A method, medium, and equipment for predicting rolling force in a finishing mill.
By establishing a physical model based on equivalent deformation resistance and equivalent contact arc length, and using the differential evolution algorithm to optimize the unknown coefficients, the deviation problem in the prediction of rolling force of the finishing mill was solved, realizing high-precision, simple and efficient rolling force prediction, which is suitable for setting the rolling force and controlling the constant elongation of the finishing mill.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- WISDRI ENG & RES INC LTD
- Filing Date
- 2026-01-28
- Publication Date
- 2026-06-02
AI Technical Summary
Existing prediction methods for rolling force in finished mills suffer from problems such as calculation bias in physical models and weak generalization ability and poor interpretability of pure data models, resulting in low prediction accuracy and difficulty in meeting the real-time and reliability requirements of industrial control.
A physical model based on equivalent deformation resistance and equivalent contact arc length is established, and the unknown coefficients are optimized by combining differential evolution algorithm. By collecting data at the rolling site and performing self-learning, a rolling force prediction method with clear physical meaning is constructed.
It achieves high-precision, simple and efficient rolling force prediction, has strong generalization ability, can significantly improve strip steel performance and reduce performance deviation, and meets the real-time and interpretability requirements of industrial control.
Smart Images
Figure CN122133441A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of finishing mill rolling technology, and in particular to a method, medium and equipment for predicting rolling force in finishing mills. Background Technology
[0002] After continuous annealing and surface galvanizing, the surface quality, shape, and microstructure of cold-rolled strip steel are difficult to meet standards, and the material remains in a completely soft state. The presence of a yield plateau severely restricts the quality of subsequent processing. Therefore, a light rolling process with an elongation of 0.5%–3% is required after annealing and galvanizing, i.e., a finishing process, to eliminate the yield plateau and minor waviness. The core functions of this process include: ① eliminating the yield plateau; ② controlling surface roughness and improving surface quality; ③ optimizing the shape; and ④ improving the material's mechanical properties and stamping formability.
[0003] The finishing mill uses a rolling force control mode to achieve constant elongation control. As a key process parameter, the accuracy of the rolling force setting directly determines the control effect: accurate setting can quickly and stably achieve the target elongation, while setting deviation will lead to prolonged adjustment time, increased fluctuation amplitude, and cause process instability and excessively long abnormal quality sections.
[0004] Field practice shows that regardless of whether the classic Stone formula or Roberts formula applicable to leveling rolling is used, or the Bland-Ford-Hill model applicable to conventional cold rolling, there is a systematic underestimation of rolling force in the hot-dip galvanizing finishing process. For interstitial steel (IF steel) with low yield strength and low elongation, the model calculation deviation is particularly significant, deviating severely from measured values. If deformation resistance and friction coefficient are calculated through self-learning, they often exhibit contradictory deviations—either the deformation resistance is higher than the experimentally measured value, or the friction coefficient exceeds the physically reasonable range, resulting in poor model interpretability and difficulty in guiding production.
[0005] While current mainstream research attempts to use big data-driven pure data models (such as neural networks and random forests) to predict the rolling force of strip mills, the practical application of these models still faces fundamental limitations. First, their generalization ability is weak: when rolling conditions exceed the coverage of the training data (such as extreme thicknesses, new steel grades, or fluctuations in process parameters), the model's prediction accuracy drops sharply, resulting in insufficient robustness. Second, in control applications, it is necessary to accurately calculate the partial derivatives of the rolling force with respect to key variables such as strip deformation resistance, friction coefficient, and entry thickness to optimize control strategies. However, the partial derivatives given by the data model often deviate significantly from the physical laws of metal plastic deformation, and may even contain sign errors, leading to directional deviations in control. More seriously, as "black box" systems, these models lack interpretability in their internal decision-making logic. Once a prediction error occurs, field engineers find it difficult to diagnose and correct it, failing to meet the stringent requirements of traceability and security in industrial control. Therefore, although their offline prediction accuracy is acceptable, they fall short in terms of real-time performance, stability, and reliability, and cannot be deployed in the field at this stage. Summary of the Invention
[0006] The purpose of this invention is to provide a method, medium, and equipment for predicting the rolling force of a smooth mill, aiming to overcome the dual bottlenecks of calculation errors based on physical models and the weak generalization ability and poor interpretability of pure data models in existing smooth mill rolling force prediction methods. This invention provides a simple and efficient calculation method with high prediction accuracy, strong generalization ability, and good interpretability for predicting the rolling force of a smooth mill. The specific technical solution is as follows:
[0007] A method for predicting the rolling force of a finishing mill, the method comprising the following steps:
[0008] S100. A physical model between rolling force and rolling process parameters is established based on equivalent deformation resistance and equivalent contact arc length.
