A model structure optimization method and device for a five-stand rolling mill

Abstract

The application discloses a model structure optimization method for a five-stand tandem rolling mill, and belongs to the technical field of rolling processes. The method comprises the following steps: obtaining a rolling model parameter table and a friction coefficient model parameter table; performing capacity expansion processing on the rolling model parameter table to generate a target rolling model parameter table; adding a steel grade search field and a thickness search field in the friction coefficient model parameter table to generate a target friction coefficient model parameter table; constructing a preset load distribution table indexed according to steel grades and thicknesses; and deploying the above three tables to a control system as a target model structure. The application improves the process parameter setting accuracy of different steel grades and thickness specifications, reduces the manual intervention rate of load distribution, suppresses the rolling mill vibration of thin specifications, and improves the line stability and product quality by optimizing the model constant table classification, adding the steel grade and thickness dimensions of the friction coefficient model, and introducing the preset load distribution table to work with the original load distribution model.

Description

Technical Field

[0001] This application relates to the field of rolling process technology, and in particular to a method and apparatus for optimizing the model structure of a five-strand rolling mill. Background Technology

[0002] With the continuous expansion of steel product types and the decreasing thickness of specifications, the traditional five-stand rolling mill model structure suffers from problems when dealing with multi-variety, thin-gauge production. These problems include insufficient precision in setting rolling process parameters, inability to match friction coefficients to individual product specifications, and reliance on manual experience for load distribution. Consequently, rolling stability is poor and product quality fluctuates significantly. Therefore, there is an urgent need for a model structure optimization method for five-stand rolling mills to solve the aforementioned technical problems. Summary of the Invention

[0003] The summary section introduces a series of simplified concepts, which will be further explained in detail in the detailed description section. This summary section is not intended to limit the key and essential technical features of the claimed technical solutions, nor is it intended to determine the scope of protection of the claimed technical solutions.

[0004] In a first aspect, this application provides a method for optimizing the model structure of a five-stand rolling mill, comprising: Obtain the rolling model parameter table and friction coefficient model parameter table from the control system of the five-stand rolling mill. The rolling model parameter table is used to store rolling process parameters indexed by steel grade group spacing, thickness group spacing, and width group spacing. The friction coefficient model parameter table is used to store friction coefficient parameters indexed by roll type, emulsion mode, and stand number. The rolling model parameter table is expanded to generate the target rolling model parameter table; Add steel type and thickness search fields to the friction coefficient model parameter table to generate the target friction coefficient model parameter table; Construct a preset load allocation table indexed by steel type and thickness, wherein the preset load allocation table is used to preferentially set the load allocation of a specified stand during the rolling process using the preset load allocation value, and the load allocation value of the remaining stands is determined based on the load allocation model; The target rolling model parameter table, the target friction coefficient model parameter table, and the preset load distribution table are deployed to the control system of the five-strand rolling mill as the target model structure of the five-strand rolling mill.

[0005] In some embodiments, expanding the rolling model parameter table to generate a target rolling model parameter table includes: The steel grade interval in the rolling model parameter table is expanded from the original position to the first preset steel grade in the smooth roll mode and the second preset steel grade in the rough roll mode, where the original position is 16. The thickness interval in the rolling model parameter table is expanded from the original level to the first preset thickness level in the smooth roll mode and the second preset thickness level in the rough roll mode. The width group spacing in the rolling model parameter table is re-divided to generate an optimized width group spacing. Based on the amplified steel grade group spacing, the amplified thickness group spacing, and the optimized width group spacing, the target rolling model parameter table is generated.

[0006] In some implementations, the first preset steel grade level is 96, the second preset steel grade level is 96, the first preset thickness level is 32, and the second preset thickness level is 32.

[0007] In some implementations, it also includes: Obtain the specification information of the strip steel to be rolled, wherein the specification information includes steel grade information, thickness information and width information; Based on the specification information, the surface condition of the target roll corresponding to the strip to be rolled is determined, wherein the surface condition of the target roll is a smooth roll condition or a rough roll condition; Based on the surface condition of the target roll, a steel grade retrieval operation is performed from the target rolling model parameter table to determine the target steel grade corresponding to the strip to be rolled.

