Method and system for high-precision dynamic compensation of cold rolling thickness of ultra-thin aluminum alloy strip

Abstract

The application discloses a cold rolling thickness high-precision dynamic compensation method and system for ultrathin aluminum alloy strips, and the method comprises the following steps: constructing a digital twin model to predict roll thermal crown and generate feedforward compensation; adopting Kalman filtering to fuse multi-sensor data to realize high-precision eccentricity real-time compensation; completing shape-thickness decoupling control through a target crown priority strategy; superimposing feedforward and feedback instructions to drive a high-frequency hydraulic servo valve to quickly adjust a roll gap; and the compensation system comprises a digital twin server, a real-time controller, a high-precision sensor group, a shape-thickness collaborative execution mechanism and a rolling mill control system. The application can simultaneously inhibit thermal crown slow time-varying and eccentricity fast time-varying interference, reduce the shape-thickness coupling degree to below 5%, stabilize the thickness precision of 0.14mm-level ultrathin aluminum strips to be within 1 mu m, reduce the overproof roll proportion from 8% to below 1.5%, and make the plate crown qualified rate reach 99%, so that the rolling precision and production stability of the ultrathin aluminum alloy strips are significantly improved, and the application is suitable for the production of high-end printing plate base ultrathin aluminum materials.

Description

Technical Field

[0001] This invention relates to the field of aluminum strip rolling technology, specifically to a method and system for high-precision dynamic compensation of cold-rolled thickness for ultra-thin aluminum alloy strips. Background Technology

[0002] In the production of ultra-thin aluminum alloy strips for printing plate substrates, 0.14mm grade CTP substrates require extremely high thickness uniformity, with a target control accuracy of ±1μm. However, in the actual rolling process, multiple factors such as mill stiffness fluctuations, strong coupling between strip shape and thickness, slow changes in roll thermal crown, periodic disturbances due to roll eccentricity, and fluctuations in rolling speed and tension combine to make thickness deviation difficult to control.

[0003] Existing thickness control technologies have limitations: 1. Traditional AGC and local compensation schemes only provide local compensation for non-steady-state stages such as acceleration and deceleration, and do not cover the continuous thickness fluctuations caused by the coupling of multiple factors under steady-state rolling, so the thickness fluctuations continue to exist. 2. The strong coupling between plate shape and thickness means that thickness adjustment can easily damage the plate shape, while plate shape adjustment can cause thickness deviation, and the two cannot work together stably. 3. Current rolling control is developing towards intelligence, but existing high-level algorithms such as deep learning do not provide millisecond-level execution paths, core decoupling logic, and field-implementable control architectures, making it difficult to achieve stable and high-precision compensation in ultra-thin strip rolling.

[0004] Therefore, developing a high-precision dynamic compensation method and system that can simultaneously handle slow time-varying thermal interference, fast time-varying eccentric interference, and plate shape-thickness coupling has become an urgent technical problem to be solved in this field. Summary of the Invention

[0005] The purpose of this invention is to provide a high-precision dynamic compensation method and system for cold-rolled thickness of ultra-thin aluminum alloy strip. By constructing a closed-loop control scheme of model prediction feedforward + data fusion feedback + multivariable decoupling, it solves the problems of thickness gradual change caused by slow time-varying change of roll thermal crown that cannot be compensated in advance, insufficient accuracy of compensation for fast time-varying disturbance of roll eccentricity, strong coupling between strip shape and thickness leading to damage to strip shape in thickness control, and traditional AGC only compensates after the fact, with lag response and poor steady-state accuracy.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A high-precision dynamic compensation method for cold-rolled thickness of ultra-thin aluminum alloy strip includes the following steps: S1 constructs a digital twin model of the rolling process, inputs rolling plan parameters, predicts the trend of roll thermal crown change, and generates thermal crown feedforward compensation. S2 collects the roll angular position signal, strip inlet thickness signal and outlet thickness signal, performs multi-sensor data fusion through Kalman filtering algorithm, calculates roll eccentricity in real time and outputs real-time eccentricity compensation amount; S3 adopts a target convexity priority strategy, based on feedback from a multi-channel strip shaper, and stabilizes the strip shape through hydraulic bending rollers and a 32-zone cooling system. At the same time, it performs strip shape-thickness decoupling control to reduce coupling interference. S4 superimposes the feedforward compensation amount with the real-time feedback compensation amount, driving the high-frequency response hydraulic servo mechanism to adjust the roll gap, thereby achieving dynamic thickness compensation.

