A method, device and equipment for pre-adjustment of a skin pass mill based on strip roughness

Abstract

The application discloses a kind of based on strip roughness's skin pass mill pre-regulation method, device and equipment, comprising: according to the first parameter kind's historical production parameter determines pre-drive model, the first production parameter of the product to be produced matched with the first parameter kind is input into pre-drive model, obtains pre-drive parameter and strip roughness target value;According to the second parameter kind's historical production parameter determines roughness prediction model, according to pre-drive parameter and roughness prediction model, determine strip roughness initial value;According to strip roughness initial value and strip roughness target value, determine roughness error;In the case where roughness error meets preset error standard, strip roughness initial value is determined as strip roughness prediction value;According to strip roughness prediction value, first production parameter and pre-drive model, determine target drive parameter, and according to target drive parameter, skin pass mill is adjusted.Can be adjusted according to strip roughness to rolling parameter.

Description

Technical Field

[0001] This invention relates to the field of metal material processing technology, and in particular to a pre-adjustment method, apparatus and equipment for a finishing machine based on the roughness of strip steel. Background Technology

[0002] A finishing mill is a device used to improve the surface quality and roughness of a product. Taking a strip galvanizing production line as an example, the finishing mill uses small deformation rolling to eliminate the yield plateau after strip annealing, thereby adjusting the surface roughness of the strip.

[0003] In existing technologies, the parameters of the finishing machine are set manually, resulting in low control precision and an inability to meet users' complex requirements for strip surface roughness. Therefore, how to accurately adjust the finishing machine is a technical problem that urgently needs to be solved. Summary of the Invention

[0004] This application provides a pre-adjustment method, apparatus, and equipment for a finishing machine based on strip roughness, which solves the technical problem of low control accuracy of the finishing machine in the prior art, making it unable to meet the complex requirements of users for the surface roughness of the strip, and achieves the technical effect of accurately adjusting the finishing machine.

[0005] In a first aspect, this application provides a pre-adjustment method for a finishing machine based on strip roughness, comprising:

[0006] The pre-driving model is determined based on the historical production parameters of the first parameter type. The first production parameters of the product to be produced that match the first parameter type are input into the pre-driving model to obtain the pre-driving parameters and the target value of strip roughness.

[0007] The roughness prediction model is determined based on the historical production parameters of the second parameter type, and the initial value of strip roughness is determined based on the pre-driving parameters and the roughness prediction model.

[0008] The roughness error is determined based on the initial and target values ​​of the strip roughness.

[0009] If the roughness error meets the preset error standard, the initial value of the strip roughness is determined as the predicted value of the strip roughness.

[0010] Based on the predicted strip roughness value, the first production parameter, and the pre-drive model, the target drive parameter is determined, and the finishing machine is adjusted according to the target drive parameter.

[0011] In some embodiments of this application, based on the foregoing scheme, the first parameter types include at least one of the following: steel grade, steel tapping mark, thickness, width, elongation, roll roughness, roll diameter, zinc coating type, strip set roughness, and production season;

[0012] Pre-drive parameters include at least one of the following: rolling force, bending roll force, inlet tension, and outlet tension;

[0013] The second parameter types include at least one of the following: steel grade, roll diameter, roll surface roughness, thickness, width, elongation, rolling force, roll bending force, inlet tension, outlet tension, and rolling kilometers.

[0014] In some embodiments of this application, based on the foregoing scheme, the pre-driving model is determined according to historical production parameters of the first parameter type, including:

[0015] The initial model is determined based on the historical production parameters of the first parameter category that matches the product to be produced;

[0016] Based on the initial predicted values ​​output by the initial model and the historical production parameters that match the initial predicted values, determine the coefficient of determination and the mean absolute percentage error.

[0017] If the coefficient of determination and the mean absolute percentage error both meet the preset model criteria, the initial model is determined as the pre-driven model.

[0018] In some embodiments of this application, based on the foregoing scheme, when the roughness error does not meet the preset error standard, the method further includes:

[0019] If the roughness error satisfies the first preset relationship with the preset error standard, the pre-driving parameters are positively adjusted according to the first preset step size, and the process returns to the step of determining the initial value of the strip roughness based on the pre-driving parameters and the roughness prediction model.

[0020] If the roughness error satisfies the second preset relationship with the preset error standard, the pre-driving parameters are negatively adjusted according to the second preset step size, and the process returns to the step of determining the initial value of strip roughness based on the pre-driving parameters and the roughness prediction model.

[0021] In some embodiments of this application, based on the foregoing scheme, when the roughness error does not meet the preset error standard, the method further includes:

[0022] If the pre-driving parameter exceeds the preset upper limit, update the strip roughness target value according to the strip roughness target value and the preset roughness lower limit value, and return to the step of determining the roughness error according to the initial strip roughness value and the strip roughness target value.

[0023] If the pre-driving parameter is lower than the preset lower limit, the strip roughness target value is updated according to the strip roughness target value and the preset roughness upper limit value, and the process returns to the step of determining the roughness error based on the initial strip roughness value and the strip roughness target value.

[0024] In some embodiments of this application, based on the foregoing scheme, when the pre-driving parameter exceeds a preset upper limit or falls below a preset lower limit, the method further includes:

[0025] The cumulative number of updates to the target strip roughness value is calculated. If the number of updates exceeds the preset number standard, the roll is replaced and the process is returned to execute the steps of determining the pre-drive model based on the historical production parameters of the first parameter type, inputting the first production parameters of the product to be produced that match the first parameter type into the pre-drive model, and obtaining the pre-drive parameters and the target strip roughness value.

