

[0011] The plate shape adaptive adjustment and bending roll control module constructs the plate shape profile time series data, identifies the edge wave, middle wave and straight state, and dynamically calculates the bending roll force distribution strategy;

[0012] The elongation automatic tracking control module calculates the deviation between the real-time elongation and the target, automatically adjusts the pressing force and the main motor linear speed, adjusts the step length, and triggers speed reduction and tension release when the limit is exceeded;

[0013] The feedback optimization and collaborative integration module builds a master-slave control architecture, inputs data to drive the model to determine operating deviations, and adjusts the control strategy.

[0014] In a preferred embodiment, in the microstructure regulation and cooling control module, the steel coil furnace information is called to identify the transverse bending risk level of the strip; then the cooling path required for the quasi-acicular ferrite structure is calculated, the cooling section is divided into multiple spray sections, and a temperature measurement zone is set.

[0015] In a preferred embodiment, in the microstructure regulation and cooling control module, the roll load and coiling outlet temperature data of the strip spray zone are collected to form a real-time multivariable input vector and construct the Avrami dynamic model.

[0016] In a preferred embodiment, in the microstructure regulation and cooling control module, the chemical composition of the steel strip is monitored as input to predict the phase fraction; a graph structure is constructed, the positive and negative sample comparison losses are designed, the normal path vs. phase transformation deviation is identified, and the phase transformation correction amount is output:

[0017] The target phase fraction, i.e., the upper limit of acicular ferrite, is set, the phase fraction deviation is calculated, and the flow rate and opening time of each nozzle section are adjusted in real time to bring the tissue drift back to the target path.

[0018] In a preferred embodiment, in the microstructure regulation and cooling control module, an extended form is constructed; the phase change equation coefficients are dynamically adjusted to adapt to batch differences; then, a tissue abnormality indication function is defined to define a tissue drift index.

[0019] In a preferred embodiment, the actual rolling force applied to the workpiece by both sides during the rolling process is collected in the electro-hydraulic closed-loop control module and input into the PLC or industrial control master station; the target rolling force value is then calculated and the PID controller parameters are optimized;

[0020] The PID output instruction is physically converted into a pressing behavior, and the dynamic response is completed through electro-hydraulic coordination; the input controller forms a new error, continues to adjust the pressing amount, and sets the dead zone tolerance range.

[0021] In a preferred embodiment, in the plate shape adaptive adjustment and bending roller control module, the height fluctuations of the strip at multiple horizontal points are detected to construct a time-series plate shape profile; the height change ratio of the edge area and the center area on both sides of the plate shape curve is calculated, and the bending roller hydraulic servo cylinder at the corresponding position is adjusted.

[0022] In a preferred embodiment, the elongation automatic tracking control module records the inlet section length Lin and the outlet section length Lout, periodically calculates the actual elongation, sets the target elongation, and calculates the elongation error to adjust the pressing force.

[0023] In a preferred embodiment, real-time sensor data, control output status and fault alarm codes from other modules are read; and a master-slave logic control tree is established.

[0024] The technical effects and advantages of the automatic control system for a hot-rolled strip skin-pass mill of the present invention are as follows:

[0025] The present invention uses multi-dimensional sensing and intelligent control technology to achieve fine adjustment of key parameters in the flattening process of hot-rolled strip, thereby improving the control accuracy and response speed. The microstructure control module realizes real-time monitoring and adjustment of complex phase change processes, optimizes the material structure, and improves the uniformity and mechanical properties of the product. The closed-loop control of the pressing force enhances the dynamic adaptability of the hydraulic system, ensuring the stability of the strip thickness and the uniformity of the plate shape. The plate shape adjustment system introduces automatic identification and real-time compensation mechanisms, which effectively reduces manual intervention and improves the efficiency and accuracy of plate shape control. The elongation tracking control combined with multi-sensor data enhances process stability and avoids local deformation and defects. The overall system realizes the coordinated work of various control links through feedback optimization strategies, improves the stability and automation level of the production process, significantly reduces equipment failure rate and energy consumption, and enhances the intelligent management capabilities of the production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of the implementation of an automatic control system for a hot-rolled strip skin-pass mill according to the present invention.