[0007] Furthermore, in S1, the digital twin model of the rolling process is constructed based on the physical characteristics of the rolling mill, the thermodynamic characteristics of the roll system, and historical rolling data. It is used to predict the rolling force distribution and the roll gap change caused by the thermal expansion of the rolls. The prediction accuracy of the digital twin model of the rolling process for the thermal crown of the rolls is ±0.3μm.

[0008] Furthermore, in S2, the roll angular position signal comes from the encoder, and the strip inlet thickness signal and outlet thickness signal come from the thickness gauges set at the inlet and outlet respectively. The Kalman filter algorithm fuses the real-time data from the encoder and the thickness gauge to output the roll eccentricity compensation amount, with a compensation accuracy of ±0.5μm.

[0009] Furthermore, in S3, the shape-thickness decoupling control is achieved through a target convexity priority strategy, combined with feedback from a multi-channel shape meter and a segmented cooling system.

[0010] Furthermore, in S3, the plate crown is first stabilized within the range of ≤0.5% by hydraulic bending rollers and 32-zone cooling, and then the thickness is fine-tuned to decouple the effect of plate shape adjustment on thickness to below 5%.

[0011] Furthermore, in S4, the response time of the high-frequency hydraulic servo mechanism is ≤5ms, and the final thickness control accuracy is stabilized at ±1μm.

[0012] A high-precision dynamic compensation system for cold-rolled thickness of ultra-thin aluminum alloy strip includes: A digital twin simulation server is used to build high-fidelity rolling models, predict roll thermal crown, and output feedforward compensation amounts. A real-time controller equipped with a Kalman filter algorithm is used for multi-sensor data fusion, eccentricity estimation, and real-time compensation calculations. High-precision sensor array, including roll angular position encoder, inlet laser thickness gauge, and outlet laser thickness gauge; The plate shape detection and execution mechanism includes a multi-channel plate shape meter, a hydraulic bending roller system, a 32-zone cooling system, and a high-frequency response hydraulic servo valve; The integrated execution subsystem is integrated into the primary and secondary control systems of the rolling mill. It is used to receive superimposed compensation commands and execute roll gap adjustments to achieve dynamic compensation.

[0013] Furthermore, the digital twin simulation server supports more than 30 minutes of pre-simulation of the rolling process, and can proactively compensate for slow time-varying thermal interference.

[0014] Furthermore, the real-time controller can automatically learn the roll eccentricity characteristics during the 1-2 roll rolling process to achieve closed-loop suppression of periodic disturbances.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: The present method and system for high-precision dynamic compensation of cold-rolled thickness of ultra-thin aluminum alloy strip has the following advantages: 1. Precise feedforward compensation for thermal convexity, with digital twin prediction accuracy of ±0.3μm, offsets thickness drift caused by thermal deformation at the source, avoiding lag in post-compensation. 2. The eccentricity compensation accuracy has been greatly improved, from the traditional ±1.5μm to ±0.5μm, which significantly suppresses periodic thickness fluctuations; 3. Strong decoupling of plate shape and thickness, controlling coupling degree <5%, plate convexity stable ≤0.5%, achieving high-precision control of both plate shape and thickness; 4. Excellent overall control performance, with a stable thickness accuracy of ±1μm for 0.14mm ultra-thin CTP substrate, the proportion of rolls with thickness deviation reduced from 8% to below 1.5%, and a board convexity qualification rate of 99%, meeting the requirements for stable mass production of high-end CTP substrates; 5. The entire process is closed-loop, realizing feedforward + feedback + decoupled collaboration, with millisecond-level response and high accuracy in both steady and unsteady states. Detailed Implementation

