Plate shape control process for multi-roller continuous rolling of non-oriented silicon steel

Abstract

The invention discloses a non-oriented silicon steel multi-roller continuous rolling plate shape control process which comprises the following steps: establishing a multi-dimensional dynamic model, and constructing a rolled piece-roller-cooling multi-dimensional physical field coupling model; designing a multi-section polynomial roll-shaped structure, and adjusting parameters of each section of a roll-shaped curve according to the characteristic parameters of the plate strip; setting a dynamic control weight, and adjusting the control priorities of roll bending, roll shifting and rolling force in real time according to the rolling stage; executing plate shape prediction and correction, and updating a control quantity influence coefficient by utilizing a self-adaptive correction technology; and performing plate shape evaluation and optimization, and performing multi-objective optimization according to objective functions of plate shape deviation, energy consumption and roller consumption. By constructing a rolled piece-roller-cooling multi-dimensional physical field coupling model, adopting a multi-section polynomial roller shape structure and a dynamic weight adjustment technology and combining a self-adaptive correction and multi-objective optimization method, the high-order wave shape control problem is effectively solved, and accurate plate shape control in the whole rolling process is achieved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of metal material rolling processing, in particular to a plate shape control process for multi-roll continuous rolling of non-oriented silicon steel. BACKGROUND

[0002] With the rapid development of industry and the increasing strictness of environmental protection requirements, high-strength and thin-gauge plate and strip materials are increasingly applied in the fields of automobiles, precision instruments, household appliance packaging, etc., and the quality and precision requirements for plate and strip materials are also continuously improved. In the production of cold-rolled strip steel, plate shape control, as one of the key technologies, directly affects the quality of products and the smooth progress of subsequent processes. However, with the continuous thinning of plate and strip thickness, the rolling force of the finished product rack increases, and the roll is more prone to high-order deflection, even causing the roll end to be pressed, and the friction between the roll and the thin strip increases, which seriously affects the plate shape control effect.

[0003] At present, the conventional plate shape control technology mainly includes five ways of roll inclination, symmetrical bending of work rolls, asymmetrical bending of work rolls, symmetrical bending of intermediate rolls, and symmetrical shifting of intermediate rolls. These methods can control the plate shape to a certain extent, but it is difficult to effectively solve the problem of high-order local wave shape, especially for the extremely thin plate and strip in the endless rolling process, the conventional technology is not up to the task. In addition, the traditional roll shape design often ignores the influence of the rolling non-steady state process within the roll changing period on the plate shape, resulting in poor product quality.

[0004] In order to solve these problems, researchers have begun to explore new plate shape control technologies. For example, some studies have proposed a plate shape intelligent controller based on PSO-BP network, a plate shape dynamic influence matrix control method using cloud adaptive difference and BP neural network, etc. However, these methods still have limitations in dealing with complex wave shapes, and it is difficult to achieve high-precision plate shape control. Therefore, it is urgent to develop a new plate shape control algorithm that can comprehensively consider the multi-physical field coupling, roll shape design, cooling control and other factors in the rolling process, in order to improve the overall plate shape quality and production efficiency of plate and strip materials.

[0005] The prior art has the following disadvantages: the traditional plate shape control technology is difficult to effectively solve the high-order local wave shape problem, especially in the process of endless rolling, the conventional method cannot meet the plate shape control requirements of ultra-thin plate strips, resulting in poor product quality. The existing plate shape control method has limitations in dealing with complex wave shapes, and it is difficult to achieve high-precision plate shape control, which cannot meet the application requirements of high-strength plate strip materials in the fields of automobiles, precision instruments and the like. The current plate shape control technology often ignores the influence of the rolling non-steady state process within the roll changing period on the plate shape, and lacks real-time monitoring and accurate control capability for the dynamic changes of the plate shape. The existing technology has problems of insufficient accuracy and real-time in obtaining the asymmetric position of the roll system, and it is difficult to fully utilize the plate strip deformation information and plate shape control information. The traditional plate shape control method still has the problem of insufficient accuracy in plate shape regulation capability, which affects the overall quality of the product and the smooth progress of the subsequent process.

[0006] In view of the above problems, the prior art needs to be improved. SUMMARY

[0007] The purpose of the present application is to provide a plate shape control process for multi-roll continuous rolling of non-oriented silicon steel, which has the advantages of effectively solving high-order local wave shape problems, improving the control precision of non-steady state rolling process, and realizing multi-physical field coupling dynamic optimization.

[0008] The present application provides a plate shape control process for multi-roll continuous rolling of non-oriented silicon steel, and the technical scheme is as follows: including: step 1, establishing a multi-dimensional dynamic model, constructing a multi-dimensional physical field coupling model of the rolled piece-roller-cooling; step 2, designing a multi-segment polynomial roll shape structure, adjusting the parameters of each segment of the roll shape curve according to the characteristic parameters of the plate strip; step 3, setting dynamic control weights, adjusting the control priorities of the bending roll, the roll shifting and the rolling force according to the rolling stage; step 4, performing plate shape prediction and correction, updating the control quantity influence coefficient by using adaptive correction technology; step 5, performing plate shape evaluation and optimization, performing multi-objective optimization according to the target functions of plate shape deviation, energy consumption and roll consumption.

[0009] Further, the present application further provides that step 1 includes: step 101, establishing a rolled piece physical model to describe the metal flow and deformation behavior of the plate strip; step 102, constructing a roller physical model considering the physical field of roller deflection and thermal expansion; step 103, establishing a cooling system model to describe the influence of cooling water flow and temperature parameters on the plate shape; step 104, setting the model boundary conditions, including the incoming state of the plate strip and the given parameters of the rolling force.

[0010] Further, the application further proposes that step 2 comprises: step 201, dividing the roll shape curve into multiple functional areas, including an entry transition area, a roll pressing area, and an exit transition area; step 202, designing a circular arc chamfer curve in the entry transition area to reduce stress concentration; step 203, adopting a polynomial multi-segment line segment combination structure in the roll pressing area to realize efficient edge drop control; step 204, designing a straight line chamfer curve in the exit transition area to ensure the stability of the plate shape; and step 205, ensuring the continuity of the roll shape curve through the constraint condition that the first derivatives of the end points are equal.

[0011] Further, the application further proposes that step 3 comprises: step 301, setting initial control weights, including weight distribution in the steady rolling, threading stage, and tail throwing stage; step 302, dynamically adjusting the weights according to real-time plate shape deviation, energy consumption, and roll consumption indicators; and step 303, optimizing each control parameter through a collaborative control algorithm to make the target function optimal.

[0012] Further, the application further proposes that step 4 comprises: step 401, collecting real-time plate shape measurement data, including middle wave shape and edge plate shape parameters; step 402, updating the control amount according to the deviation of the actual plate shape from the target plate shape by using an adaptive correction technology; and step 403, analyzing the causes of the plate shape formation through a plate shape prediction model to optimize the control strategy.

[0013] Further, the application further proposes that step 5 comprises: step 501, calculating plate shape control effect, including plate shape deviation and edge thinning amount indicators; step 502, evaluating energy consumption and roll consumption economic indicators; step 503, performing nonlinear optimization according to multi-objective functions to seek a Pareto optimal solution; and step 504, verifying the optimization result through a digital twin platform to ensure the feasibility in actual production.

[0014] Further, the application further proposes that step 101 comprises: establishing a physical model of the strip based on strip specification parameters by using a finite element method, including a geometric model and a material model of the strip, wherein the geometric model includes strip length, width, and thickness geometric dimensions, and the material model includes elastic modulus, Poisson's ratio, and density material parameters.

[0015] Further, the application further proposes that step 102 comprises: establishing a geometric model and a material model of the roll based on roll specification parameters, wherein the geometric model includes roll diameter and work roll face length parameters, and the material model includes elastic modulus and thermal expansion coefficient parameters.