[0026] In some embodiments of this application, based on the foregoing scheme, the target driving parameters are determined according to the predicted strip roughness value, the first production parameter, and the pre-driving model, including:

[0027] Based on the predicted strip roughness value, the first production parameter, and the pre-driving model, determine the predicted driving parameters corresponding to at least two consecutive production times.

[0028] If at least one set of two adjacent predicted driving parameters meet a preset difference standard, the predicted driving parameter is determined as the target driving parameter.

[0029] If at least one set of adjacent prediction driving parameters does not meet the preset difference standard, adjust the pre-driving parameters.

[0030] In some embodiments of this application, based on the foregoing scheme, before determining the pre-driving model according to historical production parameters of the first parameter type, the method further includes:

[0031] Determine whether the known parameter types of the product to be produced match the parameter types included in the historical production parameters.

[0032] Secondly, this application provides a pre-conditioning device for an optical rectifier, comprising:

[0033] The pre-driving parameter determination module is used to determine the pre-driving model based on the historical production parameters of the first parameter type. The first production parameters of the product to be produced that match the first parameter type are input into the pre-driving model to obtain the pre-driving parameters and the target value of strip roughness.

[0034] The strip roughness initial value determination module is used to determine the roughness prediction model based on the historical production parameters of the second parameter type, and to determine the initial value of strip roughness based on the pre-driving parameters and the roughness prediction model.

[0035] The roughness error determination module is used to determine the roughness error based on the initial value and target value of the strip roughness.

[0036] The strip roughness prediction value determination module is used to determine the initial value of strip roughness as the predicted value of strip roughness when the roughness error meets the preset error standard.

[0037] The adjustment module is used to determine the target driving parameters based on the predicted strip roughness value, the first production parameters, and the pre-drive model, and to adjust the finishing machine according to the target driving parameters.

[0038] Thirdly, this application provides an electronic device, comprising:

[0039] processor;

[0040] Memory used to store processor-executable instructions;

[0041] The processor is configured to execute a pre-conditioning method for a light finishing machine based on strip roughness, as provided in the first aspect.

[0042] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0043] This application provides a pre-adjustment method for a light finishing machine based on strip roughness, comprising: determining a pre-driving model based on historical production parameters of a first parameter type; inputting the first production parameters of the product to be produced, which match the first parameter type, into the pre-driving model to obtain pre-driving parameters and a target value for strip roughness; determining a roughness prediction model based on historical production parameters of a second parameter type; determining an initial value for strip roughness based on the pre-driving parameters and the roughness prediction model; determining a roughness error based on the initial value and the target value for strip roughness; determining the initial value for strip roughness as a predicted value for strip roughness if the roughness error meets a preset error standard; determining a target driving parameter based on the predicted value for strip roughness, the first production parameters, and the pre-driving model; and adjusting the light finishing machine based on the target driving parameter.

[0044] It is evident that the dual-layer architecture of the pre-driven model and the roughness prediction model, based on historical production parameters, can automatically determine the optimal rolling parameters for the finishing mill and accurately predict the surface roughness of the strip. Furthermore, combined with roughness error feedback, it can correct roughness fluctuations caused by parameter lag, enabling intelligent pre-setting and online fine-tuning of key parameters such as rolling force and tension. This significantly improves the process response speed and production continuity of the finishing mill, while reducing equipment wear and energy consumption losses caused by repeated trial and error. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 A flowchart illustrating a pre-adjustment method for a finishing machine based on strip roughness, provided for an embodiment of this application;

[0047] Figure 2 A schematic diagram illustrating the prediction accuracy of a pre-driven model provided in an embodiment of this application;

[0048] Figure 3 A schematic diagram illustrating the prediction accuracy of a roughness prediction model provided in an embodiment of this application;

[0049] Figure 4 A schematic diagram of rolling parameter prediction data for a pre-adjustment method for a finishing mill based on strip roughness provided in this application embodiment;

[0050] Figure 5a A schematic diagram illustrating the rolling force prediction accuracy of a pre-adjustment method for a finishing mill based on strip roughness, provided in an embodiment of this application;

[0051] Figure 5b A schematic diagram illustrating the bending roll force prediction accuracy of a pre-adjustment method for a finishing machine based on strip roughness, provided in an embodiment of this application;

[0052] Figure 5c A schematic diagram illustrating the inlet tension prediction accuracy of a pre-adjustment method for a finishing machine based on strip roughness, provided in an embodiment of this application;

[0053] Figure 5d A schematic diagram illustrating the exit tension prediction accuracy of a pre-adjustment method for a finishing machine based on strip roughness, provided in an embodiment of this application;

[0054] Figure 6 A schematic diagram of a pre-adjustment device for a finishing machine based on strip roughness is provided in this application embodiment;

[0055] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0056] This application provides a pre-adjustment method for a finishing machine based on strip roughness, which solves the technical problem that the existing finishing machine has low control accuracy and cannot meet the complex requirements of users for strip surface roughness.

[0057] The technical solution of this application embodiment is to solve the above-mentioned technical problems, and the general idea is as follows:

[0058] This application provides a pre-adjustment method for a light finishing machine based on strip roughness, comprising: determining a pre-driving model based on historical production parameters of a first parameter type; inputting the first production parameters of the product to be produced, which match the first parameter type, into the pre-driving model to obtain pre-driving parameters and a target value for strip roughness; determining a roughness prediction model based on historical production parameters of a second parameter type; determining an initial value for strip roughness based on the pre-driving parameters and the roughness prediction model; determining a roughness error based on the initial value and the target value for strip roughness; determining the initial value for strip roughness as a predicted value for strip roughness if the roughness error meets a preset error standard; determining a target driving parameter based on the predicted value for strip roughness, the first production parameters, and the pre-driving model; and adjusting the light finishing machine based on the target driving parameter.