[0009] S200: Collect actual rolling process parameters and actual rolling force of multiple coils at the rolling site;
[0010] S300. Based on the physical model, the calculated rolling force of the multiple coils is obtained according to the actual rolling process parameters of the multiple coils, and the proportional deviation between the calculated rolling force of the multiple coils and the actual rolling force of the multiple coils is used as the objective function.
[0011] S400. Determine the unknown coefficients in the physical model based on the actual rolling process parameters and objective function of the multi-coil using the differential evolution algorithm;
[0012] S500. Substitute the unknown coefficients determined by the solution into the physical model for prediction of the rolling force of the finishing mill.
[0013] Further, step S100 includes:
[0014] S110. Obtain rolling process parameters and unit design parameters;
[0015] S120. Calculate the equivalent contact arc length and equivalent deformation resistance based on the rolling process parameters and unit design parameters.
[0016] S130. Calculate the rolling force based on the equivalent contact arc length and equivalent deformation resistance.
[0017] Furthermore, the rolling process parameters include strip entry thickness, strip width, strip elongation, entry tension, exit tension, average diameter of work rolls, rolling speed, strip yield strength, and coefficient of friction between strip and work rolls; the unit design parameters include maximum rolling speed, maximum diameter of work rolls, and maximum strip elongation.
[0018] The formula for calculating the equivalent contact arc length is as follows:
[0019]
[0020] in, For the equivalent contact arc length, These are the coefficients of the first model. For strip elongation, This represents the maximum elongation of the strip. The average diameter of the working rolls. For reduction rate, The coefficient of friction between the strip and the work roll is denoted as . The strip entry thickness is given; the reduction rate is calculated using the following formula: ;
[0021] The formula for calculating equivalent deformation resistance is as follows:
[0022]
[0023] in, For equivalent deformation resistance, The yield strength of the strip steel. These are the coefficients of the second model. For rolling speed, For maximum rolling speed, The maximum diameter of the working rolls. These are the coefficients of the third model. For the inlet tensile stress, These are the coefficients of the fourth model. The outlet tensile stress is given by the formulas below; the inlet and outlet tensile stresses are calculated as follows:
[0024]
[0025] in, For inlet tension, For strip width, For export tension, The thickness is the export thickness; the formula for calculating the export thickness is as follows: ;
[0026] The formula for calculating rolling force is as follows:
[0027]
[0028] in, This refers to the rolling force.
[0029] Furthermore, considering the influence of the bending roll force on the rolling force, the rolling force is corrected using the following formula:
[0030]
[0031] in, The corrected rolling force, The coefficients of the fifth model, For the bending roller force, These are the coefficients of the sixth model.
[0032] Furthermore, considering the influence of thickness on the rolling force, the rolling force is corrected using the following formula:
[0033]
[0034] in, The corrected rolling force, These are the coefficients of the seventh model. These are the coefficients of the eighth model. =1mm, which is used to eliminate units.
[0035] Furthermore, the rolling force is corrected by simultaneously considering the effects of bending roll force and thickness on the rolling force, as shown in the following formula:
[0036]
[0037] in, The corrected rolling force, The coefficients of the fifth model, For the bending roller force, The coefficients of the sixth model, These are the coefficients of the seventh model. These are the coefficients of the eighth model. =1mm, which is used to eliminate units.
[0038] Furthermore, in step S300, the absolute value of the ratio deviation between the calculated rolling force and the actual rolling force for each coil is taken, and the data of the 10% portion with the largest absolute value of the ratio deviation is removed. The average value of the remaining absolute values of the ratio deviation is taken as the objective function.