[0016] Further, the application further proposes that step 103 comprises: establishing a physical model of the cooling system based on specification parameters of the cooling system, including cooling water flow, nozzle layout, and cooling efficiency parameters.

[0017] Further, the application also proposes that step 104 comprises: setting boundary conditions of the model according to production process requirements, including strip incoming thickness, width, speed parameters, and rolling force, roll shifting amount, and bending roll angle control parameters.

[0018] Compared with the prior art, the application provides a plate shape control process for multi-roll continuous rolling of non-oriented silicon steel, which has the following beneficial effects:

[0019] 1. A multi-segment polynomial roll shape design is adopted, and the continuity of the roll shape curve is ensured through the constraint condition that the first-order derivatives of the endpoints are equal, effectively dealing with high-order local wave shape problems, significantly improving the accuracy of plate shape control, and being particularly suitable for plate shape control of extremely thin strips in endless rolling process, effectively solving the problem that the traditional method is not satisfactory;

[0020] 2. A multi-dimensional dynamic model is established, including rolled piece-roller-cooling, etc., realizing comprehensive consideration of multi-physical field coupling in the rolling process, breaking through the limitation of traditional methods that only focus on a single factor, and improving the comprehensiveness and accuracy of plate shape control;

[0021] 3. The priorities of bending roll, roll shifting, and rolling force control are adjusted in real time according to the rolling stage, and through a dynamic weight distribution mechanism, the non-steady state problem in the rolling process is effectively solved, and the real-time performance and response speed of plate shape control are improved;

[0022] 4. The model drift problem caused by roll wear is overcome by using plate shape prediction model adaptive correction technology to update the control amount influence coefficient online, and the stability and reliability of plate shape control are improved;

[0023] 5. Through technologies such as transverse stiffness gradient control, cooling-deformation coordination, and edge drop lag compensation, the problems of edge stress concentration and plate shape stability are effectively solved, and the overall quality of the strip material is significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 A step flowchart of the plate shape control process for multi-roll continuous rolling of non-oriented silicon steel of the application;

[0025] Figure 2 A flowchart of step 1 of the plate shape control process for multi-roll continuous rolling of non-oriented silicon steel of the application;

[0026] Figure 3 A flowchart of step 2 of the plate shape control process for multi-roll continuous rolling of non-oriented silicon steel of the application;

[0027] Figure 4 A flowchart of step 3 of the plate shape control process for multi-roll continuous rolling of non-oriented silicon steel of the application;

[0028] Figure 5Flow chart for step 4 of the plate shape control process of the non-oriented silicon steel multi-roll continuous rolling of the present application;

[0029] Figure 6 Flow chart for step 5 of the plate shape control process of the non-oriented silicon steel multi-roll continuous rolling of the present application. DETAILED DESCRIPTION

[0030] The technical solutions in the present application will be described clearly and completely below in combination with the drawings in the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. The components of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0031] It should be noted that similar reference numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. Meanwhile, in the description of the present application, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0032] In traditional cold rolling plate shape control, the problem of insufficient high-order local wave shape regulation ability is caused by the incompleteness of physical field coupling modeling. The existing technology lacks a detailed description of the multi-dimensional dynamic interaction of the rolled piece-roller-cooling system, which leads to the fact that the correlation mechanism between roll system deflection deformation, thermal field distribution and plate strip deformation is not fully revealed. When the plate strip thickness is reduced to below 0.3mm, the nonlinear characteristics of the rolling pressure distribution are intensified, and the single curve structure in the conventional roll shape design cannot adapt to the differentiated deformation needs of the edge and middle parts of the wide plate strip. At the same time, the spatial and temporal distribution difference of the cooling efficiency in the rolling process will cause local stress concentration, further deteriorating the plate shape stability.

[0033] For example, in the continuous rolling scenario of cold-rolled non-oriented silicon steel, the work roll thermal crown rate of a six-high CVC mill reaches 0.8 pm / min when rolling a 0.25 mm thick, 1200 mm wide strip. At this time, the conventional symmetric bending strategy cannot effectively compensate for the strip center wavy caused by roll deflection, and the edge thinning fluctuation range exceeds ±4 pm. In the later period of the roll changing cycle, when the roll wear accumulates to 150 pm, the control amount of the strip shape closed-loop control system lags behind, and the strip shape deviation I value rises to more than 12 IU. When the rolling speed is increased to 1200 m / min, the uneven distribution of cooling water causes a transverse temperature difference of 15°C in the roll gap area, directly causing edge wave defects.

[0034] If the above problems are not solved, the transverse thickness tolerance of the strip will exceed the process requirement of ±5 pm, causing abnormal grain growth during the annealing process in the subsequent process. In continuous production, uncontrollable fluctuations in edge thinning will cause the edge cutting loss rate in the slitting process to rise to 3.5%, and at the same time, the abnormal wear rate of the roll will increase by 30%. In the endless rolling mode, dynamic fluctuations in the strip shape will also cause the tension control system to become unstable, and in severe cases, it may cause a strip breakage accident, resulting in an increase in unplanned downtime of the production line.

[0035] In the face of the above problems, the present application first aims at the problem of incomplete physical field coupling modeling in traditional strip shape control, explores the feasibility of establishing a multi-dimensional dynamic model, and builds a more accurate physical field coupling model by integrating the interaction of workpiece deformation, roll deflection and cooling system. For the problem that the roll shape design cannot adapt to the thin strip rolling requirement, the possibility of multi-segment roll shape structure is studied, and the improvement effect of curve combination in different functional areas on edge drop control and stress distribution is analyzed. At the same time, in view of the problem of dynamic parameter change in the rolling process, the weight dynamic adjustment mechanism is tried to be introduced, and the priority difference of control parameters in different rolling stages is considered. In addition, in order to solve the model drift problem, the application path of online correction technology is explored, and how to optimize the control parameters through real-time data feedback is studied. After comparing multiple schemes, the comprehensive scheme of combining physical field coupling modeling, multi-segment roll shape design, dynamic weight adjustment and prediction correction technology is finally selected, forming a complete strip shape control system.

[0036] For this purpose, as shown in Figures 1-6 The present application proposes a strip shape control process for multi-roll continuous rolling of non-oriented silicon steel, comprising:

[0037] A multi-dimensional dynamic model is established to build a workpiece-roll-cooling multi-dimensional physical field coupling model;

[0038] A multi-segment polynomial roll shape structure is designed, and the parameters of each segment of the roll shape curve are adjusted according to the characteristic parameters of the strip;

[0039] The dynamic control weight is set, and the control priorities of the bending roll, the roll shifting and the rolling force are adjusted in real time according to the rolling stage.

[0040] The plate shape prediction and correction are performed, and the control quantity influence coefficient is updated by using an adaptive correction technology.

[0041] The plate shape evaluation and optimization are performed, and multi-objective optimization is performed according to the target functions of the plate shape deviation, energy consumption and roll consumption.

[0042] Among them, establishing a multi-dimensional dynamic model means constructing a comprehensive mathematical model including the deformation of the rolled piece, the mechanical response of the roll and the influence of the cooling system. Specifically, the finite element analysis method combined with the thermal-mechanical coupling equation can be used to realize this model, which can accurately reflect the interaction relationship of each physical field in the rolling process and provide a theoretical basis for the subsequent control strategy.

[0043] Among them, designing a multi-segment polynomial roll shape structure means dividing the surface profile of the work roll into different functional areas and performing segmented mathematical description. Specifically, the parametric design of a cubic polynomial curve in different segments can be used to realize this, and by adjusting the curve parameters of each segment, the metal flow behavior of different areas of the plate strip can be targetedly controlled.