[0008] In some implementations, the step of performing a steel grade retrieval operation from the target rolling model parameter table based on the surface condition of the target roll to determine the target steel grade corresponding to the strip to be rolled includes: Using the surface condition of the target roll as the first search condition, the target mode branch in the target rolling model parameter table is determined; Under the target mode branch, determine whether there is a steel grade code corresponding to the strip to be rolled; When the steel grade code exists, the steel grade code is used as the second search condition to query the steel grade corresponding to the steel grade code from the target rolling model parameter table, and the queried steel grade is determined as the target steel grade. The steel grade code is a unique identifier constructed based on the steel type information, the thickness information and the width information, and is used to query the unique corresponding steel grade. When the steel grade code does not exist, the tap mark and yield strength of the strip to be rolled are obtained, and the tap mark and yield strength are used as the third search conditions to query the corresponding steel grade from the target rolling model parameter table, and the queried steel grade is determined as the target steel grade; Query the rolling process parameters corresponding to the target steel grade from the target rolling model parameter table.

[0009] In some embodiments, the step of adding a steel grade search field and a thickness search field to the friction coefficient model parameter table to generate a target friction coefficient model parameter table includes: Construct a storage structure for friction coefficient parameters with roll type, steel grade, thickness, stand number, and emulsion mode as joint indexes; For each set of index values ​​in the composite index, obtain the corresponding friction coefficient parameter; The friction coefficient parameters are written into the friction coefficient parameter storage structure to generate the target friction coefficient model parameter table.

[0010] In some implementations, load allocation values ​​are determined based on a load allocation model, including: Obtain the initial load distribution values ​​for each rack; The initial value of the load distribution is verified to generate a verification result. The verification process includes rolling force limit verification, rolling force balance verification between the first stand and the second stand, and power balance verification between the second stand and the fifth stand. Based on the verification results, regression calculations are performed on the initial load distribution value using historical strip rolling data to generate an optimized load distribution value, and the optimized load distribution value is determined as the load distribution value for each stand.

[0011] In some embodiments, the number of designated stands is determined based on the surface condition of the target roll, wherein when the surface condition of the target roll is a smooth roll, the number of designated stands is less than or equal to 4; and when the surface condition of the target roll is a rough roll, the number of designated stands is less than or equal to 3.

[0012] In some implementations, it also includes: Based on the steel grade information and the thickness information, the maximum rolling speed of the five-strand rolling mill is determined.

[0013] Secondly, this application proposes a model structure optimization device for a five-stand rolling mill, comprising: The parameter acquisition unit is used to acquire the rolling model parameter table and the friction coefficient model parameter table in the control system of the five-stand mill. The rolling model parameter table is used to store rolling process parameters indexed by steel grade group spacing, thickness group spacing and width group spacing. The friction coefficient model parameter table is used to store friction coefficient parameters indexed by roll type, emulsion mode and stand number. An expansion processing unit is used to expand the rolling model parameter table to generate a target rolling model parameter table. The retrieval expansion unit is used to add steel type retrieval field and thickness retrieval field to the friction coefficient model parameter table to generate the target friction coefficient model parameter table. A rolling distribution unit is used to construct a preset load distribution table indexed by steel type and thickness. The preset load distribution table is used to prioritize the load distribution setting of a specified stand during the rolling process, and the load distribution value of the remaining stands is determined based on the load distribution model. The model deployment unit is used to deploy the target rolling model parameter table, the target friction coefficient model parameter table, and the preset load distribution table to the control system of the five-strand rolling mill, as the target model structure of the five-strand rolling mill.

[0014] Thirdly, an electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program stored in the memory to implement the steps of the model structure optimization method for a five-roll mill described in any of the first aspects.

[0015] Fourthly, this application also proposes a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of the model structure optimization method for a five-strand rolling mill as described in any of the first aspects.