[0059] It is evident that the dual-layer architecture of the pre-driven model and the roughness prediction model, based on historical production parameters, can automatically determine the optimal rolling parameters and accurately predict the surface roughness of the strip. Furthermore, by incorporating roughness error feedback, it corrects the roughness fluctuation problem caused by parameter lag, enabling intelligent pre-setting and online fine-tuning of key parameters such as rolling force and tension. This significantly improves the process response speed and production continuity of the finishing mill, while reducing equipment wear and energy consumption losses caused by repeated trial and error.

[0060] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0061] First, it should be clarified that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0062] As the core equipment of the cold-rolled strip galvanizing production line, the finishing mill eliminates the yield plateau of strip after annealing through micro-rolling deformation, optimizes processing performance, and precisely controls the surface roughness and micromorphology of strip, which plays a decisive role in the mechanical properties, shape accuracy and surface quality of the product.

[0063] However, in existing technologies, the rolling force setting model does not incorporate strip surface roughness parameters and lacks online roughness detection devices, resulting in insufficient roughness control accuracy during actual rolling. This makes it difficult to meet the stringent requirements of high-end users such as automotive steel sheet manufacturers for diverse surface morphologies and narrow roughness windows. On the other hand, the traditional manual setting mode relies on operational experience, and the adjustment of rolling force and tension parameters is lagging, causing excessive elongation at the strip head and significant fluctuations in elongation during acceleration and deceleration. At the same time, the frequent roll changes caused by uncontrolled roughness further exacerbate the loss of production efficiency.

[0064] Taking a domestic automotive sheet galvanizing production line as an example, its finishing machine parameters are entirely controlled by manual intervention. The surface roughness dispersion, sheet shape defects and elongation instability caused by rolling force setting errors are prominent, which seriously restricts the stability of product quality and the level of production line intelligence.

[0065] To address the aforementioned problems, embodiments of this application provide a pre-adjustment method for a finishing machine based on strip roughness. For example... Figure 1 The diagram shown is a flowchart of a pre-adjustment method for a finishing machine based on strip roughness provided in an embodiment of this application, including steps S1-S5.

[0066] Step S1: Determine the pre-driving model based on the historical production parameters of the first parameter type, and input the first production parameters of the product to be produced that match the first parameter type into the pre-driving model to obtain the pre-driving parameters and the target value of strip roughness.

[0067] Step S2: Determine the roughness prediction model based on the historical production parameters of the second parameter type, and determine the initial value of strip roughness based on the pre-driving parameters and the roughness prediction model;

[0068] Step S3: Determine the roughness error based on the initial value and target value of the strip roughness.

[0069] Step S4: If the roughness error meets the preset error standard, the initial value of the strip roughness is determined as the predicted value of the strip roughness.

[0070] Step S5: Determine the target driving parameters based on the predicted strip roughness value, the first production parameters, and the pre-driving model, and adjust the finishing machine according to the target driving parameters.

[0071] Regarding step S1, the pre-driving model is determined based on the historical production parameters of the first parameter type. The first production parameters of the product to be produced that match the first parameter type are input into the pre-driving model to obtain the pre-driving parameters and the target value of strip roughness.

[0072] It should be noted that the first parameter type refers to the core process parameters that affect the strip roughness and rolling stability. The first parameter type can include at least one of the following: steel grade, tap mark, thickness, width, elongation, roll roughness, roll diameter, zinc coating type, strip set roughness, and production season.

[0073] Historical production parameters encompass multi-dimensional process data recorded during past production processes, thus forming the data foundation for model training.

[0074] The input values ​​of the pre-driven model are the first production parameters of the product to be produced, and the output values ​​are the pre-driven parameters. The first production parameters are known parameters in the contract plan for the product to be produced. The pre-driven parameters may include at least one of the following: rolling force, bending force, inlet tension, and outlet tension.

[0075] For example, when the product to be produced is one that has never been produced before, one or more of its steel grade, contract thickness, contract width, and elongation setting values ​​may be unknown and cannot be accurately queried from historical production parameters. Therefore, a pre-drive model is determined based on the steel grade, tap mark, thickness, width, elongation, roll roughness, roll diameter, zinc coating type, strip roughness setting, and production season from historical production parameters. The known first production parameters of the product to be produced are input into the pre-drive model to preset the rolling force, bending force, inlet tension, and outlet tension even when one or more of the steel grade, contract thickness, contract width, and elongation setting values ​​are unknown.

[0076] The target value for strip roughness is the desired surface roughness of the strip, which can be determined based on historical production parameters or the contract plan for the products to be produced.

[0077] Regarding step S1, the pre-driving model is determined based on the historical production parameters of the first parameter type, including steps S11-S13.

[0078] Step S11: Determine the initial model based on the historical production parameters of the first parameter type that matches the product to be produced;

[0079] Step S12: Determine the coefficient of determination and the mean absolute percentage error based on the initial predicted value output by the initial model and the historical production parameters that match the initial predicted value.

[0080] Step S13: If the coefficient of determination and the mean absolute percentage error both meet the preset model criteria, the initial model is determined as the pre-driven model.

[0081] Regarding step S11, the initial model is determined based on the historical production parameters of the first parameter type that matches the product to be produced.