[0027] Figure 2 This is a timing diagram of an automatic control system for a hot-rolled strip skin-pass mill according to the present invention. DETAILED DESCRIPTION

[0028] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0029] Example

[0030] The present invention discloses an automatic control system for a hot-rolled strip skin-pass mill. Figure 1 As shown, it includes: microstructure regulation and cooling control module, pressing force electro-hydraulic closed-loop control module, plate shape adaptive adjustment and bending roll control module, elongation automatic tracking control module and feedback optimization and collaborative integration module, and signal connections between each module.

[0031] like Figure 2 As shown in the figure, the microstructure regulation and cooling control module is mainly used to effectively reduce the stress difference between the upper and lower yield points of the material by precisely controlling the cooling path and the microstructure evolution process at the end stage of hot rolling, inhibit the occurrence of transverse bending defects, and ensure that the strip has the target structure state before entering the leveler.

[0032] Specifically, in the microstructure control and cooling control module, coil heat information is retrieved from the MES or automation platform to extract chemical composition data, including C, Mn, Nb, Ti, Al, Si, P, and S. Next, a steel grade identification algorithm is used to match the steel grade against a database of high-yield steel grades preset by the mill. This database contains risk levels for cross-bends predicted by composition ratio and lattice parameters. For example, based on the logical rules of C content > 0.08% and Mn+Si > 1.5%, the risk levels are divided into low, medium, and high levels. This determines whether the hot-rolled strip is a high-risk yield steel grade. If this is determined to be a high-risk yield steel grade, the high-risk steel grade will automatically trigger the activation flag of the microstructure control link. This enables early identification of high-yield-sensitive steel grades and ensures that enhanced control strategies are adopted for downstream cooling parameters.

[0033] Furthermore, the online hot rolling process model is called, the steel grade composition is input, and the empirical CCT curve library preset in the industrial PC is embedded, and the corresponding curve is selected. Then, the phase transformation dynamics formula JMAK model is used to calculate the cooling path required for the quasi-acicular ferrite structure, which serves as a reference for temperature control in the cooling section. The details are as follows:

[0034] X(t)=1-exp(-kt n );

[0035] Where X represents the phase transformation volume fraction, k and n represent the phase transformation parameters specific to the steel grade;

[0036] Then determine the following target temperature control parameter range:

[0037] Initial cooling temperature: finishing rolling outlet temperature;

[0038] Termination cooling temperature: critical temperature for the formation of acicular ferrite;

[0039] Cooling time: Keep within 5 to 10 seconds to ensure complete tissue transformation;

[0040] Cooling rate: needs to reach 30-70℃ / s to avoid the formation of pearlite or granular bainite.

[0041] Furthermore, based on the cooling path set in the above steps, the control structure is sent from the PLC controller to the cooling water zone actuator; the cooling section is divided into multiple spray sections, such as: high-pressure section, low-pressure section, and mixed section, each section is equipped with a flow control valve and solenoid valve;

[0042] The parameters of each segment are set as follows:

[0043] Flow control: Through the combined regulation of solenoid valve + VFD pump, the water volume accuracy per unit time can reach ±5L / min;

[0044] Spraying time control: set the time window Δti;

[0045] At the same time, set the temperature feedback channel, set up an infrared thermometer at the inlet section and an infrared temperature measurement at the outlet section;

[0046] Furthermore, the cooling parameters are controlled in a real-time closed-loop manner through control logic to ensure that the temperature curve fits the phase change trajectory. The control logic is as follows:

[0047]

[0048] Furthermore, the coiling temperature is precisely controlled at the cooling end. At the same time, a temperature measuring area is set before coiling to ensure that the coiling temperature meets the target value. If the temperature in the temperature measuring area is too high, low-pressure spray cooling is performed at the end. If the temperature in the temperature measuring area is too low, the residence time is increased by automatic deceleration, or some nozzles are closed to increase the temperature, so as to achieve the target organization and reduce the yield point stress difference to below the critical value, thereby fundamentally avoiding the occurrence of transverse bending defects and ensuring that the material has a low yield difference organizational structure before entering the leveling machine.