[0039] Furthermore, S400 includes the following steps:
[0040] S410. Initialization: Set the population size, mutation factor, crossover probability, and maximum number of iterations, and randomly generate an initial unknown coefficient vector population.
[0041] S420, Fitness Assessment: Substitute the unknown coefficient vector into the physical model to calculate the predicted rolling force corresponding to the actual rolling process parameters of each coil, and calculate the fitness value of each individual according to the objective function.
[0042] S430, Mutation Operation: Perform a mutation operation on each individual in the current population to generate a mutation vector;
[0043] S440, Crossover operation: Cross the mutation vector with the target individual to generate an experimental vector;
[0044] S450, Boundary Processing: Perform boundary reflection or reset processing on coefficients in the test vector that exceed the preset value range;
[0045] S460, Selection Operation: Compare the fitness values of the experimental vectors and the target individuals, and select the better ones to enter the next generation of the population;
[0046] S470, Iteration termination judgment: Repeat steps S420-S460 until the maximum number of iterations is reached or the fitness value reaches the target fitness value;
[0047] S480, Optimal Coefficient Output: The individual with the best fitness value in the final population is used as the unknown coefficient vector of the physical model determined by the solution.
[0048] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the above-described method for predicting the rolling force of a light-finishing mill.
[0049] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the above-described method for predicting the rolling force of a light-finishing mill.
[0050] The present invention provides a method, medium, and equipment for predicting rolling force in a finishing mill, which has the following beneficial effects:
[0051] This invention establishes a physical model between rolling force and rolling process parameters based on equivalent deformation resistance and equivalent contact arc length. Actual rolling process parameters and actual rolling forces of multiple coils are collected at the rolling mill. The calculated rolling force of the multiple coils is obtained based on the physical model according to the actual rolling process parameters, and the proportional deviation between the calculated rolling force and the actual rolling force of the multiple coils is used as the objective function. The unknown coefficients in the physical model are determined by solving the differential evolution algorithm based on the actual rolling process parameters and the objective function. The determined unknown coefficients are substituted into the physical model for predicting the rolling force of the finishing mill. Thus, based on the metal plastic deformation mechanism, a mathematical and physical model of the rolling force of the finishing mill with clear physical meaning is constructed. This model is simple to calculate, highly accurate, has strong generalization ability, and good interpretability. When used for setting the rolling force and controlling the constant elongation rate of the finishing mill, it can significantly improve the performance of strip steel and reduce the occurrence of strip steel performance deviations. Attached Figure Description
[0052] Figure 1 A flowchart illustrating a method for predicting rolling force in a finishing mill, provided in an embodiment of the present invention;
[0053] Figure 2 This is a table of actual rolling process parameters and actual rolling force data collected on-site from multiple coils in the verification embodiments of the present invention;
[0054] Figure 3 A comparison chart of the predicted deviations of rolling force for various schemes;
[0055] Figure 4 This is a graph showing the change in self-learning fitness in the differential evolution algorithm.
[0056] Figure 5 This is a structural block diagram of a computer device according to an embodiment of the present invention. Detailed Implementation
[0057] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. The advantages and features of the present invention will become clearer from the following description. It should be noted that the drawings are all in a very simplified form and use non-precise proportions, and are only used to facilitate and clearly illustrate the purpose of the embodiments of the present invention.
[0058] Example 1
[0059] This embodiment provides a method for predicting the rolling force of a finishing mill. (See reference...) Figure 1 As shown, the method includes the following steps:
[0060] S100. A physical model is established between rolling force and rolling process parameters based on equivalent deformation resistance and equivalent contact arc length.
[0061] In one embodiment, step S100 includes:
[0062] S110. Obtain rolling process parameters and unit design parameters;
[0063] Specifically, the rolling process parameters include strip entry thickness, strip width, strip elongation, entry tension, exit tension, average diameter of work rolls, rolling speed, strip yield strength, and friction coefficient between strip and work rolls; the unit design parameters include maximum rolling speed, maximum diameter of work rolls, and maximum strip elongation.