[0082] An initial model is constructed based on historical production data related to the first type of parameters (such as thickness, width, elongation, roll roughness, and roll diameter) of the product to be manufactured. Feature analysis of the historical production data reveals that these parameters typically exhibit a normal distribution within a certain production cycle and possess significant characterizing power, providing a statistical basis for the rationality of the model's input features. Comparative studies can be conducted on the applicability of different machine learning algorithms (such as random forests, support vector machines, or neural networks). For example, by considering the algorithm's ability to handle high-dimensional data, its efficiency in fitting nonlinear relationships, and its adaptability in noise resistance, the modeling method that best matches the data distribution characteristics and the product to be manufactured can be selected.

[0083] This process requires constructing a data sample set encompassing multidimensional historical production data, and dividing it into training and validation sets according to time series (e.g., using an 8:2 ratio). The validation set prioritizes recent historical production data (e.g., within one month of the current production date) to ensure the model effectively captures dynamic changes in process conditions and avoids prediction biases caused by equipment aging or environmental factors. The resulting initial model possesses generalization and timeliness, providing a foundation for subsequent parameter predictions.

[0084] Regarding step S12, the determination coefficient and mean absolute percentage error are determined based on the initial predicted value output by the initial model and the historical production parameters that match the initial predicted value.

[0085] Coefficient of determination R 2 R can quantify the model's ability to explain data fluctuations. 2 The closer the coefficient of determination (R²) is to 1, the stronger the model's ability to explain fluctuations in actual values. 2 This can be expressed by the following formula:

[0086]

[0087] Among them, y i Let i be the actual value of the i-th observation. The predicted value for the i-th observation. is the average of the actual values ​​of all observations, where n is the total number of observations.

[0088] Mean Absolute Percentage Error (MAPE) measures the relative error level of predicted values. A smaller MAPE indicates a lower average percentage deviation between predicted and actual values, and thus higher model accuracy. MAPE can be expressed by the following formula:

[0089]

[0090] Among them, y i Let i be the actual value of the i-th observation. Let be the predicted value of the i-th observation, and n be the total number of observations.

[0091] Regarding step S13, if both the coefficient of determination and the mean absolute percentage error meet the preset model criteria, the initial model is determined as the pre-driven model.

[0092] It should be noted that for different types of parameters output by the pre-driven model, different preset model standards can be set for each type of parameter.

[0093] Taking rolling force, bending force, inlet tension and outlet tension as examples, the corresponding preset model standards are shown in Table 1.

[0094] Table 1 Preset Model Standards

[0095] <![CDATA[R 2 ]]> 0.987 0.988 0.996 0.996 MAPE 2.5％ 5.0％ 1.3％ 1.1％

[0096] Furthermore, before determining the pre-driving model based on historical production parameters of the first parameter type, the method also includes:

[0097] Determine whether the known parameter types of the product to be produced match the parameter types included in the historical production parameters.

[0098] In the case of a perfect match, the pre-driving parameters and the target value of strip roughness are determined based on the matching historical production parameters, and the roughness prediction model is determined directly based on the historical production parameters of the second parameter type. The initial value of strip roughness is then determined based on the pre-driving parameters and the roughness prediction model.

[0099] Specifically, the known parameters of the product to be produced (such as steel grade, contract thickness, contract width, etc.) are compared with the parameters in the historical production parameter database. If all parameters can be completely matched in the historical records (for example, the product has been produced before and its steel grade name, contract specifications, elongation setting value, and user requirements are completely stored in the historical production parameters), then the optimal process parameters (such as rolling force, bending roll force, inlet tension, and outlet tension) in the historical production parameters are determined as the pre-drive parameters and automatically sent to the production line for execution.

[0100] For example, Tables 2 and 3 are rolling parameter lookup tables provided in the embodiments of this application. All parameters of the product to be produced can be completely matched in the historical records, and the corresponding pre-driving parameters are shown in Tables 2 and 3.

[0101] Table 2 Rolling parameters for product A to be manufactured

[0102]

[0103] Table 3 Rolling parameters for product B to be manufactured

[0104]

[0105] Regarding step S2, the roughness prediction model is determined based on the historical production parameters of the second parameter type, and the initial value of strip roughness is determined based on the pre-driving parameters and the roughness prediction model.

[0106] The second parameter can include at least one of the following: steel grade, roll diameter, roll surface roughness, thickness, width, elongation, rolling force, roll bending force, inlet tension, outlet tension, and rolling kilometers.

[0107] The steps for determining the roughness prediction model can refer to the steps for determining the pre-driven model described above. Using multidimensional feature variables strongly correlated with strip surface roughness (such as steel grade, roll diameter, roll surface roughness, thickness, width, elongation, rolling force, bending force, inlet tension, outlet tension, and rolling mileage), and employing the same historical sample set as the pre-driven model, and dividing it into training and validation sets, a mapping relationship between process settings and strip surface roughness is established, i.e., the roughness prediction model.

[0108] The dynamic parameters output by the pre-driven model, such as rolling force, bending force, inlet tension, and outlet tension, or historical production parameters that perfectly match the product to be produced, are then used as input values ​​for the roughness prediction model to determine the initial value of the strip roughness.

[0109] Regarding step S3, the roughness error is determined based on the initial value and the target value of the strip roughness.

[0110] Roughness error reflects the degree of deviation of roughness under the current process parameter settings, and can be expressed by the following formula.

[0111]

[0112] Where σ is the roughness error, and Rα t Rα is the target roughness value for the strip steel. s This represents the initial value for the roughness of the strip.

[0113] Regarding step S4, if the roughness error meets the preset error standard, the initial value of the strip roughness is determined as the predicted value of the strip roughness.