[0049] It should be noted that when the microstructure control and cooling module executes controlled cooling, the strip has just left the finishing mill and entered the high-low pressure multi-stage spray zone. At this time, the temperature drop rate of the steel plate is extremely sensitive to the formation of acicular ferrite. If the raw material composition fluctuates or the nozzle is partially clogged, the real-time microstructure will deviate from the predicted path. Existing online measurement relies primarily on roll load and coil outlet temperature, which cannot directly capture the austenite-ferrite transformation dynamics. The placement of ultrasonic or magnetostrictive sensors is limited by high temperatures and cooling water shock, resulting in microstructure composition and grain size only being able to be measured offline. Furthermore, most mills only use a dual closed-loop "spray flow-time" control loop, combining CCT curves with empirical temperature drop models. This lacks in-situ observation and feedback of real-time lattice phase transformations, making it impossible to identify sudden nozzle failures or batch differences in steel grades. High-temperature ultrasonic-electromagnetic hybrid sensor arrays or fiber Bragg gratings based on spectral reflection are needed to achieve subsecond microstructure parameter inversion and form a micro-circuit feedback loop with the cooling PLC.

[0050] If structural drift is ignored, the yield point stress difference increases, leading to transverse folding defects. This can lead to excessive flattening and pressurization, making wave repair difficult, or even the entire roll being scrapped, resulting in both economic losses and disrupted production line cycles. Therefore, in this embodiment, a four-dimensional time series dataset of "temperature-load-acoustic-magnetic" is constructed, and a self-supervised graph convolutional network (GCN) is used to learn phase transformation dynamics, predicting phase fraction and grain size in real time. An abnormal fraction threshold is then used to trigger local flow compensation in the spray zone.

[0051] The specific steps are as follows:

[0052] Step S1: An EMAT-FBG composite sensor array is arranged above the strip spray zone to establish a "temperature-load-acoustic-magnetic" four-dimensional data channel. An EMAT (electromagnetic acoustic transducer) that can generate high-frequency shear waves and sense differences in grain size and phase density, and an FBG (fiber Bragg grating) that can sense temperature, stress, and strain modulated reflection wavelengths and characterize residual stress and microstructural disturbances are used to synchronously collect process parameters such as the roll load Fr and the coiling outlet temperature Tu in the strip spray zone, and form a real-time multivariable input vector to achieve comprehensive perception of the thermal-stress field and structural changes.

[0053] Furthermore, the corresponding CCT curve and phase transformation path are selected based on the steel composition, and the substructure transformation target is defined as acicular ferrite. Then, the Avrami kinetic model is constructed:

[0054] X γ→α (t)=1-exp[-K(q,C)t n ];

[0055] Where q represents the cooling rate, K(q, C) represents the phase change rate factor obtained by experimental regression, and t represents time, which represents the tissue development process;

[0056] Sensors then monitor the chemical composition in real time and feed it into the model to predict the phase fractions;

[0057] Furthermore, we construct a graph structure G = (V, E), where each node is a location channel, and edges at different time points form spatiotemporal coupling. We then design a positive and negative sample contrast loss Lcontrast to allow the network to self-supervise the identification of "normal path vs phase transition deviation", according to the formula:

[0058]

[0059] Where, zi = GCNEnc (xi = phase fraction);

[0060] Output phase change correction amount:

[0061] δX(s,t)=σ(Wz(s,t)+b);

[0062] Correct the status of the organization in real time:

[0063]

[0064] Realize double closed-loop residual coupling of data-driven and physical prediction.