[0064] The yield strength is related to the strip steel grade, which can be obtained by testing and inspection of finished steel coils of the same grade. The friction coefficient is related to the concentration of finishing solution and the roughness of the work roll. Generally speaking, the concentration of finishing solution and the roughness of the work roll are usually fixed for the same unit, so the friction coefficient is fixed. The friction coefficient can be determined by looking up a table in the database.
[0065] S120. Calculate the equivalent contact arc length and equivalent deformation resistance based on the rolling process parameters and unit design parameters.
[0066] Specifically, the formula for calculating the equivalent contact arc length is as follows:
[0067]
[0068] in, For the equivalent contact arc length, These are the coefficients of the first model. For strip elongation, This represents the maximum elongation of the strip. The average diameter of the working rolls. For reduction rate, The coefficient of friction between the strip and the work roll is denoted as . The strip entry thickness is given; the reduction rate is calculated using the following formula: .
[0069] The formula for calculating equivalent deformation resistance is as follows:
[0070]
[0071] in, For equivalent deformation resistance, The yield strength of the strip steel. These are the coefficients of the second model. For rolling speed, For maximum rolling speed, The maximum diameter of the working rolls. These are the coefficients of the third model. For the inlet tensile stress, These are the coefficients of the fourth model. The outlet tensile stress is given by the formulas below; the inlet and outlet tensile stresses are calculated as follows:
[0072]
[0073] in, For inlet tension, For strip width, For export tension, The thickness is the export thickness; the formula for calculating the export thickness is as follows: .
[0074] S130. Calculate the rolling force based on the equivalent contact arc length and equivalent deformation resistance.
[0075] Specifically, the formula for calculating rolling force is as follows:
[0076]
[0077] in, This refers to the rolling force.
[0078] In one embodiment, the rolling force is corrected to account for the effect of the bending roll force on the rolling force, as shown in the following formula:
[0079]
[0080] in, The corrected rolling force, The coefficients of the fifth model, For the bending roller force, These are the coefficients of the sixth model.
[0081] In one embodiment, the rolling force is corrected to account for the effect of thickness on the rolling force, as shown in the following formula:
[0082]
[0083] in, The corrected rolling force, These are the coefficients of the seventh model. These are the coefficients of the eighth model. =1mm, which is used to eliminate units.
[0084] In one embodiment, the rolling force is corrected by considering the effects of bending roll force and thickness on the rolling force, as shown in the following formula:
[0085]
[0086] in, The corrected rolling force, The coefficients of the fifth model, For the bending roller force, The coefficients of the sixth model, These are the coefficients of the seventh model. These are the coefficients of the eighth model. =1mm, which is used to eliminate units.
[0087] Based on the aforementioned mathematical and physical model of the rolling force in a polished mill, values are assigned to the relevant coefficients to reduce the number of coefficients and achieve dimensionality reduction. Methods for determining these values include theoretical methods, empirical methods, and experimental methods. For example: , It is a working condition factor, which cannot be determined. K4 and K2 are both taken as 0.5 based on theory and experience. If the effect of the bending roller is not considered, , All values are 0; considering the bending roller, , This is also a working condition factor, which cannot be determined. If the effect of thickness is not considered... , All values are 0; considering thickness, , These are also operating condition coefficients, which cannot be determined. Furthermore, rolling process parameters, including strip entry thickness, strip width, strip elongation, entry tension, exit tension, average diameter of the work rolls, rolling speed, and bending force, can be measured and are considered known. Unit design parameters, including maximum rolling speed, maximum work roll diameter, and maximum strip elongation, are all design parameters and can be considered constants. In this case, the number of unknown coefficients becomes 2 to 6.
[0088] S200: Collect actual rolling process parameters and actual rolling force of multiple coils at the rolling site.
[0089] In one embodiment, actual rolling data from multiple coils is collected at the rolling mill, with collection intervals ranging from 500ms to 1000ms. This data includes all rolling process parameters required for the calculations in the aforementioned model, such as: strip entry thickness, strip width, strip elongation, entry tension, exit tension, average diameter of the work rolls, rolling speed, and bending force. The data from each coil is averaged to obtain a set of data for each coil.