[0114] The roughness error conforming to the preset error standard means that the absolute value of the roughness error is less than the preset error standard m, that is, |σ|≤m.

[0115] Furthermore, if the roughness error does not meet the preset error standard, the method also includes steps S41-S42.

[0116] Step S41: If the roughness error satisfies the first preset relationship with the preset error standard, the pre-driving parameters are positively adjusted according to the first preset step size, and the process returns to the step of determining the initial value of the strip roughness based on the pre-driving parameters and the roughness prediction model.

[0117] Step S42: If the roughness error satisfies the second preset relationship with the preset error standard, the pre-driving parameters are negatively adjusted according to the second preset step size, and the process returns to the step of determining the initial value of the strip roughness based on the pre-driving parameters and the roughness prediction model.

[0118] Regarding step S41, if the roughness error satisfies the first preset relationship with the preset error standard, the pre-driving parameters are positively adjusted according to the first preset step size, and the process returns to the step of determining the initial value of the strip roughness based on the pre-driving parameters and the roughness prediction model.

[0119] Specifically, the first preset relationship is that the roughness error exceeds the preset error standard, i.e., σ>m. The pre-driving parameter adjusted in step S41 can be the inlet tension, i.e., the tension applied to the strip when it enters the finishing mill (rolling mill) during cold rolling, used to control the elongation and surface quality of the strip. The inlet tension is gradually increased according to a preset step size k1% (each adjustment is k1% of the current inlet tension value, for example, 2%) to improve rolling smoothness and achieve the purpose of adjusting the surface roughness of the strip.

[0120] Furthermore, if the pre-driving parameter exceeds the preset upper limit, the strip roughness target value is updated based on the strip roughness target value and the preset roughness lower limit value, and the process returns to the step of determining the roughness error based on the initial strip roughness value and the strip roughness target value.

[0121] Because excessive inlet tension increases the risk of strip breakage or abnormal roll wear, the strip roughness target value is updated after the inlet tension exceeds the preset upper limit. Rα t Rα is the original target value for strip roughness. low This is the preset lower limit value for roughness.

[0122] Regarding step S42, if the roughness error satisfies the second preset relationship with the preset error standard, the pre-driving parameters are negatively adjusted according to the second preset step size, and the process returns to the step of determining the initial value of the strip roughness based on the pre-driving parameters and the roughness prediction model.

[0123] Specifically, the second preset relationship is that the roughness error is lower than the negative of the preset error standard, i.e., σ < -m. The pre-driving parameter adjusted in step S42 can be the inlet tension. The inlet tension is gradually reduced according to a preset step size k2% (the adjustment range is k2% of the current inlet tension value each time, for example, 2%) to improve the rolling flatness and achieve the purpose of adjusting the surface roughness of the strip.

[0124] Furthermore, if the pre-driving parameter is lower than the preset lower limit, the strip roughness target value is updated based on the strip roughness target value and the preset roughness upper limit value, and the process returns to the step of determining the roughness error based on the initial strip roughness value and the strip roughness target value.

[0125] If reducing the inlet tension fails to prevent the strip roughness from decreasing (reaching the preset lower limit), it indicates that the current process conditions cannot meet the target strip roughness value. Production standards need to be relaxed to reduce production accidents caused by excessive pursuit of a smooth surface. Therefore, once the inlet tension falls below the preset lower limit, the target strip roughness value is updated to [value missing]. Rα t Rα is the original target value for strip roughness. high This is the preset upper limit value for roughness.

[0126] Furthermore, if the pre-driving parameter exceeds a preset upper limit or falls below a preset lower limit, the method further includes:

[0127] The cumulative number of updates to the target strip roughness value is calculated. If the number of updates exceeds the preset number standard, the roll is replaced and the process is returned to execute the steps of determining the pre-drive model based on the historical production parameters of the first parameter type, inputting the first production parameters of the product to be produced that match the first parameter type into the pre-drive model, and obtaining the pre-drive parameters and the target strip roughness value.

[0128] For example, if the preset number of updates is 5, and the number of updates exceeds 5, it is determined that the current roll condition cannot meet the process requirements. If production continues, it may lead to excessive wear of the roll, further increasing the risk of uncontrolled strip roughness. Therefore, the roll needs to be replaced and the process returns to step S1. The pre-drive model is then redefined based on the new roll, and subsequent steps S1-S5 are executed sequentially.

[0129] Regarding step S5, the target driving parameters are determined based on the predicted strip roughness value, the first production parameters, and the pre-driving model, and the finishing machine is adjusted according to the target driving parameters.

[0130] Based on the predicted strip roughness value, the first production parameter, and the pre-driving model, the target driving parameter is determined, including steps S51-S53.

[0131] Step S51: Based on the predicted strip roughness value, the first production parameter, and the pre-driving model, determine the predicted driving parameters corresponding to at least two consecutive production times.

[0132] Step S52: If at least one set of two adjacent predicted driving parameters meet the preset difference standard, the predicted driving parameter is determined as the target driving parameter.

[0133] Step S53: If at least one set of two adjacent prediction driving parameters does not meet the preset difference standard, adjust the pre-driving parameters.

[0134] Regarding steps S51-S53, at least two sets of predictive driving parameters are generated within a continuous production time window (e.g., every 10 seconds or every coil of strip segmentation). The dynamic evolution trend of process parameters is captured through time series prediction, providing a data basis for stability verification.

[0135] Specifically, the predicted strip roughness value obtained from the aforementioned steps and the first production parameter are input into the pre-driven model, and at least two sets of predicted driving parameters are output within a continuous production time window (e.g., every 10 seconds or every coil of strip segmentation). If the difference between at least one set of two adjacent predicted driving parameters is within a threshold (e.g., the difference in rolling force is less than a preset difference standard n%, such as 5%), then the predicted driving parameters are determined to be stable and can be used as target driving parameters to adjust the finishing mill.