[0065] After that, the target phase fraction, i.e., the upper limit of acicular ferrite, is set, and the phase fraction deviation is calculated. At the same time, a closed-loop control formula is executed on the PLC controller every T1 period to adjust the flow rate and opening time of each nozzle segment in real time to bring the tissue drift back to the target path. The specific formula is as follows:

[0066]

[0067] Where Qspray represents the total flow rate of the spray area; kp, kd, kf represent the rolling coefficients;

[0068] Step S2: Set the Avrami parameter K,n to the system extension state and construct the system state extension form;

[0069] The UKF is then used to fuse the [vs,λB] measurements and dynamically adjust the phase change equation coefficients to adapt to batch differences.

[0070] Furthermore, the tissue abnormality indication function first defines the tissue drift index Ψ:

[0071]

[0072] If N consecutive frames satisfy Ψ>Ψcrit, it is determined that: local nozzle blockage, steel batch mutation, or spraying failure in the controlled cooling zone; and the emergency feedback logic is triggered: the "strong cooling bypass" of the control spray system is started, or the flow is redistributed to achieve local cooling enhancement.

[0073] Furthermore, SS-GCN was continuously trained in the cloud using a window of M=10^4 time series data, and a <2MB model GCNl ite was obtained by distillation, which enabled the small model to be deployed in the PLC-ARM controller to meet the inference response time of high-speed strip steel.

[0074] The electro-hydraulic closed-loop control module of the pressing force controls the rolling pressure to achieve uniform pressing and maintain stable thickness.

[0075] Specifically, in the electro-hydraulic closed-loop control module of the pressing force, a strain gauge pressure sensor is first installed between the upper and lower support rollers and the frame to collect the actual rolling force applied to the rolled piece on both sides (left and right) during the rolling process in real time. At the same time, the sampling period is controlled within 10 to 50 ms to meet the high-speed rolling response requirements.

[0076] Afterwards, the analog signal output by the sensor is converted into a digital signal by the A / D conversion module and enters the PLC or industrial control master station. Before entering the control logic, the signal needs to undergo a first-order low-pass filter or median filter to eliminate high-frequency interference and spike abnormal values ​​caused by vibration.

[0077] Furthermore, the process model calculates the target rolling force value based on the steel type, thickness, width and temperature, and constructs the following closed-loop control logic:

[0078] e(t)=F target -F actual ;

[0079]

[0080] Where, e(t) represents the rolling force error, and u(t) represents the control output signal (i.e., the press-down correction amount);

[0081] Afterwards, the PID controller parameters are optimized online using the Ziegler-Nichols critical proportion method or offline using the model reference auto-tuning (MRAC) algorithm to ensure fast system response, no overshoot, and no steady-state error.

[0082] Furthermore, the PID output command is physically converted into a press-down action, achieving a dynamic "error-action" response through electro-hydraulic coordination. Specifically, the control signal u(t) is output by the controller and input into the electro-hydraulic proportional valve. The proportional valve adjusts the valve core opening via the electrical signal, controlling the flow of hydraulic oil and driving the hydraulic cylinders on both sides to operate synchronously. Simultaneously, the action of the hydraulic cylinders moves the upper work roll up and down, dynamically adjusting the roll gap and creating a press-down action on the strip. Both hydraulic cylinders are equipped with LVDT displacement sensors to collect their response displacement in real time, ensuring symmetrical press-down or performing necessary wedge-shaped pressure differential compensation.

[0083] Then, the pressure sensor obtains the rolling force change, inputs it into the controller to form a new error, and continues to adjust the pressing amount. The displacement sensor measures the actual response displacement of the hydraulic cylinder and compares it with the control target to form an internal feedback verification closed loop of the control system, and sets the dead zone tolerance interval. If the error exceeds the threshold, the fine-tuning PID will be activated; if an abnormality occurs, it enters the "protection pressing mode" to prevent overpressure or curling defects, forming a self-closed-loop pressing force control mechanism, making adaptive compensation for errors, and improving thickness consistency.

[0084] The plate shape adaptive adjustment and bending roller control module dynamically compensates for the middle waves and side waves to achieve the target plate shape.