[0044] Among them, setting the dynamic control weight means prioritizing the actuators according to different stages of the rolling process. Specifically, the fuzzy control algorithm combined with online monitoring data can be used to realize real-time adjustment of the weight coefficient. This mechanism can balance the control conflicts between the bending roll force, the roll shifting amount and the rolling force.

[0045] Among them, performing plate shape prediction and correction means feedback correction of the control model based on online measurement data. Specifically, the recursive least squares method can be used to update the influence coefficient matrix to realize this. This technology can compensate for the model deviation caused by roll wear and working condition changes.

[0046] Among them, performing plate shape evaluation and optimization means establishing a comprehensive evaluation system including quality indicators and economic indicators. Specifically, the genetic algorithm can be used for multi-objective optimization solution, and the optimal process parameter combination can be determined through Pareto frontier analysis.

[0047] The core innovation of the present application lies in accurately describing the interaction of physical fields in the rolling process by establishing a multi-dimensional coupled model, realizing local stress concentration control by combining segmented roll shape design, and adapting to the control requirements of different rolling stages by using a dynamic weight distribution mechanism, and finally forming a closed-loop optimization system to improve the plate shape control precision and stability.

[0048] The working process and principle of the present application are as follows: firstly, a multi-dimensional dynamic model is established, and a multi-dimensional physical field coupling model of the rolled piece-roller-cooling is constructed. The model includes a rolled piece physical model, a roller physical model and a cooling system model, and the overall description of the rolling process is realized by setting the model boundary conditions. Among them, the rolled piece physical model describes the metal flow and deformation behavior of the strip, the roller physical model considers the physical field of roller deflection and thermal expansion, and the cooling system model describes the influence of cooling water flow and temperature parameters on the shape of the strip.

[0049] Then, a multi-segment polynomial roll shape structure is designed, and the parameters of each segment of the roll shape curve are adjusted according to the characteristic parameters of the strip. The roll shape curve is divided into an inlet transition zone, a roll pressure zone and an outlet transition zone. The inlet transition zone adopts a circular arc chamfer curve to reduce stress concentration, the roll pressure zone uses a multi-segment polynomial line segment combination structure to realize efficient edge drop control, and the outlet transition zone adopts a straight line chamfer curve to ensure the stability of the strip shape. The continuity of the roll shape curve is ensured by the constraint condition that the first-order derivatives of the end points are equal.

[0050] Then, the dynamic control weight is set, and the control priority of the bending roll, the roll shifting and the rolling force is adjusted in real time according to the rolling stage. The initial control weight includes the weight distribution of the steady-state rolling, the threading stage and the tailing stage. The weight is dynamically adjusted according to the real-time strip shape deviation, energy consumption and roll consumption indicators, and the control parameters are optimized through the collaborative control algorithm to make the objective function optimal.

[0051] When performing strip shape prediction and correction, real-time strip shape measurement data is collected, including the middle wave shape and the edge strip shape parameters. The control amount is updated according to the deviation between the actual strip shape and the target strip shape by using the adaptive correction technology. The causes of the strip shape are analyzed in reverse through the strip shape prediction model, and the control strategy is optimized.

[0052] Finally, the strip shape is evaluated and optimized, the strip shape control effect is calculated, including the strip shape deviation and the edge thinning amount indicators. The energy consumption and roll consumption economic indicators are evaluated, and the non-linear optimization is carried out according to the multi-objective function to seek the Pareto optimal solution. The optimization result is verified through the digital twin platform to ensure the feasibility in actual production.

[0053] As a preferred embodiment, the scheme of the present application is implemented as follows:

[0054] In the multi-roll continuous rolling process of non-oriented silicon steel, a multi-dimensional dynamic model is first established. The rolled piece physical model is constructed by using the finite element method, including the geometric model and the material model of the strip. The geometric model includes the length, width and thickness of the strip, and the material model includes the elastic modulus, Poisson's ratio and density. The roller physical model includes the geometric model and the material model of the roller, wherein the geometric model includes the roller diameter and the working roll face length, and the material model includes the elastic modulus and the thermal expansion coefficient. The cooling system model includes the cooling water flow, the nozzle layout and the cooling efficiency parameters.

[0055] When designing the multi-segment polynomial roll shape structure, the roll shape curve is divided into an entry transition zone, a roll compression zone and an exit transition zone. The entry transition zone adopts a circular arc chamfer curve, the roll compression zone uses a multi-segment polynomial line segment combination structure, and the exit transition zone adopts a straight line chamfer curve. The continuity of the roll shape curve is ensured through the constraint condition that the first-order derivatives of the end points are equal.

[0056] When setting the dynamic control weight, the initial control weight includes the weight distribution of the steady rolling, the threading stage and the tailing stage. The weight is dynamically adjusted according to the real-time plate shape deviation, energy consumption and roll consumption indicators, and the bending roll, roll shifting and rolling force control parameters are optimized through the collaborative control algorithm.

[0057] When performing plate shape prediction and correction, the central wave shape and the edge plate shape parameters are adopted. The control amount is updated according to the deviation between the actual plate shape and the target plate shape by using the adaptive correction technology. The plate shape forming reason is analyzed reversely through the plate shape prediction model, and the control strategy is optimized.

[0058] When performing plate shape evaluation and optimization, the plate shape deviation and the edge thinning amount indicators are calculated, and the energy consumption and roll consumption economic indicators are evaluated. The Pareto optimal solution is sought through the nonlinear optimization of the multi-objective function. The optimization result is verified through the digital twin platform to ensure the feasibility in actual production.

[0059] Through the above scheme, the present application realizes the accurate control of the plate shape in the multi-roll continuous rolling process of non-oriented silicon steel. The establishment of the multi-dimensional dynamic model improves the comprehensive description ability of the rolling process, the design of the multi-segment polynomial roll shape structure improves the edge drop control effect, the setting of the dynamic control weight enhances the adaptability to different rolling stages, the plate shape prediction and correction technology improves the control precision, and the multi-objective optimization method realizes the comprehensive balance of the plate shape quality, energy consumption and roll consumption. The comprehensive application of these technical means effectively solves the high-order local wave shape problem, improves the plate shape control precision, adapts to the rolling demand of the ultra-thin plate, improves the energy consumption and roll consumption indicators, and provides technical support for the high-quality production of non-oriented silicon steel.

[0060] In some of the above schemes of the present application, the coupling relationship between the subsystems is not fully considered when establishing the multi-dimensional dynamic model, which leads to insufficient simulation precision of the roll deflection deformation and the temperature field change of the model, and it is difficult to accurately reflect the interaction mechanism between the rolled piece and the roll during the rolling process, which affects the precision of the subsequent control strategy.

[0061] The present application further proposes a technical scheme including establishing a rolled piece physical model, constructing a roll physical model, establishing a cooling system model and setting a model boundary condition.

[0062] The physical model of the rolled piece adopts a finite element method, and includes a geometric model and a material model. The geometric model covers the length, width and thickness parameters of the strip, and the material model includes the elastic modulus, Poisson's ratio and density parameters. The physical model of the roll includes the geometric parameters of the roll diameter and the working roll face length, and the material parameters of the elastic modulus and the thermal expansion coefficient. The cooling system model includes the cooling water flow, nozzle layout and cooling efficiency parameters. The boundary conditions of the model include the thickness, width and speed parameters of the incoming strip, and the rolling force, roll shifting amount and bending roll angle control parameters. The data interaction of the subsystem models is realized through a parameter coupling interface. The rolled piece deformation data is mapped to the roll model through a stress-strain relationship, and the roll thermal expansion amount is transmitted to the cooling system model through a thermal force coupling equation.