[0136] The following two specific embodiments illustrate in detail a pre-adjustment method for a finishing machine based on strip roughness provided in this application.

[0137] Example 1

[0138] In the bright finishing mill rolling process of a galvanizing production line, when the product to be produced in the production plan is a product that has already been produced, and its steel grade name, contract thickness, contract width, elongation setting value, and end user are all in the historical production data, the optimal process parameters (such as rolling force, bending roll force, and inlet / outlet tension) in the historical production parameters are determined as pre-drive parameters and automatically sent to the production line for execution.

[0139] If the products scheduled for production in the production plan are products that have never been produced before, and one or more of their steel grade name, contract thickness, contract width, or elongation setting values ​​are unknown, making it impossible to accurately query them in historical production parameters, then a pre-driven model is determined based on the steel grade, tap mark, thickness, width, elongation, roll roughness, roll diameter, zinc coating type, strip roughness setting, and production season from the historical production parameters.

[0140] In this galvanizing production line, there are a total of 56,727 data samples of historical production parameters. 80% of the data (45,382 data samples) were selected as the training set and 20% of the data (11,345 data samples) were selected as the validation set.

[0141] The input features of the pre-driven model are external grade, strip mark, contract thickness, contract width, elongation setting, roll diameter, strip roughness setting, production season (determined based on the start time of production), product target roughness, and roll length. The output features of the pre-driven model are rolling force, roll bending force, inlet tension, and outlet tension.

[0142] Taking rolling force as an example, the prediction accuracy of the pre-driven model is as follows: Figure 2 As shown in the figure, the horizontal axis represents production time, and the vertical axis represents the average rolling force. The blue curve represents the actual rolling force data during the actual production process, while the red curve represents the setpoint rolling force data output by the pre-driven model. The coefficient of determination and mean absolute percentage error of this pre-driven model are shown in Table 4.

[0143] Table 4 Prediction Accuracy Table

[0144] <![CDATA[R 2 ]]> 0.987 0.988 0.996 0.996 MAPE 2.5％ 5.0％ 1.3％ 1.1％

[0145] A roughness prediction model was established using the same historical sample set as the pre-driven model, divided into training and validation sets. The input features of the roughness prediction model included external grade, tap mark, contract thickness, contract width, elongation setpoint, roll diameter, roll surface roughness, season, target product roughness, roll length, and the rolling force, bending force, inlet tension, and outlet tension output by the pre-driven model. The output feature of the roughness prediction model was the initial value of the strip roughness.

[0146] The prediction accuracy of the roughness prediction model is as follows: Figure 3 As shown, the horizontal axis represents production time, and the vertical axis represents roughness. The blue curve represents the actual roughness data during the actual production process, while the red curve represents the predicted roughness data output by the roughness prediction model. The coefficient of determination R of this roughness prediction model is... 2 =0.804, and the mean absolute percentage error (MAPE) is 6.24%, indicating that it can effectively explain 80.4% of the roughness fluctuations, and the average relative error is controlled within 6.24%.

[0147] It is evident that the dual-layer architecture of the pre-driving model and the roughness prediction model based on historical production parameters can automatically determine the optimal rolling parameters and accurately predict the surface roughness of the strip steel. This enables the automatic setting and distribution of parameters such as the rolling force of the galvanized wire finishing mill, reducing the rate of manual intervention, minimizing roll changes caused by roughness discrepancies, and stabilizing the control of product quality on the production line.

[0148] Example 2

[0149] After applying the pre-adjustment method for a bright finisher based on strip roughness provided in this application, the predicted rolling parameters of a galvanizing production line are as follows: Figure 4 As shown. After adjusting the finishing mill using the predicted rolling parameters as the target driving parameters, Figure 5a , Figure 5b , Figure 5c and Figure 5d The chart shows the matching between the actual and predicted values ​​of rolling force, bending roll force, inlet tension, and outlet tension. The horizontal axis represents the corresponding data type, and the vertical axis represents the numerical values. The orange curve represents the predicted value, and the blue curve represents the actual value. The predicted values ​​basically match the actual values. It is evident that the pre-adjustment method for a finishing mill based on strip roughness provided in this application accurately predicts the rolling parameters of the finishing mill.

[0150] In summary, this application provides a pre-adjustment method for a finishing machine based on strip roughness, comprising: determining a pre-driving model based on historical production parameters of a first parameter type; inputting the first production parameters of the product to be produced, which match the first parameter type, into the pre-driving model to obtain pre-driving parameters and a target value for strip roughness; determining a roughness prediction model based on historical production parameters of a second parameter type; determining an initial value for strip roughness based on the pre-driving parameters and the roughness prediction model; determining a roughness error based on the initial value and the target value for strip roughness; determining the initial value for strip roughness as a predicted value for strip roughness if the roughness error meets a preset error standard; determining a target driving parameter based on the predicted value for strip roughness, the first production parameters, and the pre-driving model; and adjusting the finishing machine based on the target driving parameter.

[0151] It is evident that the dual-layer architecture of the pre-driven model and the roughness prediction model, based on historical production parameters, can automatically determine the optimal rolling parameters and accurately predict the surface roughness of the strip. Furthermore, by incorporating roughness error feedback, it corrects the roughness fluctuation problem caused by parameter lag, enabling intelligent pre-setting and online fine-tuning of key parameters such as rolling force and tension. This significantly improves the process response speed and production continuity of the finishing mill, while reducing equipment wear and energy consumption losses caused by repeated trial and error.