[0085] Specifically, in the plate shape adaptive adjustment and bending roll control module, a laser displacement sensor array is first arranged above the strip exit section to detect the ups and downs of the strip at multiple points in the horizontal direction (X-axis) in real time. At the same time, an infrared / photoelectric width sensor is synchronously configured to accurately obtain the horizontal width of the strip and perform normalization processing. Then, the data collected by the LDS is combined with the width data to construct a time-series plate shape profile diagram, realize real-time monitoring of the strip shape, and obtain ups and downs data at multiple positions, providing a data basis for plate shape type judgment.

[0086] Furthermore, an edge-to-center ratio analysis algorithm is deployed within the central control unit. This algorithm calculates the height variation ratio between the edge and center areas on both sides of the plate curve. If the edges are higher and the center is lower, it is identified as edge wave; if the center is higher and lower on both sides, it is identified as center wave. If the curvature is close to 0, it is identified as a straight plate. At the same time, the edge-to-center ratio results are updated every T seconds to form a wave shape recognition sequence. This enables automatic recognition and classification of different types of wave shapes, providing directional judgment criteria for subsequent roll bending control.

[0087] Afterwards, based on the judgment results, the PLC issues a control instruction to adjust the bending roller hydraulic servo cylinder at the corresponding position:

[0088] For edge waves: increase the bending roller pressure on both sides and apply force toward the center of the strip;

[0089] For the middle waves: increase the bending roller pressure in the middle and apply force to both sides.

[0090] At the same time, the bending roll hydraulic cylinder is equipped with a displacement / force sensor to achieve closed-loop position and force feedback; and the control system uses PID regulation + feedforward compensation to enable the bending roll force to dynamically track the target value.

[0091] Finally, the RMS amplitude change rate of the strip shape within a fixed time period T2 is calculated; if the change rate exceeds the limit, the secondary bending roll correction is started; at the same time, the wave shape judgment, bending roll force output, and strip parameters of this cycle are recorded for subsequent training data accumulation.

[0092] The elongation automatic tracking control module controls the elongation of the strip to improve flatness and mechanical consistency.

[0093] Specifically, in the elongation automatic tracking control module, a high-precision LVDT displacement sensor or rotary encoder is first installed at the strip inlet and outlet to record the inlet section length Lin and the outlet section length Lout. Then, the actual elongation ε is periodically calculated according to the formula:

[0094]

[0095] Furthermore, the target elongation εtarget is set based on the process parameter table or database, and the elongation error Δε is calculated:

[0096] Δε=ε target -ε actual ;

[0097] According to the deviation results, the pressing force is automatically adjusted to increase the pressing force to promote extension or the main motor linear speed ratio is adjusted to increase the outlet speed to achieve stretching.

[0098] Afterwards, the fuzzy PID or Bang-Bang controller is called to determine the pressure correction step size based on the extension deviation and trend. If the deviation continues to exceed the limit, the emergency speed reduction or tension release mechanism is activated to avoid steel strip tearing or accumulation. The elongation closed-loop control is achieved to ensure that the strip still maintains the set flatness and strain stability under the dynamic rolling state.

[0099] The feedback optimization and collaborative integration module dynamically optimizes process parameters, adjusts priorities, issues alarms, and learns.

[0100] Specifically, in the feedback optimization and collaborative integration module, real-time sensor data, control output status, and fault alarm codes from other modules are first read through centralized collection;

[0101] The master-slave logic control tree of the above module data is established through the central controller PLC + industrial PC; the master-slave control tree structure is as follows:

[0102] PLC is the main control unit;

[0103] The submodule is set as a slave;

[0104] All communications are synchronized and time-stamped in real time.

[0105] Afterwards, the production data of each cycle is input into the data-driven model to determine the deviation of the operating status; if the strategy is found to be invalid or has long-term deviations, the expert adjustment template in the historical database is retrieved; fuzzy control or reinforcement learning is used to automatically correct the control strategy, so that the system has the ability of "perception-decision-learning-correction" to achieve intelligent collaborative control.