[0063] Specifically, the physical model of the rolled piece is discretized into three-dimensional unit bodies through finite element meshing, and an elastoplastic model is used for the material constitutive relationship. The metal flow distribution in the rolling process is obtained through iterative calculation. The physical model of the roll adopts an axisymmetric modeling method. The working roll face length parameter and the strip width form a proportional relationship of 1.2-1.5 times. The thermal expansion coefficient parameter is set to 11.7x10 -6 / ℃ according to the roll material. The cooling system model simulates the functional relationship between the nozzle spray angle and the cooling water coverage area through a CFD method. The cooling efficiency parameter is set to 0.85-0.92. In the boundary conditions, the incoming strip thickness parameter is set to 0.3-0.5 mm, and the rolling force parameter is updated in real time through a rolling force calculation formula. Through the multi-physical field data interaction mechanism, the rolled piece deformation, roll deflection and cooling water temperature parameters form a closed loop feedback, and the dynamic behavior of the rolling process is accurately simulated.

[0064] As a preferred embodiment, the scheme of the application is implemented as follows: when establishing the physical model of the rolled piece, the finite element software ANSYS is used to establish the strip metal flow and deformation behavior model. Further, when constructing the physical model of the roll, the numerical calculation method is used to establish the physical field model of the roll deflection and thermal expansion. When establishing the cooling system model, the fluid mechanics equation is used to calculate the influence of the cooling water flow and temperature parameters on the strip shape. Specifically, when setting the boundary conditions of the model, the incoming strip state parameters and rolling force given parameters are input. For example, for the incoming strip state, the strip thickness, width and speed parameters are input, and for the rolling force given parameters, the rolling force, roll shifting amount and bending roll angle control parameters are input.

[0065] By the technical scheme, the application realizes accurate description of plate strip metal flow and deformation behavior by establishing a physical model of a rolled piece, realizes accurate calculation of a physical field of roll deflection and thermal expansion by constructing a physical model of a roll, realizes quantitative analysis of the influence of cooling water flow and temperature parameters on plate shape by establishing a cooling system model, and realizes accurate input of plate strip incoming state and given parameters of rolling force by setting boundary conditions of the model. Thus, comprehensive consideration of multi-physical field coupling in the rolling process is realized, and the comprehensiveness and accuracy of plate shape control are improved.

[0066] In some schemes of the application, stress concentration and insufficient edge drop control in the roll shape design process lead to unstable plate strip edge quality and affect the overall plate shape control effect.

[0067] The application further proposes that the roll shape curve is divided into multiple functional regions, including an inlet transition region, a roll pressing region, and an outlet transition region; a circular arc chamfer curve is designed in the inlet transition region to reduce stress concentration; a polynomial multi-segment line segment combination structure is adopted in the roll pressing region to realize efficient edge drop control; a straight line chamfer curve is designed in the outlet transition region to ensure plate shape stability; and continuity of the roll shape curve is ensured through a constraint condition of equal first-order derivatives of end points.

[0068] The radius of the circular arc chamfer curve in the inlet transition region is controlled in a range of 15-30 mm, and a specific value is dynamically adjusted according to the plate strip width; the roll pressing region adopts a curve form of a combination of a cubic polynomial and a quadratic polynomial, and polynomial coefficients are optimized in real time through plate strip edge drop feedback; the angle of the straight line chamfer in the outlet transition region is set to 45-60 degrees, and the chamfer length accounts for 8-12% of the total length of the roll surface; and equal first-order derivatives of end points are realized through establishment of a system of simultaneous equations, and parameter constraint conditions are set at connection positions of the curves.

[0069] Specifically, when the circular arc chamfer curve is adopted in the inlet transition region, the curvature radius forms a matching relationship with the plate strip inlet speed, and when the plate strip enters the rolling region at a speed of 0.8-1.2 m / s, the design can effectively disperse concentrated stress generated by metal flow. In the combined polynomial curve of the roll pressing region, the cubic polynomial segment controls deformation of the middle part of the plate strip, and the quadratic polynomial segment controls edge thinning, and a common tangent point coordinate is set at a connection position of the two curves to control the edge drop rate in a range of 0.5-1.2 μm / mm. The straight line chamfer of the outlet transition region forms geometric cooperation with a subsequent guide device, and when the plate strip thickness is less than 0.3 mm, an angle of 58±2 degrees is selected for the chamfer to eliminate tail warping. Derivative values are calculated at connection positions of the curves through numerical differentiation, and Newton iteration is used to solve the constraint equation set to ensure the second-order continuous characteristics of the roll surface curve.

[0070] As a preferred embodiment, the schemes of the application are implemented as follows:

[0071] The roll shape curve is divided into multiple functional areas, including an entry transition area, a roll pressing area, and an exit transition area. A circular arc chamfer curve is designed in the entry transition area to reduce stress concentration. A polynomial multi-segment line segment combination structure is adopted in the roll pressing area to achieve efficient edge drop control. A straight line chamfer curve is designed in the exit transition area to ensure the stability of the plate shape. The continuity of the roll shape curve is ensured by the constraint condition that the first-order derivatives of the end points are equal.

[0072] Specifically, the circular arc chamfer curve of the entry transition area adopts a quadratic curve equation, and the curve length is 5% of the total length of the roll. The roll pressing area adopts a three-segment polynomial curve combination, and the length of each curve is 30% of the total length of the roll, corresponding to the edge, the middle, and the other edge, respectively. The straight line chamfer curve of the exit transition area has a length of 5% of the total length of the roll. The first-order derivatives of the end points of each curve are ensured to be equal through a numerical optimization method, ensuring the continuity of the overall roll shape curve.

[0073] Further, the polynomial curve of the roll pressing area can adopt a cubic polynomial function, and the coefficients thereof are obtained through least squares fitting. The coefficient adjustment range of the edge curve is relatively large to achieve flexible edge drop control. The coefficient adjustment range of the middle curve is relatively small to maintain the flatness of the middle of the strip.

[0074] Therefore, through the refined roll shape design, the transverse thickness distribution of the strip can be effectively controlled, the stress concentration at the edge is reduced, and the plate shape control precision is improved.

[0075] Through the above technical solutions, the continuity and smoothness of the roll shape curve are achieved, the stress concentration is effectively reduced, and the precision and stability of the plate shape control are improved. The multi-segment polynomial roll shape structure enhances the edge drop control capability and improves the transverse thickness distribution of the strip. The special design of the entry and exit transition areas reduces the stress fluctuation when the strip enters and exits the roll gap, and improves the stability of the rolling process. The overall roll shape design scheme has strong adaptability and can be flexibly adjusted according to the characteristic parameters of different specifications of the strip to meet the needs of various rolling conditions.

[0076] In some of the above schemes of the present application, the dynamic weight distribution mechanism needs to deal with the non-steady state problem of the rolling process, but in actual application, there are problems such as lack of process stage adaptability of weight initialization, dynamic adjustment response lag, and insufficient multi-parameter collaborative optimization, resulting in insufficient real-time performance of plate shape control.

[0077] The present application further proposes a technical scheme including setting an initial control weight, adjusting the control priorities of the bending roll, the roll shifting, and the rolling force in real time according to the rolling stage, and optimizing each control parameter through a collaborative control algorithm to make the target function optimal.

[0078] In the initial control weight distribution, the bending roll force weight coefficient is 0.5, the roll shifting weight coefficient is 0.3, and the rolling force weight coefficient is 0.2 in the steady rolling stage; the roll shifting weight is increased to 0.5 and the rolling force weight is reduced to 0.1 in the threading stage; and the bending roll weight is increased to 0.6 in the tailing stage. The dynamic adjustment process is triggered when the plate shape deviation exceeds the set threshold of 5 μm, and the plate shape data is collected by the online monitoring system every 200 ms. When the edge thinning deviation reaches ± 3 μm, the roll shifting control weight is automatically increased by 0.1. The collaborative control algorithm uses a quadratic programming method with constraints to quantify the plate shape deviation, energy consumption and roll consumption into normalized parameters in the range of 0-1. The optimal weight combination is solved by the Lagrange multiplier method.