[0090] S300. Based on the physical model, the calculated rolling force of the multiple coils is obtained according to the actual rolling process parameters of the multiple coils, and the proportional deviation between the calculated rolling force of the multiple coils and the actual rolling force of the multiple coils is used as the objective function.
[0091] In one embodiment, the absolute value of the ratio deviation between the calculated rolling force and the actual rolling force for each coil is taken, and the data of the 10% with the largest absolute value of the ratio deviation is discarded. The average value of the remaining absolute values of the ratio deviation is used as the objective function. Here, the ratio deviation between the calculated rolling force and the actual rolling force refers to the ratio obtained by dividing the difference between the calculated rolling force and the actual rolling force by the actual rolling force.
[0092] S400. Based on the actual rolling process parameters and objective function of the multi-coil rolls, the unknown coefficients in the physical model are determined by solving the differential evolution algorithm.
[0093] In one embodiment, S400 includes the following steps:
[0094] S410. Initialization: Set the population size, mutation factor, crossover probability, and maximum number of iterations, and randomly generate an initial unknown coefficient vector population.
[0095] S420, Fitness Assessment: Substitute the unknown coefficient vector into the physical model to calculate the predicted rolling force corresponding to the actual rolling process parameters of each coil, and calculate the fitness value of each individual according to the objective function.
[0096] S430, Mutation Operation: Perform a mutation operation on each individual in the current population to generate a mutation vector;
[0097] S440, Crossover operation: Cross the mutation vector with the target individual to generate an experimental vector;
[0098] S450, Boundary Processing: Perform boundary reflection or reset processing on coefficients in the test vector that exceed the preset value range;
[0099] S460, Selection Operation: Compare the fitness values of the experimental vectors and the target individuals, and select the better ones to enter the next generation of the population;
[0100] S470, Iteration termination judgment: Repeat steps S420-S460 until the maximum number of iterations is reached or the fitness value reaches the target fitness value;
[0101] S480, Optimal Coefficient Output: The individual with the best fitness value in the final population is used as the unknown coefficient vector of the physical model determined by the solution.
[0102] S500. Substitute the unknown coefficients determined by the solution into the physical model for prediction of the rolling force of the finishing mill.
[0103] Verification Example:
[0104] Existing patent CN119806043A mentions that 49 sets of actual production data of a certain leveling mill were selected, and the classic Roberts leveling rolling force calculation formula was used to compare with the actual rolling force in production. The result obtained by the Roberts calculation model differed from the actual leveling rolling force data by about 6.9 times, indicating that the Roberts calculation model was too small compared with the actual value.
[0105] Data from 658 rolls of DX56D mild steel were collected on-site. (See reference.) Figure 2 As shown, the collected data items include thickness, width, speed, elongation, bending force, rolling force, inlet tension, and outlet tension. Representative values were selected for each coil, resulting in 658 sets of actual data. The yield strength of this grade was measured at 145 MPa on-site, and the coefficient of friction is typically taken as 0.2. Using the classic Roberts model, the predicted rolling force is concentrated around 60% to 70% of the actual rolling force. If a reverse calculation is performed, the classic Roberts model yields a friction coefficient of 0.33, or an increased yield strength of 254.8 MPa, neither of which conforms to normal process experience.
[0106] See Figure 3 , 4 As shown, the Roberts model recommends a coefficient of 51.6 MPa. If the proposed solution is used for self-learning based on real data, the optimized coefficient is 75.9 MPa. At this point, the fitness (average rolling force deviation) is 11.62%. Using the proposed solution without rolling force correction, the rolling force prediction deviation is 10.74%; using the proposed solution considering thickness for rolling force correction, the deviation is 10.59%; using the proposed solution considering roll bending force for rolling force correction, the deviation is 9.87%; and using the proposed solution considering both thickness and roll bending force for rolling force correction, the deviation is 9.73%. The comparison shows that the proposed solution significantly improves accuracy.