[0106] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0107] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0108] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application of the technical solution and the invention constraints. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0109] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0110] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0111] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An automatic control system for a hot strip skin pass mill, characterized in that ； It includes: microstructure regulation and cooling control module, pressing force electro-hydraulic closed-loop control module, plate shape adaptive adjustment and bending roll control module, elongation automatic tracking control module, feedback optimization and collaborative integration module, and signal connection between each module; The microstructure regulation and cooling control module is mainly used to extract the chemical composition and process information of steel grades, build a multi-dimensional sensor data channel, predict the microstructure evolution using the phase change kinetics model, and control the spray cooling flow and time in different zones. The electro-hydraulic closed-loop control module for the pressing force is mainly used to calculate the target rolling force, adjust the electro-hydraulic proportional valve in a closed loop to control the action of the hydraulic cylinder, and dynamically compensate for the pressing error; The plate shape adaptive adjustment and bending roll control module constructs the plate shape profile time series data, identifies the edge wave, middle wave and straight state, and dynamically calculates the bending roll force distribution strategy; The elongation automatic tracking control module calculates the deviation between the real-time elongation and the target, automatically adjusts the pressing force and the main motor linear speed, adjusts the step length, and triggers speed reduction and tension release when the limit is exceeded; The feedback optimization and collaborative integration module builds a master-slave control architecture, inputs data to drive the model to determine operating deviations, and adjusts the control strategy.

2. The automatic control system for a hot strip skin-pass mill according to claim 1, characterized in that: In the microstructure regulation and cooling control module, the steel coil furnace information is called to identify the risk level of transverse bending of the strip; then the cooling path required for the quasi-acicular ferrite structure is calculated, the cooling section is divided into multiple spray sections, and temperature measurement areas are set.

3. The automatic control system for a hot strip skin-pass mill according to claim 2, characterized in that: In the microstructure regulation and cooling control module, the roll load and coiling outlet temperature data in the strip spray zone are collected to form a real-time multivariable input vector and construct the Avrami dynamic model.

4. The automatic control system for a hot strip skin-pass mill according to claim 3, characterized in that: In the microstructure regulation and cooling control module, the chemical composition of the strip is monitored as input to predict the phase fraction. A graph structure is constructed, and the positive and negative sample comparison losses are designed to identify the normal path vs. phase transformation deviation, and output the phase transformation correction value: The target phase fraction, i.e., the upper limit of acicular ferrite, is set, the phase fraction deviation is calculated, and the flow rate and opening time of each nozzle section are adjusted in real time to bring the tissue drift back to the target path.

5. The automatic control system for a hot strip skin-pass mill according to claim 4, characterized in that: In the microstructure regulation and cooling control module, an extended form is constructed; the phase change equation coefficients are dynamically adjusted to adapt to batch differences; then, the tissue abnormality indication function is defined to define the tissue drift index.

6. The automatic control system for a hot strip skin-pass mill according to claim 5, characterized in that: In the electro-hydraulic closed-loop control module for the pressing force, the actual rolling force applied to the workpiece on both sides during the rolling process is collected and input into the PLC or industrial control master station; the target rolling force value is then calculated and the PID controller parameters are optimized; The PID output instruction is physically converted into a pressing behavior, and the dynamic response is completed through electro-hydraulic coordination; the input controller forms a new error, continues to adjust the pressing amount, and sets the dead zone tolerance range.

7. The automatic control system for a hot strip skin-pass mill according to claim 6, characterized in that: In the plate shape adaptive adjustment and bending roll control module, the height fluctuations of the strip at multiple horizontal points are detected to construct a time-series plate shape profile; the height change ratio between the edge area and the center area on both sides of the plate shape curve is calculated, and the bending roll hydraulic servo cylinder at the corresponding position is adjusted.

8. The automatic control system for a hot strip skin-pass mill according to claim 7, characterized in that: In the elongation automatic tracking control module, the inlet section length Lin and the outlet section length Lout are recorded, the actual elongation is calculated periodically, the target elongation is set, and the elongation error is calculated to adjust the pressing force.

9. The automatic control system for a hot strip skin-pass mill according to claim 8, characterized in that: Read real-time sensor data, control output status and fault alarm codes from other modules; establish a master-slave logic control tree.

Images