[0079] Specifically, the initial weight distribution establishes the baseline parameters of different rolling stages based on historical production data. In the steady stage, the bending roll force is focused on maintaining the flatness of the plate shape. In the threading stage, the roll shifting control is strengthened to cope with the deformation of the strip head. In the tailing stage, the bending roll compensation is enhanced to release the stress of the tail. In the dynamic adjustment process, the middle wave height and edge thinning amount data are obtained in real time by the plate shape detector. When the edge thinning deviation exceeds ± 3 μm, the fuzzy logic module preset in the control system increases the roll shifting weight coefficient by 0.1, while proportionally reducing other weight parameters. The collaborative control algorithm recalculates the objective function after each weight adjustment. The objective function is set to the plate shape deviation weight coefficient 0.6, the energy consumption coefficient 0.3, and the roll consumption coefficient 0.1. The control parameter combination that minimizes the objective function is solved by matrix operation. Finally, the optimized bending roll pressure value, roll shifting displacement and rolling force set value are output to the actuator.

[0080] As a preferred embodiment, the scheme of the application is implemented as follows:

[0081] The initial control weight is set, including the weight distribution of the steady rolling, threading stage and tailing stage. In the steady rolling stage, the bending roll control weight is set to 0.4, the roll shifting control weight is set to 0.3, and the rolling force control weight is set to 0.3. In the threading stage, the bending roll control weight is set to 0.5, the roll shifting control weight is set to 0.3, and the rolling force control weight is set to 0.2. In the tailing stage, the bending roll control weight is set to 0.3, the roll shifting control weight is set to 0.4, and the rolling force control weight is set to 0.3.

[0082] The weight is dynamically adjusted according to the real-time plate shape deviation, energy consumption and roll consumption indicators. The plate shape deviation indicators include middle wave degree and edge wave degree, the energy consumption indicators include rolling power and cooling water consumption, and the roll consumption indicators include roll wear amount. When the plate shape deviation exceeds the set threshold, the bending roll control weight is increased; when the energy consumption indicator exceeds the set threshold, the roll shifting control weight is increased; and when the roll consumption indicator exceeds the set threshold, the rolling force control weight is increased.

[0083] The control parameters are optimized by the cooperative control algorithm, so that the objective function is optimal. The fuzzy neural network algorithm is adopted to minimize the plate shape deviation, minimize the energy consumption and minimize the roll consumption as the objective function, and the bending roll angle, the roll displacement and the rolling force are jointly optimized. Iterative calculation is performed until the objective function converges or the maximum iteration number is reached.

[0084] Through the above technical solutions, the dynamic optimization adjustment of the plate shape control parameters is realized, and the precision and stability of the plate shape control are improved. By adjusting the control weight in real time, the system can quickly respond to various changes in the rolling process and effectively suppress the generation of plate shape defects. The application of the cooperative control algorithm makes the control parameters coordinated with each other, avoids the negative effects caused by single parameter adjustment, and realizes the comprehensive optimization of plate shape, energy consumption and roll consumption.

[0085] In some schemes of the application, when the plate shape is evaluated and optimized, the plate shape deviation, energy consumption and roll consumption indicators cannot be effectively balanced, and the optimization result lacks actual production verification mechanism, which may cause the theoretical optimization scheme to be unmatched with the actual working condition.

[0086] The application further provides a scheme including calculating the plate shape control effect, evaluating the energy consumption and roll consumption economic indicators, performing nonlinear optimization according to the multi-objective function, and verifying the optimization result through the digital twin platform.

[0087] When the plate shape control effect is calculated, the plate shape deviation and edge thinning indicators are included, the transverse thickness distribution and edge thinning degree are quantitatively analyzed, and the plate shape quality evaluation benchmark is established. When the economic indicators are evaluated, the electric energy consumption data and roll wear data in the rolling process are collected to form a quantifiable cost analysis model. When the nonlinear optimization is performed, the objective function is constructed based on the weight distribution of the plate shape deviation, energy consumption and roll consumption, and the gradient descent method or genetic algorithm is used to seek the Pareto optimal solution. When the digital twin platform is verified, the optimized control parameters are input into the virtual rolling model to simulate the plate shape response in the actual production environment.

[0088] Specifically, the strip shape deviation is collected by a laser range finder in real time, and the standard deviation is calculated as an evaluation index. The edge thinning is determined by measuring the thickness difference between the edge 50 mm position and the center thickness. The energy consumption data is collected by an electric power sensor to measure the real-time power of the main motor and auxiliary system. The roll consumption data is obtained based on the roll surface wear detection device. In the multi-objective optimization process, the strip shape deviation is not more than 2IU, the energy consumption is reduced by 15%, and the roll consumption is reduced by 20% as the constraint condition, and the NSGA-II algorithm is used for iterative optimization. The digital twin platform imports the roll stiffness characteristic curve and the roll thermal expansion model to simulate the strip shape change under different working conditions and verify the feasibility of the optimization scheme. For example, when the strip shape deviation is optimized to 1.5IU, the energy consumption is reduced by 12% and the roll consumption is reduced by 18% at the same time, and the stability of the parameter combination under the condition of rolling speed 800m / min is verified by the twin model.

[0089] As a preferred embodiment, the scheme of the application is implemented as follows:

[0090] In the strip shape control process, real-time strip shape measurement data is collected. Specifically, the middle wave shape and edge shape parameters are collected by a strip shape meter installed at the outlet of the rolling mill. The middle wave shape is obtained by measuring the thickness change near the center line of the strip, and the edge shape parameters are determined by measuring the thickness distribution of the edge region on both sides of the strip.

[0091] Further, the control quantity is updated using an adaptive correction technique. Thus, according to the deviation of the actual strip shape and the target strip shape, the control parameters are dynamically adjusted. For example, when the middle wave shape is detected to be out of the preset range, the work roll bending force is increased; when the edge shape is abnormal, the roll shifting amount is adjusted. This adaptive correction process continuously optimizes the control parameters through an iterative algorithm to minimize the strip shape deviation.

[0092] Specifically, the causes of strip shape formation are analyzed by a strip shape prediction model. The strip shape prediction model is established based on the finite element analysis method, considering factors such as rolling force, roll deformation, strip deformation, etc. By inputting the measured strip shape data into the prediction model, the causes of strip shape deviation are deduced in reverse, such as roll wear, uneven cooling, etc. According to the analysis results, the control strategy is optimized, such as adjusting the rolling force distribution, optimizing the cooling water flow, etc., so as to improve the strip shape control precision.

[0093] By the technical scheme, real-time optimization and accurate adjustment of the plate shape control are achieved. Due to the adoption of the adaptive correction technology, the control parameters can be dynamically adjusted according to the actual production situation, effectively overcoming the uncertainty factors in the rolling process. Meanwhile, through the reverse analysis of the plate shape prediction model, the root causes of the plate shape defects are deeply excavated, providing a basis for formulating a targeted control strategy. The closed-loop feedback mechanism significantly improves the accuracy and stability of the plate shape control, effectively reduces the plate shape defects, and improves the product quality. In addition, the method also has strong adaptability and can cope with the rolling of plate strips of different specifications and materials, meeting the increasingly high requirements of high-end manufacturing industry on plate shape precision.

[0094] In some schemes of the application, there is a problem of balancing the plate shape quality index and the economic index in the multi-objective optimization process, and the actual production feasibility of the optimization scheme lacks effective verification means, which may easily lead to a mismatch between the theoretical optimization result and the field working condition.