[0016] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] A high-precision dynamic compensation method for cold-rolled thickness of ultra-thin aluminum alloy strip includes the following steps: S1 constructs a digital twin model of the rolling process, inputs rolling plan parameters, predicts the trend of roll thermal crown change, and generates thermal crown feedforward compensation. The digital twin model of the rolling process is constructed based on the physical characteristics of the rolling mill, the thermodynamic characteristics of the roll system, and historical rolling data. It is used to predict the rolling force distribution and the roll gap change caused by the thermal expansion of the roll. The prediction accuracy of the digital twin model of the rolling process for roll thermal crown is ±0.3μm. S2 acquires the roll angular position signal, strip inlet thickness signal, and outlet thickness signal. It performs multi-sensor data fusion through a Kalman filter algorithm, calculates the roll eccentricity in real time, and outputs the real-time eccentricity compensation amount. The roll angular position signal comes from the encoder, and the strip inlet thickness signal and outlet thickness signal come from thickness gauges set at the inlet and outlet, respectively. The Kalman filter algorithm fuses the real-time data from the encoder and thickness gauges and outputs the roll eccentricity compensation amount with a compensation accuracy of ±0.5μm. S3 employs a target convexity-first strategy, based on feedback from a multi-channel shape gauge. It stabilizes the strip shape through hydraulic bending rollers and a 32-segment cooling system, while simultaneously implementing shape-thickness decoupling control to reduce coupling interference. This shape-thickness decoupling control is achieved through the target convexity-first strategy, combined with multi-channel shape gauge feedback and a segmented cooling system. First, the shape convexity is stabilized within ≤0.5% using hydraulic bending rollers and 32-segment cooling. Then, thickness fine-tuning is performed, decoupling the impact of shape adjustment on thickness to below 5%. S4 superimposes the feedforward compensation amount and the real-time feedback compensation amount to drive the high-frequency response hydraulic servo mechanism to adjust the roll gap, thereby achieving dynamic thickness compensation. The response time of the high-frequency response hydraulic servo mechanism is ≤5ms, and the final thickness control accuracy is stabilized at ±1μm.

[0018] A high-precision dynamic compensation system for cold-rolled thickness of ultra-thin aluminum alloy strip includes: The digital twin simulation server is used to build high-fidelity rolling models, predict roll thermal crown, and output feedforward compensation. It supports more than 30 minutes of pre-simulation of the rolling process and can proactively compensate for slow time-varying thermal interference. The real-time controller, equipped with a Kalman filter algorithm, is used for multi-sensor data fusion, eccentricity estimation and real-time compensation calculation. It can automatically learn the roll eccentricity characteristics within 1-2 rolls of rolling process and achieve closed-loop suppression of periodic disturbances. High-precision sensor array, including roll angular position encoder, inlet laser thickness gauge, and outlet laser thickness gauge; The plate shape detection and execution mechanism includes a multi-channel plate shape meter, a hydraulic bending roller system, a 32-zone cooling system, and a high-frequency response hydraulic servo valve; The integrated execution subsystem is integrated into the primary and secondary control systems of the rolling mill. It is used to receive superimposed compensation commands and execute roll gap adjustments to achieve dynamic compensation.

[0019] Example 1: Production of 1050A-H18CTP sheet base with specifications of 0.14mm × 1250mm using a six-roll CVC cold rolling mill A high-precision dynamic compensation method for cold-rolled thickness of ultra-thin aluminum alloy strip includes the following steps: (1) System preparation and model pre-run After the rolling plan is issued, the digital twin module loads parameters such as target thickness 0.14mm, width 1250mm, initial roll temperature 38℃, and rolling speed 600m / min; it simulates the rolling process for the next 30 minutes and predicts that the thermal crown of the support roll will increase by 0.07mm in the 18th minute of rolling; the system automatically generates a continuous feedforward compensation curve; (2) Rolling start-up and dynamic compensation Threading → Establishing tension → Mill speed-up operation; 1-2 rolls before rolling: The Kalman filter module quickly acquires encoder signals, learns and locks roll eccentricity characteristics, including eccentricity ±8μm, period corresponding to roll circumference, and outputs eccentricity compensation commands in real time to suppress thickness fluctuations caused by eccentricity. Rolling to the 18th minute: The digital twin feedforward compensation is automatically superimposed to offset the thickness variation deviation caused by the increase in thermal crown; 3) Interference response and coordinated control During rolling, a cooling nozzle suddenly became blocked, resulting in a localized 5I-shaped wave pattern. Priority action of the plate shape control module: Adjust the cooling flow of adjacent zones and correct the bending roller force to quickly eliminate waviness; The thickness control module synchronously senses the coupling effect and finely adjusts the roll gap in the reverse direction through a decoupling algorithm; finally, the plate shape is restored to qualified, and the exit thickness deviation is always kept within ±0.8μm.

[0020] (4) Effect verification Results of online inspection and offline sampling for a continuous rolling shift (8 hours): Strip thickness accuracy: 0.14±0.001mm; Percentage of rolls with thickness exceeding tolerance: 1.5%; Plate convexity pass rate: 99%; It fully meets the high-end quality requirements of ultra-thin CTP substrates.