[0152] Based on the same inventive concept, embodiments of this application also provide, as follows: Figure 6 The pre-adjustment device for a finishing machine based on strip roughness, as shown, includes:

[0153] The pre-driving parameter determination module 61 is used to determine the pre-driving model based on the historical production parameters of the first parameter type, input the first production parameters of the product to be produced that match the first parameter type into the pre-driving model, and obtain the pre-driving parameters and the target value of strip roughness.

[0154] The strip roughness initial value determination module 62 is used to determine the roughness prediction model based on the historical production parameters of the second parameter type, and to determine the initial value of the strip roughness based on the pre-driving parameters and the roughness prediction model.

[0155] The roughness error determination module 63 is used to determine the roughness error based on the initial value of the strip roughness and the target value of the strip roughness.

[0156] The strip roughness prediction value determination module 64 is used to determine the initial value of strip roughness as the predicted value of strip roughness when the roughness error meets the preset error standard.

[0157] The adjustment module 65 is used to determine the target driving parameters based on the predicted strip roughness value, the first production parameters and the pre-drive model, and to adjust the finishing machine according to the target driving parameters.

[0158] Furthermore, the device also includes a pre-driving model determination module for:

[0159] The initial model is determined based on the historical production parameters of the first parameter category that matches the product to be produced;

[0160] Based on the initial predicted values ​​output by the initial model and the historical production parameters that match the initial predicted values, determine the coefficient of determination and the mean absolute percentage error.

[0161] If the coefficient of determination and the mean absolute percentage error both meet the preset model criteria, the initial model is determined as the pre-driven model.

[0162] Furthermore, the device also includes a pre-drive parameter adjustment module for:

[0163] When the roughness error and the preset error standard are in the first preset relationship, the pre-driving parameters are positively adjusted according to the preset step size, and the process returns to the step of determining the initial value of strip roughness based on the pre-driving parameters and the roughness prediction model.

[0164] If the roughness error and the preset error standard have a second preset relationship, the pre-driving parameters are negatively adjusted according to the preset step size, and the process returns to the step of determining the initial value of the strip roughness based on the pre-driving parameters and the roughness prediction model.

[0165] Furthermore, the device also includes a strip roughness target value update module, used for:

[0166] If the pre-driving parameter exceeds the preset upper limit, update the strip roughness target value according to the strip roughness target value and the preset roughness lower limit value, and return to the step of determining the roughness error according to the initial strip roughness value and the strip roughness target value.

[0167] If the pre-driving parameter is lower than the preset lower limit, the strip roughness target value is updated according to the strip roughness target value and the preset roughness upper limit value, and the process returns to the step of determining the roughness error based on the initial strip roughness value and the strip roughness target value.

[0168] Furthermore, the device also includes an update frequency adjustment module for:

[0169] The cumulative number of updates to the target strip roughness value is calculated. If the number of updates exceeds the preset number standard, the roll is replaced and the process is returned to execute the steps of determining the pre-drive model based on the historical production parameters of the first parameter type, inputting the first production parameters of the product to be produced that match the first parameter type into the pre-drive model, and obtaining the pre-drive parameters and the target strip roughness value.

[0170] Furthermore, the device also includes a target drive parameter determination module, used for:

[0171] Based on the predicted strip roughness value, the first production parameter, and the pre-driving model, determine the predicted driving parameters corresponding to at least two consecutive production times.

[0172] If at least one set of two adjacent predicted driving parameters meet a preset difference standard, the predicted driving parameter is determined as the target driving parameter.

[0173] If at least one set of adjacent prediction driving parameters does not meet the preset difference standard, adjust the pre-driving parameters.

[0174] Furthermore, the device also includes a matching module for:

[0175] Determine whether the known parameter types of the product to be produced match the parameter types included in the historical production parameters.

[0176] Based on the same inventive concept, embodiments of this application also provide, as follows: Figure 7 An electronic device shown includes:

[0177] Processor 71;

[0178] Memory 72 is used to store executable instructions of processor 71;

[0179] The processor 71 is configured to execute a pre-conditioning method for finishing the strip based on the roughness of the strip, as described above.

[0180] Based on the same inventive concept, this application also provides a non-transitory computer-readable storage medium, which, when the instructions in the storage medium are executed by the processor 71 of an electronic device, enables the electronic device to perform a pre-adjustment method for finishing a strip based on the roughness of the strip as described above.

[0181] Since the electronic device described in this embodiment is an electronic device used to implement the information processing method in the embodiments of this application, those skilled in the art can understand the specific implementation methods and various variations of the electronic device in this embodiment based on the information processing method described in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. Any electronic device used by those skilled in the art to implement the information processing method in the embodiments of this application falls within the scope of protection of this application.