[0095] The application further proposes a technical scheme including plate shape control effect calculation, economic index evaluation, multi-objective nonlinear optimization, and digital twin verification.

[0096] The plate shape control effect calculation includes quantitative analysis of the plate shape deviation and the edge thinning amount, and accurate measurement is achieved by setting a deviation threshold. The economic index evaluation involves independent accounting of energy consumption and roll consumption, and a cost model under unit output is established. The multi-objective nonlinear optimization uses a weight distribution algorithm to handle multiple constraint conditions, and the optimal solution set is determined through Pareto frontier analysis. The digital twin verification dynamically verifies by constructing a virtual rolling environment and simulating actual production parameters.

[0097] Specifically, first, in the plate shape control effect calculation stage, the plate shape profile data is obtained through an online measurement system, and the root mean square value of the longitudinal plate shape deviation and the percentage of the edge thinning amount are calculated, wherein the plate shape deviation threshold is set to ±5IU, and the allowable range of the edge thinning amount is ±0.8%. Then, in the energy consumption evaluation link, the energy consumption of a single coil of steel is calculated by integrating the power curve of the main motor of the rolling mill, and the roll consumption evaluation is based on the wear model to calculate the roughness change rate of the working roll surface. Subsequently, a comprehensive objective function including the plate shape deviation, the energy consumption per ton of steel, and the roll surface wear rate is established in the multi-objective optimization process, and an improved NSGA-II algorithm is used for iterative solution to generate a Pareto optimal solution set. Finally, the preferred scheme is imported into the digital twin platform, and the temperature fluctuation, rolling force fluctuation, and other working conditions in actual production are simulated through a virtual rolling process to verify the feasibility of the scheme. For example, in a certain embodiment, the optimization scheme reduces the edge thinning amount to 0.6%, while reducing the energy consumption per ton of steel by 12.3%, and the digital twin verification confirms that the scheme can still remain stable under the rolling speed fluctuation ±10% working condition.

[0098] As a preferred embodiment, the scheme of the present application is implemented as follows:

[0099] In the plate shape control process, the plate shape control effect is first calculated. Specifically, the thickness distribution of the plate strip is measured by a shape meter, and the plate shape deviation index is calculated. For example, the I-unit plate shape deviation index can be used, and the calculation formula is: I = (h - h0) / h0 x 105, where h is the actual thickness and h0 is the target thickness. At the same time, the thickness of the edge region of the plate strip is measured, and the edge thinning index is calculated.

[0100] Further, the energy consumption and roll consumption economic indexes are evaluated. The energy consumption index can be calculated by measuring the power consumption of the main motor of the rolling mill, and the roll consumption index is evaluated by measuring the change amount of the roll diameter. For example, the unit output energy consumption index kWh / t and the unit output roll consumption index mm / kt can be used to quantitatively evaluate.

[0101] Thus, according to the multi-objective function, a nonlinear optimization is performed to seek a Pareto optimal solution. The multi-objective function can include the plate shape deviation, the edge thinning, the energy consumption and the roll consumption indexes, and a weighted summation method is used to construct the objective function. Through an optimization method such as a genetic algorithm or a particle swarm algorithm, the optimal control parameter combination is found.

[0102] Finally, the optimization result is verified through a digital twin platform. The control parameters obtained by optimization are input into the digital twin model to simulate the rolling process and predict the plate shape control effect. If the simulation result meets the requirements, the optimized parameters are applied to the actual production; if the simulation result is not ideal, the optimization step is returned, the optimization algorithm or the objective function weight is adjusted, and the optimization is re-performed.

[0103] Through the above technical scheme, the present application realizes multi-objective comprehensive optimization of plate shape control effect, energy consumption and roll consumption. By calculating the plate shape deviation and edge thinning indexes, the plate shape control effect can be accurately evaluated. Combined with the evaluation of energy consumption and roll consumption economic indexes, the production efficiency and cost factors are comprehensively considered. The nonlinear optimization algorithm is used to seek a Pareto optimal solution, which can find the best balance point among multiple objectives. The verification link of the digital twin platform ensures the feasibility and effectiveness of the optimization result in actual production. This method of combining multi-objective optimization and digital twin verification improves the precision and stability of plate shape control, while taking into account the production efficiency and economy, and provides a high-efficiency and reliable plate shape control strategy for the multi-roll continuous rolling process of non-oriented silicon steel.

[0104] In some of the above schemes of the present application, when establishing the physical model of the rolled piece, the traditional method has insufficient accuracy in representing the geometric characteristics and material properties of the plate strip, resulting in deviations between the simulation results of metal flow and deformation behavior and the actual production conditions, affecting the accuracy of the subsequent multi-physical field coupling model.

[0105] The application further proposes to establish a physical model of the strip according to the strip specification parameters by using the finite element method, including a geometric model and a material model of the strip. The geometric model includes the geometric dimensions of the length, width and thickness of the strip, and the material model includes the material parameters of the elastic modulus, Poisson's ratio and density.

[0106] The application of the finite element method realizes accurate solution of the stress and strain field by discretizing the strip into a unit grid. When the geometric model is constructed, the spatial dimensions of the three-dimensional solid model are set according to the measured length, width and thickness values of the incoming strip. In the material model setting, the elastic modulus uses the measured value of the silicon steel material at the rolling temperature, the Poisson's ratio is determined according to the anisotropic experimental data of the material, and the density parameter is obtained according to the material composition. For example, for a strip with a thickness of 0.3 mm, the grid size in the length direction of the geometric model is set to 5 mm, and the width direction is divided into fine grids with a size of 0.1 mm.

[0107] Specifically, the establishment process of the finite element model includes importing the geometric profile data of the strip and generating a parameterized three-dimensional solid. In the material property definition stage, the elastic modulus curve of the silicon steel material under different temperature conditions is input, and the constitutive equation of the anisotropic material is set. By constraining the displacement boundary conditions in the rolling direction, the plastic flow of the metal under the action of the roller is simulated. Among them, the accurate setting of the Poisson's ratio can effectively reflect the coupling relationship between the transverse contraction and the longitudinal extension of the strip, and the dynamic adjustment of the elastic modulus can represent the material hardening effect in the rolling process. The model obtains the stress distribution cloud diagram of each node of the strip through iterative calculation, provides deformation behavior prediction data for subsequent roll shape optimization, and thus improves the simulation accuracy of the physical model of the rolled piece to the actual rolling process.

[0108] As a preferred embodiment, the scheme of the application is implemented as follows:

[0109] A physical model of the strip is established according to the strip specification parameters by using the finite element method. The physical model includes a geometric model and a material model of the strip. The geometric model includes the geometric dimensions of the length, width and thickness of the strip. Specifically, the length of the strip is set to 10000 mm, the width is set to 1200 mm, and the thickness is set to 0.5 mm. The material model includes the material parameters of the elastic modulus, Poisson's ratio and density. Further, the elastic modulus is set to 210 GPa, the Poisson's ratio is set to 0.3, and the density is set to 7850 kg / m 3 . Thus, by establishing the physical model of the strip, the deformation behavior of the strip in the rolling process can be accurately described.

[0110] By the technical scheme, the physical characteristics of the strip can be accurately described, and accurate basic data can be provided for subsequent shape control. Since the finite element method is used to establish the physical model, the complex deformation behavior of the strip in the rolling process can be more comprehensively considered, thereby improving the accuracy and stability of the shape control. In addition, by setting specific geometric dimensions and material parameters, the model is closer to the actual production situation, which is beneficial to improve the practicability and reliability of the shape control.

[0111] In some schemes of the application, the geometric parameters and material characteristics of the roll are not accurately considered in the roll modeling process, resulting in insufficient model accuracy, which cannot effectively reflect the deflection deformation and thermal expansion effect of the roll in the rolling process, thereby affecting the accuracy of the shape control.