[0021] Example 2: Production of 0.12mm × 1000mm 3003 battery aluminum foil base material using a four-roll irreversible cold rolling mill A high-precision dynamic compensation method for cold-rolled thickness of ultra-thin aluminum alloy strip includes the following steps: (1) System preparation and model pre-run Switching to 3003 alloy, 0.12mm thickness, and 1000mm width specifications, the digital twin model is automatically initialized based on historical data of the same type, and quickly reconstructed by combining the current roll temperature, rolling force limit, and tension regime; a 20-minute rolling process is pre-simulated, predicting that the work roll thermal crown will increase by 0.05mm in the 12th minute of rolling, and generating the corresponding feedforward compensation curve; (2) Rolling start-up and dynamic compensation After startup, the rolling speed is kept steady at 450 m / min. The real-time controller completes the eccentricity feature extraction during the first roll rolling process, with an eccentricity of ±6 μm. The Kalman filter fuses the thickness gauge and encoder signals in real time, and eccentricity compensation is applied throughout the process, resulting in no periodic ripple in the exit thickness. (3) Interference response and coordinated control During the rolling process, due to fluctuations in coiling tension, the speed varies slightly in the range of 420-480m / min; the plate shape-thickness decoupling module calculates the speed-tension-thickness coupling relationship in real time and simultaneously corrects the bending roll force and roll gap compensation; the plate crown is controlled within 0.3%-0.45% throughout the process, and the thickness deviation does not exceed ±0.9μm.

[0022] (4) Effect verification After 6 hours of continuous production, the product stabilized within 30 minutes of changing specifications. Online testing and offline sampling results are as follows: Thickness accuracy: 0.12±0.001mm; Percentage of defective rolls: 1.2%; Plate shape qualification rate: 99.2%; It meets the high precision requirements of aluminum foil base material for power batteries.

[0023] Example 3: Production of 0.16mm × 1500mm 8011 air conditioner foil blanks using a six-roll HC cold rolling mill A high-precision dynamic compensation method for cold-rolled thickness of ultra-thin aluminum alloy strip includes the following steps: (1) System preparation and model pre-run For a wide product with a target thickness of 0.16mm and a width of 1500mm, the digital twin model focuses on predicting the uneven distribution of thermal convexity in the width direction; it predicts that the thermal convexity in the middle will be larger at the 22nd minute of rolling, and generates a middle strengthening compensation curve in advance. (2) Rolling start-up and dynamic compensation The system employs a dual-thickness gauge redundancy configuration and uses Kalman filtering to fuse the two thickness signals, enhancing anti-interference capabilities. The roll eccentricity is ±7μm, and after compensation, the thickness fluctuation caused by eccentricity is reduced to within ±0.4μm. (3) Interference response and coordinated control For a width of 1500mm, a 32-zone full-area cooling fine adjustment is enabled; when slight wavy edges occur, the shape system prioritizes the correction of edge cooling and bending rollers, while the thickness system simultaneously decouples and compensates to ensure that the thickness difference between the center and the edge is ≤0.8μm. (4) Effect verification Results of online testing and offline sampling after 8 hours of continuous production: Thickness accuracy: 0.16±0.001mm; Full-width thickness uniformity: ≤0.7μm; Excess rate: 1.0%; fully meets the production requirements for high-end air conditioning foil blanks.

[0024] The core innovation of this invention lies in constructing a closed-loop system integrating digital twin simulation, real-time Kalman filtering fusion, and shape-thickness decoupled control, achieving integrated dynamic compensation for multi-source errors such as roll eccentricity and thermal expansion through feedforward and feedback. Specifically: 1. Digital Twin-Driven Multi-Factor Coupled Compensation Model: Unlike schemes that only compensate for acceleration and deceleration, this invention first establishes a high-fidelity digital twin model based on software such as ANSYS / DEFORM and a large amount of historical data, incorporating rolling force, tension, speed, and the thermal behavior of the roll system. This model can proactively predict slow time-varying disturbances such as changes in roll thermal crown (compensation accuracy ±0.3μm), achieving feedforward compensation; 2. Real-time data fusion and eccentricity compensation based on Kalman filtering: During the rolling process, the Kalman filtering algorithm is used to fuse data from multiple sensors such as high-precision encoders and laser thickness gauges in real time to perform optimal estimation and real-time compensation for roll eccentricity (periodic fast time-varying interference), improving the eccentricity compensation accuracy from the common ±1.5μm to ±0.5μm; 3. Coordinated Decoupling Control of Plate Shape and Thickness: An innovative target crown priority strategy is adopted, and multi-channel plate shape meter data is used to reduce the coupling between plate shape control and thickness control to below 5% through algorithms. Combined with precision grinding of support roller crown (crown 0.05-0.08mm) and precise control of 32-zone cooling nozzles, stable conditions for thickness control are created while ensuring plate crown ≤0.5%.