[0182] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0183] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0184] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0185] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0186] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0187] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A pre-adjustment method for a finishing machine based on strip roughness, characterized in that, include: The pre-driving model is determined based on the historical production parameters of the first parameter type. The first production parameters of the product to be produced that match the first parameter type are input into the pre-driving model to obtain the pre-driving parameters and the target value of strip roughness. The roughness prediction model is determined based on the historical production parameters of the second parameter type, and the initial value of strip roughness is determined based on the pre-driving parameters and the roughness prediction model. The roughness error is determined based on the initial roughness value of the strip and the target roughness value of the strip; If the roughness error meets the preset error standard, the initial value of the strip roughness is determined as the predicted value of the strip roughness. Based on the predicted strip roughness value, the first production parameter, and the pre-driven model, the target driving parameter is determined, and the finishing machine is adjusted according to the target driving parameter; The first parameter types include at least one of the following: steel grade, steel exit mark, thickness, width, elongation, roll roughness, roll diameter, zinc coating type, strip set roughness, and production season; The pre-drive parameters include at least one of the following: rolling force, bending force, inlet tension, and outlet tension; The second parameter includes at least one of the following: steel grade, roll diameter, roll surface roughness, thickness, width, elongation, rolling force, roll bending force, inlet tension, outlet tension, and rolling kilometers. The step of determining the pre-driving model based on historical production parameters of the first parameter type includes: An initial model is determined based on the historical production parameters of the first parameter category that match the product to be produced; Based on the initial predicted values ​​output by the initial model and the historical production parameters that match the initial predicted values, determine the coefficient of determination and the mean absolute percentage error. If both the determination coefficient and the mean absolute percentage error meet the preset model criteria, the initial model is determined as the pre-driven model. Before determining the pre-driving model based on historical production parameters of the first parameter type, the method further includes: Determine whether the known parameter types of the product to be produced match the parameter types included in the historical production parameters.

2. The pre-adjustment method for finishing machines based on strip roughness as described in claim 1, characterized in that, If the roughness error does not meet the preset error standard, the method further includes: If the roughness error and the preset error standard satisfy the first preset relationship, the pre-driving parameter is positively adjusted according to the first preset step size, and the process returns to the step of determining the initial value of strip roughness based on the pre-driving parameter and the roughness prediction model. If the roughness error satisfies the second preset relationship with the preset error standard, the pre-driving parameter is negatively adjusted according to the second preset step size, and the process returns to the step of determining the initial value of strip roughness based on the pre-driving parameter and the roughness prediction model.

3. The pre-adjustment method for finishing machines based on strip roughness as described in claim 2, characterized in that, If the roughness error does not meet the preset error standard, the method further includes: If the pre-driving parameter exceeds the preset upper limit value, the target roughness value of the strip steel is updated according to the target roughness value of the strip steel and the preset lower limit value of the roughness, and the process returns to the step of determining the roughness error based on the initial roughness value of the strip steel and the target roughness value of the strip steel. If the pre-driving parameter is lower than the preset lower limit, the target roughness value of the strip is updated according to the target roughness value of the strip and the preset upper limit value of the roughness, and the process returns to the step of determining the roughness error based on the initial roughness value of the strip and the target roughness value of the strip.

4. The pre-adjustment method for finishing machines based on strip roughness as described in claim 3, characterized in that, If the pre-driving parameter exceeds the preset upper limit or is lower than the preset lower limit, the method further includes: The number of times the target strip roughness value is updated is accumulated. If the number of updates exceeds the preset number standard, the roll is replaced and the process is returned to execute the steps of determining the pre-drive model based on the historical production parameters of the first parameter type, inputting the first production parameters of the product to be produced that match the first parameter type into the pre-drive model, and obtaining the pre-drive parameters and the target strip roughness value.

5. The pre-adjustment method for finishing machines based on strip roughness as described in claim 1, characterized in that, The step of determining the target driving parameters based on the predicted strip roughness value, the first production parameter, and the pre-driving model includes: Based on the predicted strip roughness value, the first production parameter, and the pre-driving model, determine the predicted driving parameters corresponding to at least two consecutive production times. If at least one set of two adjacent predicted driving parameters meet a preset difference standard, the predicted driving parameter is determined as the target driving parameter; If the preset difference standard is not met between at least one set of two adjacent predicted driving parameters, the pre-driving parameters are adjusted.

6. A pre-adjustment device for a finishing machine based on strip roughness, characterized in that, include: The pre-driving parameter determination module is used to determine the pre-driving model based on the historical production parameters of the first parameter type, and input the first production parameters of the product to be produced that match the first parameter type into the pre-driving model to obtain the pre-driving parameters and the target value of strip roughness. The strip roughness initial value determination module is used to determine the roughness prediction model based on the historical production parameters of the second parameter type, and to determine the initial value of strip roughness based on the pre-driving parameters and the roughness prediction model. A roughness error determination module is used to determine the roughness error based on the initial roughness value of the strip and the target roughness value of the strip; The strip roughness prediction value determination module is used to determine the initial value of the strip roughness as the predicted value of the strip roughness when the roughness error meets the preset error standard. The adjustment module is used to determine the target driving parameters based on the predicted strip roughness value, the first production parameters, and the pre-driving model, and to adjust the finishing machine according to the target driving parameters; The first parameter types include at least one of the following: steel grade, steel exit mark, thickness, width, elongation, roll roughness, roll diameter, zinc coating type, strip set roughness, and production season; The pre-drive parameters include at least one of the following: rolling force, bending force, inlet tension, and outlet tension; The second parameter includes at least one of the following: steel grade, roll diameter, roll surface roughness, thickness, width, elongation, rolling force, roll bending force, inlet tension, outlet tension, and rolling kilometers. It also includes a pre-driven model determination module, used for: An initial model is determined based on the historical production parameters of the first parameter category that match the product to be produced; Based on the initial predicted values ​​output by the initial model and the historical production parameters that match the initial predicted values, determine the coefficient of determination and the mean absolute percentage error. If both the determination coefficient and the mean absolute percentage error meet the preset model criteria, the initial model is determined as the pre-driven model. It also includes a matching module for: Determine whether the known parameter types of the product to be produced match the parameter types included in the historical production parameters.

7. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute a pre-conditioning method for finishing a strip based on the roughness of a strip as described in any one of claims 1 to 5.

Images