[0112] The application further establishes a geometric model and a material model of the roll. The geometric model includes roll diameter and work roll face length parameters, and the material model includes elastic modulus and thermal expansion coefficient parameters.

[0113] The roll diameter parameter is obtained by measuring the roll system configuration data of the actual rolling mill, and its value range is 400-800 mm, which is used to calculate the rigidity coefficient of the roll. The work roll face length parameter is designed according to the matching of the strip steel width specification, and is usually 1.1-1.3 times of the maximum width of the strip steel, so as to ensure the effective coverage of the strip steel edge in the rolling process. The elastic modulus parameter adopts the material parameter of 210-220 GPa of high-strength alloy steel, which accurately represents the rigidity characteristics of the roll. The thermal expansion coefficient parameter selects a value range of 12-14*10^-6 / ℃, which reflects the radial expansion law of the roll under temperature rise.

[0114] Specifically, the roll diameter parameter directly affects the calculation accuracy of the bending rigidity of the roll, and the work roll face length parameter is used to determine the projection area of the roll and the strip. The elastic modulus parameter is used to construct the stress-strain relationship of the roll under stress, and the deflection deformation of the roll is calculated by finite element simulation. The thermal expansion coefficient parameter is combined with the temperature field distribution in the rolling process to predict the change of the thermal expansion amount of the work surface of the roll. In the geometric model of the roll, the diameter parameter and the work roll face length parameter form a two-dimensional geometric constraint condition, and a three-dimensional modeling software is used to construct a roll entity model. In the material model, the elastic modulus parameter and the thermal expansion coefficient parameter are input as boundary conditions for solving the physical field, and are coupled with the thermal-mechanical multi-field analysis module. When the rolling pressure acts on the roll, the model calculates the elastic deformation of the roll based on the elastic modulus parameter, and at the same time, the geometric size of the roll is corrected according to the real-time temperature data through the thermal expansion coefficient parameter, so as to realize the comprehensive simulation of the deflection deformation and thermal expansion of the roll.

[0115] As a preferred embodiment, the scheme of the application is implemented as follows: in the construction process of the roll physical model, the CAD model data of the roll is imported through a three-dimensional modeling software, and the diameter of the work roll is set to 650 mm and the length of the work roll surface is set to 2080 mm in the geometric model. The material model is defined by using an elastic-plastic constitutive equation, wherein the elastic modulus is set to 206 GPa and the thermal expansion coefficient is set to 12*10^-6 / ℃. After the model is constructed, the geometric model and the material model are coupled through a finite element analysis software to establish a stress-strain analysis model of the roll. In actual application, the model is connected with a rolling process monitoring system through a data interface to collect rolling pressure distribution data in real time for model verification.

[0116] Through the above technical scheme, the application can accurately simulate the mechanical deformation and thermal deformation behavior of the roll in the rolling process, effectively improve the matching degree of the roll physical model and the actual working condition, provide a model basis for solving the problems of large prediction deviation of roll deflection deformation and inaccurate thermal expansion compensation, and thus establish a reliable numerical analysis platform for subsequent optimization of the plate shape control strategy.

[0117] In some schemes of the application described above, the physical fields of roll deflection and thermal expansion need to be considered when constructing the roll physical model. However, in the actual modeling process, there is a problem that the geometric parameters and material characteristic parameters of the roll are incomplete, which leads to the fact that the model cannot accurately reflect the actual working state of the roll and affects the plate shape control accuracy.

[0118] The application further proposes a technical scheme of establishing a geometric model and a material model in the roll physical model, wherein the geometric model contains roll diameter and work roll surface length parameters, and the material model includes elastic modulus and thermal expansion coefficient parameters.

[0119] The roll diameter parameter is used to determine the contact arc length and deformation resistance distribution law of the roll, and the work roll surface length parameter and the strip width parameter jointly determine the rolling contact area range. The elastic modulus parameter is used to calculate the elastic deformation amount of the roll under the action of the rolling force, and the thermal expansion coefficient parameter is used to simulate the size fluctuation of the roll diameter caused by temperature change. The geometric model and the material model are coupled and calculated through finite element grid nodes, the work roll surface length parameter is set to maintain a proportional relationship of 1.05-1.15 times with the opening degree of the mill stand, and the roll diameter parameter is dynamically matched according to the calculated value of the rolling force.

[0120] Specifically, the roll geometry model is constructed by a three-dimensional solid modeling method, the roll diameter parameter is input to generate a basic cylinder, and the effective working area is determined in combination with the working roll surface length parameter. The material model adopts the isotropic assumption, the elastic modulus parameter is used to construct the stiffness matrix, and the thermal expansion coefficient parameter is introduced into the temperature field-stress field coupling equation. During the rolling process, when the roll temperature rises, the thermal expansion coefficient parameter is used to calculate the thermal expansion increment of the roll diameter, and the increment value is fed back to the rolling force calculation module for compensation and correction. For example, when the working roll surface length is 1500 mm and the roll diameter is 600 mm, the combination of the elastic modulus 210 GPa and the thermal expansion coefficient 12*10^-6 / ℃ can improve the model prediction accuracy by 18.6%. By accurately matching the roll geometry parameters and material parameters, the model can accurately predict the roll deflection deformation, providing reliable data support for subsequent shape control.

[0121] As a preferred embodiment, the scheme of the application is implemented as follows: in the modeling process of the cooling system of the cold rolling mill, first, the rated flow parameter of the cooling water circulation system is obtained to determine the cooling water volume flow range through the nozzle per unit time. According to the number of nozzles distributed along the roll axis, a spatial layout model of the nozzles is established using a three-dimensional coordinate system, wherein the nozzle spacing is arranged equidistantly along the roll axis direction, and the jet angle forms a preset inclination angle with the roll surface. Further, in combination with the heat conduction equation of the cooling water and the roll surface, the cooling efficiency parameter is quantified as a function relationship between the cooling water flow and the roll temperature change rate. The accuracy of the model is verified by CFD fluid dynamics simulation, and the cooling efficiency distribution coefficient of the nozzle coverage area is dynamically corrected according to the measured temperature data.

[0122] Through the above technical scheme, the application realizes accurate modeling of the cooling system action mechanism, and can accurately quantify the correlation between the cooling water flow distribution, nozzle layout form and plate strip thermal deformation. By establishing a physical model containing dynamic parameters of cooling efficiency, the problem of mismatching between cooling intensity and plate shape response in the traditional method is effectively solved, and the local stress concentration and plate shape warping phenomenon caused by uneven cooling are avoided, which significantly improves the dynamic adjustment accuracy of the plate shape control system.

[0123] In some schemes of the application described above, there is a problem of inaccurate boundary condition setting in the multi-dimensional dynamic model construction process, especially in the non-steady state stage of rolling, the existing method fails to fully consider the coupling relationship between the incoming material parameters and the dynamic control parameters, resulting in that the model cannot accurately reflect the complex working conditions in actual production.

[0124] The application further proposes that when setting the boundary conditions of the model, the plate strip incoming thickness, width and speed parameters, and the rolling force, roll shifting amount and bending roll angle control parameters are included.

[0125] The incoming thickness is determined by the initial thickness measurement of the strip, which is used as the initial input for the material flow calculation; the incoming width is obtained in real time by the edge detection device, which is used to determine the roll contact area; and the incoming speed is collected by the speed sensor, which is used as the reference for the dynamic parameter update of the rolling process. The rolling force is set according to the pressure feedback value of the rolling mill hydraulic system, the roll shifting amount is monitored by the displacement sensor to monitor the axial position of the work roll, and the bending angle is obtained in real time by the angle encoder to obtain the bending angle of the roll system. These parameters are transmitted to the model calculation unit through the data bus to form a closed-loop control system. For example, when the thickness fluctuation exceeds the set threshold, the rolling force parameter is automatically adjusted by 1.5-2.0 MN per millimeter of thickness difference.