[0025] This invention can stabilize the thickness control accuracy of 0.14mm strip at ±1μm (international advanced level), and reduce the proportion of rolls with thickness deviation from 8% to below 2%, effectively supporting the stable mass production of ultra-thin CTP substrates.

[0026] The above are merely embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A high-precision dynamic compensation method for cold-rolled thickness of ultra-thin aluminum alloy strip, characterized in that, Includes the following steps: S1 constructs a digital twin model of the rolling process, inputs rolling plan parameters, predicts the trend of roll thermal crown change, and generates thermal crown feedforward compensation. S2 collects the roll angular position signal, strip inlet thickness signal and outlet thickness signal, performs multi-sensor data fusion through Kalman filtering algorithm, calculates roll eccentricity in real time and outputs real-time eccentricity compensation amount; S3 adopts a target convexity priority strategy, based on feedback from a multi-channel strip shaper, and stabilizes the strip shape through hydraulic bending rollers and a 32-zone cooling system. At the same time, it performs strip shape-thickness decoupling control to reduce coupling interference. S4 superimposes the feedforward compensation amount with the real-time feedback compensation amount, driving the high-frequency response hydraulic servo mechanism to adjust the roll gap, thereby achieving dynamic thickness compensation.

2. The method for high-precision dynamic compensation of cold-rolled thickness for ultra-thin aluminum alloy strip according to claim 1, characterized in that, In S1, the digital twin model of the rolling process is constructed based on the physical characteristics of the rolling mill, the thermodynamic characteristics of the roll system, and historical rolling data. It is used to predict the rolling force distribution and the roll gap change caused by the thermal expansion of the rolls. The prediction accuracy of the digital twin model of the rolling process for the thermal crown of the rolls is ±0.3μm.

3. The method for high-precision dynamic compensation of cold-rolled thickness for ultra-thin aluminum alloy strip according to claim 1, characterized in that, In S2, the roll angular position signal comes from the encoder, and the strip inlet thickness signal and outlet thickness signal come from the thickness gauges set at the inlet and outlet respectively. The Kalman filter algorithm fuses the real-time data from the encoder and thickness gauge to output the roll eccentricity compensation amount with a compensation accuracy of ±0.5μm.

4. The method for high-precision dynamic compensation of cold-rolled thickness for ultra-thin aluminum alloy strip according to claim 1, characterized in that, In S3, the shape-thickness decoupling control is achieved through a target convexity priority strategy, combined with feedback from a multi-channel shape meter and a segmented cooling system.

5. The method for high-precision dynamic compensation of cold-rolled thickness for ultra-thin aluminum alloy strip according to claim 4, characterized in that, First, the convexity of the plate is stabilized within ≤0.5% by hydraulic bending rollers and 32-zone cooling. Then, the thickness is fine-tuned to decouple the effect of plate shape adjustment on thickness to below 5%.

6. The method for high-precision dynamic compensation of cold-rolled thickness for ultra-thin aluminum alloy strip according to claim 1, characterized in that, In S4, the response time of the high-frequency hydraulic servo mechanism is ≤5ms, and the final thickness control accuracy is stable at ±1μm.

7. A compensation system for implementing the high-precision dynamic compensation method for cold-rolled thickness of ultra-thin aluminum alloy strip according to any one of claims 1-6, characterized in that, A digital twin simulation server is used to build high-fidelity rolling models, predict roll thermal crown, and output feedforward compensation amounts. A real-time controller equipped with a Kalman filter algorithm is used for multi-sensor data fusion, eccentricity estimation, and real-time compensation calculations. High-precision sensor array, including roll angular position encoder, inlet laser thickness gauge, and outlet laser thickness gauge; The plate shape detection and execution mechanism includes a multi-channel plate shape meter, a hydraulic bending roller system, a 32-zone cooling system, and a high-frequency response hydraulic servo valve; The integrated execution subsystem is integrated into the primary and secondary control systems of the rolling mill. It is used to receive superimposed compensation commands and execute roll gap adjustments to achieve dynamic compensation.

8. The high-precision dynamic compensation system for cold-rolled thickness of ultra-thin aluminum alloy strip according to claim 7, characterized in that, The digital twin simulation server supports more than 30 minutes of pre-simulation of the rolling process and can proactively compensate for slow time-varying thermal interference.

9. The high-precision dynamic compensation system for cold-rolled thickness of ultra-thin aluminum alloy strip according to claim 7, characterized in that, The real-time controller can automatically learn the roll eccentricity characteristics during the 1-2 roll rolling process to achieve closed-loop suppression of periodic disturbances.