[0107] This invention considers the influence of rolling speed on deformation resistance, resulting in a rolling force trend that conforms to reality. By taking into account the influence of elongation on the length of the deformation zone, the rolling force is made close to the on-site calculation, achieving a highly interpretable rolling force calculation scheme for a finished mill. This scheme is simple and fast to calculate, and subsequent back-calculation of elongation from rolling force is extremely straightforward, facilitating further control optimization. A single calculation takes less than 10ms, meeting online control requirements. Furthermore, due to its simple structure, fast calculation, and rapid self-learning of correlation coefficients, it exhibits strong real-time performance in on-site applications.
[0108] Compared to existing technologies, this invention offers significant advantages: First, it is computationally simple and efficient: the model is built based on analytical expressions, requiring no complex iterations, and can even be directly embedded into a PLC control system, consuming minimal computational resources and facilitating maintenance; second, it boasts high prediction accuracy: after self-learning, the relative error of rolling force prediction can be stably controlled within ±5%, significantly reducing setting deviations compared to mainstream classical formulas; third, it exhibits strong scalability: even after specification changes, the calculation results still maintain high accuracy and excellent interpretability, with generalization capabilities far exceeding those of pure data models. This method combines mechanistic interpretability with data adaptability, providing a reliable and practical engineering solution for the intelligent control of the finishing process.
[0109] Example 2
[0110] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method for predicting the rolling force of a smooth mill.
[0111] The storage medium may be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium may also include combinations of the above types of memory.
[0112] Example 3
[0113] This embodiment provides a computer device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the above-described method for predicting the rolling force of a smooth mill.
[0114] like Figure 5As shown, the computer device 70 may include: at least one processor 71, such as a CPU (Central Processing Unit), at least one communication interface 73, a memory 74, and at least one communication bus 72. The communication bus 72 is used to enable communication between these components. The communication interface 73 may include a display screen and a keyboard; optionally, the communication interface 73 may also include a standard wired interface or a wireless interface. The memory 74 may be high-speed RAM (Random Access Memory) or non-volatile memory, such as at least one disk storage device. Optionally, the memory 74 may also be at least one storage device located remotely from the aforementioned processor 71. The memory 74 stores application programs, and the processor 71 calls the program code stored in the memory 74 to execute any of the above-described method steps.
[0115] The communication bus 72 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 72 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.
[0116] The memory 74 may include volatile memory, such as random-access memory (RAM); the memory may also include non-volatile memory, such as flash memory, hard disk drive (HDD) or solid-state drive (SSD); the memory 74 may also include a combination of the above types of memory.
[0117] The processor 71 can be a central processing unit (CPU), a network processor (NP), or a combination of CPU and NP.
[0118] The processor 71 may further include a hardware chip. This hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.
[0119] Optionally, the memory 74 is also used to store program instructions. The processor 71 can call the program instructions to implement the rolling force prediction method for a finishing mill as described in this invention.
[0120] Those skilled in the art should understand that the present invention can be implemented in many other specific forms without departing from the spirit and scope of the invention. Any changes or modifications made by those skilled in the art based on the embodiments of the present invention and the above disclosure shall fall within the protection scope of the claims.
Claims
1. A method for predicting rolling force in a finishing mill, characterized in that, The method includes the following steps: S100. A physical model between rolling force and rolling process parameters is established based on equivalent deformation resistance and equivalent contact arc length. S200: Collect actual rolling process parameters and actual rolling force of multiple coils at the rolling site; S300. Based on the physical model, the calculated rolling force of the multiple coils is obtained according to the actual rolling process parameters of the multiple coils, and the proportional deviation between the calculated rolling force of the multiple coils and the actual rolling force of the multiple coils is used as the objective function. S400. Determine the unknown coefficients in the physical model based on the actual rolling process parameters and objective function of the multi-coil using the differential evolution algorithm; S500. Substitute the unknown coefficients determined by the solution into the physical model for prediction of the rolling force of the finishing mill.
2. The method for predicting rolling force in a finishing mill according to claim 1, characterized in that, Step S100 includes: S110. Obtain rolling process parameters and unit design parameters; S120. Calculate the equivalent contact arc length and equivalent deformation resistance based on the rolling process parameters and unit design parameters. S130. Calculate the rolling force based on the equivalent contact arc length and equivalent deformation resistance.