[0126] Specifically, the incoming thickness parameter of the strip directly affects the initial set value of the rolling force calculation module. When the incoming thickness deviation is detected, the rolling force control parameter is adjusted through the feedforward compensation algorithm to ensure the force balance of the roll. The strip width parameter and the roll shifting amount form a linkage mechanism. When the strip edge position deviates, the roll shifting amount is dynamically compensated according to the width deviation value at a rate of 0.8-1.2 mm / mm. The incoming speed parameter and the bending angle control form a time coupling. During the speed improvement stage, the bending angle is pre-adjusted at a rate of 0.05° / m / min to avoid the lag of the roll gap shape. The control parameters of rolling force, roll shifting amount, and bending angle are determined by the orthogonal test method to determine the interaction weight coefficient. Under the working conditions of thickness fluctuation ±0.1 mm and width deviation ±5 mm, the plate shape deviation can be controlled within 3 IU. This parameter setting method effectively solves the prediction error problem of the traditional model in the non-steady state rolling stage by establishing a mathematical mapping relationship between the incoming state and the control parameter.

[0127] As a preferred embodiment, the scheme of the present application is implemented as follows: In the process parameter setting stage of the cold rolling production line, according to the rolling specification of 0.3 mm thickness and 1200 mm width of non-oriented silicon steel strip, the boundary conditions of the rolling model are set. The incoming parameters of the strip are collected in real time by the entrance thickness gauge, and the initial thickness is set to 2.5 mm and the width is set to 1250 mm. The inlet speed is kept at 500 m / min. The rolling force parameter is set to 12000 kN according to the rolling mill stiffness curve, the roll shifting amount is adjusted to ±50 mm stroke range through the roll position sensor feedback, and the bending angle is set to 0.8°-1.2° dynamic adjustment interval according to the rolling pressure distribution model. The control parameters are transmitted to the rolling mill actuator through the distributed control system, and a closed-loop control loop is formed with the plate shape detection system.

[0128] Through the technical scheme, the application realizes accurate boundary condition setting of a rolling process physical model, and solves the problem of plate shape prediction deviation caused by the traditional model ignoring dynamic changes of actual production parameters. Through real-time acquisition of incoming material state data and linkage with rolling control parameters, consistency of the model calculation result and actual production conditions is ensured, and the self-adaptability of the plate shape control system is effectively improved. Meanwhile, the collaborative setting of rolling force, roll shifting amount and bending roll angle parameters provides an accurate input benchmark for a multi-physical field coupling model, and significantly improves the stability and reliability of the plate shape control strategy.

[0129] The above merely describes the embodiments of the application and is not used to limit the protection scope of the application. For those skilled in the art, the application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. A process for shape control of a non-oriented silicon steel in a multi-roll continuous rolling, characterized in that, The application relates to a method for optimizing the plate shape control of a strip mill, which comprises the following steps: step 1, establishing a multi-dimensional dynamic model, and constructing a physical field coupling model of a rolled piece-roller-cooling system; step 2, designing a multi-segment polynomial roll shape structure, and adjusting the parameters of each segment of the roll shape curve according to the characteristic parameters of the strip; step 3, setting a dynamic control weight, and adjusting the control priorities of the roll bending, roll shifting and rolling force in real time according to the rolling stage; step 4, performing plate shape prediction and correction, and updating the control quantity influence coefficient by using an adaptive correction technology; step 5, performing plate shape evaluation and optimization, and performing multi-objective optimization according to the target functions of the plate shape deviation, energy consumption and roll consumption. The step 1 comprises the following steps: step 101, establishing a rolled piece physical model, and describing the metal flow and deformation behavior of the strip; step 102, constructing a roller physical model, and considering the physical field of roller deflection and thermal expansion; step 103, establishing a cooling system model, and describing the influence of the cooling water flow and temperature parameters on the plate shape; and step 104, setting the model boundary conditions, including the incoming state of the strip and the given parameters of the rolling force. The step 2 comprises the following steps: step 201, dividing the roll shape curve into multiple functional areas, including an entrance transition area, a roll pressing area and an exit transition area; step 202, designing a circular arc chamfer curve in the entrance transition area to reduce stress concentration; step 203, adopting a multi-segment polynomial line segment combination structure in the roll pressing area to realize efficient edge drop control; step 204, designing a straight line chamfer curve in the exit transition area to ensure the stability of the plate shape; and step 205, ensuring the continuity of the roll shape curve through the constraint condition that the first-order derivatives of the end points are equal. The step 3 comprises the following steps: step 301, setting an initial control weight, including the weight distribution of the steady-state rolling, the strip threading stage and the tailing stage; step 302, dynamically adjusting the weight according to the real-time plate shape deviation, energy consumption and roll consumption indexes; and step 303, optimizing the control parameters through a collaborative control algorithm to make the target function optimal. The step 4 comprises the following steps: step 401, collecting real-time plate shape measurement data, including the middle wave shape and the edge plate shape parameters; step 402, updating the control quantity according to the deviation between the actual plate shape and the target plate shape by using an adaptive correction technology; and step 403, analyzing the reasons for the plate shape formation through a plate shape prediction model to optimize the control strategy. The step 5 comprises the following steps: step 501, calculating the plate shape control effect, including the plate shape deviation and the edge thinning amount index; step 502, evaluating the energy consumption and roll consumption economic indexes; step 503, performing nonlinear optimization according to the multi-objective function to seek a Pareto optimal solution; and step 504, verifying the optimization result through a digital twin platform to ensure the feasibility in actual production.

2. The process for shape control of a non-oriented silicon steel in a multi-roll continuous rolling according to claim 1, characterized in that, The step 101 comprises the following steps: according to the strip specification parameters, a physical model of the strip is established by using a finite element method, including a geometric model and a material model of the strip, wherein the geometric model contains the geometric dimensions of the length, width and thickness of the strip, and the material model includes the material parameters of the elastic modulus, Poisson's ratio and density. The step 102 comprises the following steps: according to the roller specification parameters, a geometric model and a material model of the roller are established, wherein the geometric model contains the roller diameter and the working roll surface length parameters, and the material model includes the parameters of the elastic modulus and the thermal expansion coefficient. ​ ​ ​ 3. The process for shape control of non-oriented silicon steel in multi-roll continuous rolling according to claim 1, characterized in that, ​ ​ ​ ​ ​ ​ 4. The process for shape control of non-oriented silicon steel in multi-roll continuous rolling according to claim 1, characterized in that, ​ ​ ​ ​ 5. The process for shape control of non-oriented silicon steel in multi-roll continuous rolling according to claim 1, characterized in that, ​ ​ ​ ​ 6. The process for shape control of non-oriented silicon steel in multi-roll continuous rolling according to claim 1, characterized in that, ​ ​ ​ ​ ​ 7. The process for shape control of non-oriented silicon steel in multi-roll continuous rolling according to claim 2, characterized in that, ​ ​ 8. The process for shape control of non-oriented silicon steel in multi-roll continuous rolling according to claim 2, characterized in that, ​ ​ 9. The process for shape control of non-oriented silicon steel in multi-roll continuous rolling according to claim 2, characterized in that, The step 103 includes: According to the specification parameters of the cooling system, a physical model of the cooling system is established, including cooling water flow, nozzle layout, and cooling efficiency parameters.

10. The process for shape control of a non-oriented silicon steel sheet in a multi-roll continuous rolling according to claim 1, characterized in that, The step 104 includes: According to the production process requirements, the boundary conditions of the model are set, including the plate strip incoming thickness, width, speed parameters, and rolling force, roll shifting amount, and bending roll angle control parameters.

Images