3. The method for predicting rolling force in a finishing mill according to claim 2, characterized in that, Rolling process parameters include strip entry thickness, strip width, strip elongation, entry tension, exit tension, average diameter of work rolls, rolling speed, strip yield strength, and coefficient of friction between strip and work rolls; unit design parameters include maximum rolling speed, maximum diameter of work rolls, and maximum strip elongation. The formula for calculating the equivalent contact arc length is as follows: in, For the equivalent contact arc length, These are the coefficients of the first model. For strip elongation, This represents the maximum elongation of the strip. The average diameter of the working rolls. For reduction rate, The coefficient of friction between the strip and the work roll is denoted as . The strip entry thickness is given; the reduction rate is calculated using the following formula: ; The formula for calculating equivalent deformation resistance is as follows: in, For equivalent deformation resistance, The yield strength of the strip steel. These are the coefficients of the second model. For rolling speed, For maximum rolling speed, The maximum diameter of the working rolls. These are the coefficients of the third model. For the inlet tensile stress, These are the coefficients of the fourth model. The outlet tensile stress is given by the formulas below; the inlet and outlet tensile stresses are calculated as follows: in, For inlet tension, For strip width, For export tension, The thickness is the export thickness; the formula for calculating the export thickness is as follows: ; The formula for calculating rolling force is as follows: in, This refers to the rolling force.
4. The method for predicting rolling force in a finishing mill according to claim 3, characterized in that, The rolling force is corrected to account for the influence of the bending roll force on the rolling force, as shown in the following formula: in, The corrected rolling force, The coefficients of the fifth model, For the bending roller force, These are the coefficients of the sixth model.
5. The method for predicting rolling force in a finishing mill according to claim 3, characterized in that, The rolling force is corrected to account for the effect of thickness on the rolling force, as shown in the following formula: in, The corrected rolling force, These are the coefficients of the seventh model. These are the coefficients of the eighth model. =1mm, which is used to eliminate units.
6. The method for predicting rolling force in a finishing mill according to claim 3, characterized in that, The rolling force is modified by considering the influence of bending roll force and thickness on the rolling force, as shown in the following formula: in, The corrected rolling force, The coefficients of the fifth model, For the bending roller force, The coefficients of the sixth model, These are the coefficients of the seventh model. These are the coefficients of the eighth model. =1mm, which is used to eliminate units.
7. The method for predicting rolling force in a finishing mill according to claim 1, characterized in that, In step S300, the absolute value of the ratio deviation between the calculated rolling force and the actual rolling force for each coil is taken, and the data of the 10% portion with the largest absolute value of the ratio deviation is removed. The average value of the remaining absolute values of the ratio deviation is taken as the objective function.
8. The method for predicting rolling force of a finishing mill according to any one of claims 1 to 7, characterized in that, S400 includes the following steps: S410. Initialization: Set the population size, mutation factor, crossover probability, and maximum number of iterations, and randomly generate an initial unknown coefficient vector population. S420, Fitness Assessment: Substitute the unknown coefficient vector into the physical model to calculate the predicted rolling force corresponding to the actual rolling process parameters of each coil, and calculate the fitness value of each individual according to the objective function. S430, Mutation Operation: Perform a mutation operation on each individual in the current population to generate a mutation vector; S440, Crossover operation: Cross the mutation vector with the target individual to generate an experimental vector; S450, Boundary Processing: Perform boundary reflection or reset processing on coefficients in the test vector that exceed the preset value range; S460, Selection Operation: Compare the fitness values of the experimental vectors and the target individuals, and select the better ones to enter the next generation of the population; S470, Iteration termination judgment: Repeat steps S420-S460 until the maximum number of iterations is reached or the fitness value reaches the target fitness value; S480, Optimal Coefficient Output: The individual with the best fitness value in the final population is used as the unknown coefficient vector of the physical model determined by the solution.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for predicting the rolling force of a smooth mill as described in any one of claims 1-8.
10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method for predicting rolling force of a polished mill as described in any one of claims 1-8.