[0016] In summary, the model structure optimization method for a five-stand rolling mill provided in this application obtains the rolling model parameter table and friction coefficient model parameter table in the five-stand rolling mill control system, which is the basic data organization method of the existing model structure. The rolling model parameter table is expanded to generate a target rolling model parameter table, extending the original steel grade group spacing and thickness group spacing. This allows products of different steel grades and thicknesses, which were previously forced to share the same parameter range due to insufficient grading, to obtain a more refined process parameter index, improving the accuracy of rolling process parameter settings and avoiding deviations in rolling force and tension settings caused by parameter mixing. Steel grade and thickness search fields are added to the friction coefficient model parameter table to generate a target friction coefficient model parameter table. This ensures that the friction coefficient model no longer relies solely on roll type, emulsion mode, and stand number, but instead configures model parameters separately according to the actual rolling characteristics of different steel grades and thicknesses. This allows for the calculation of rolling force and forward slip value for each stand, improving the strip shape and thickness control quality. A preset load allocation table indexed by steel grade and thickness was constructed. During the rolling process, this table prioritizes the load allocation settings for designated stands using the preset load allocation values, while the load allocation values ​​for the remaining stands are still determined based on the original load allocation model. This approach allows process engineers to directly apply their experience-based allocation values ​​for specific steel grades and thicknesses to critical stands (such as stands containing thin-gauge tin-plated substrates prone to vibration), avoiding frequent manual intervention by operators. It also retains the power balance calculation capability for non-designated stands, effectively reducing process risks such as mill vibration and scratches, and improving the stability and automation level of the rolling process. Finally, the generated target rolling model parameter table, target friction coefficient model parameter table, and preset load allocation table were deployed to the control system of the five-stand mill as the target model structure, enabling the above optimizations to be applied to actual production and achieving high-quality and stable rolling of multiple varieties and wide specifications of products. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A schematic flowchart of a model structure optimization method for a five-stand rolling mill provided in this application embodiment; Figure 2 A schematic diagram of the thickness group spacing expansion of the rolling model parameter table provided in the embodiments of this application; Figure 3 A schematic diagram of the expanded steel grade group spacing in the rolling model parameter table provided in the embodiments of this application; Figure 4A schematic diagram of a model structure optimization device for a five-strand rolling mill provided in this application embodiment; Figure 5 This is a schematic diagram of an electronic device structure for optimizing the model structure of a five-roll mill, provided as an embodiment of this application. Detailed Implementation

[0018] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them.

[0019] This application relates to a model structure optimization method for a five-strand rolling mill. In the cold rolling production process of the steel industry, a five-strand rolling mill is used to roll hot-rolled strip steel to target specifications. With the increasing market demand for high-precision, thin-gauge, and multi-variety products, production lines need to frequently switch between producing different steel grades such as plain plates, tin-plated substrates, and secondary cold-rolled materials, with thickness specifications covering a wide range from 0.09mm to 1.0mm. However, the original model structure of the five-strand rolling mill was mainly designed for a large-batch, low-variety production mode. Its internally stored rolling process parameters, friction coefficient parameters, and load distribution logic are all based on a limited and fixed mathematical model. When the production line shifts to multi-variety, small-batch, and thin-gauge production, the shortcomings of the original model structure in terms of parameter setting accuracy and process adaptation flexibility gradually become apparent. Specifically, this manifests as: sharing the same parameter level for different steel grades and thicknesses leading to process setting deviations; the inability to match the friction coefficient according to the individual product type affecting the rolling force and forward slip calculation accuracy; excessive reliance on manual intervention based on operator experience for load distribution; and the maximum rolling speed only considering equipment capacity while ignoring the vibration risks of different products. Therefore, this application aims to optimize the model structure of a five-strand rolling mill to meet the high-quality and stable rolling requirements of multi-variety, thin-gauge products.

[0020] To more clearly illustrate the technical solution of this application, some terms involved in this application are defined and explained below.

[0021] The rolling model parameter table in this application refers to a database table pre-stored in the five-stand rolling mill control system. This table uses steel grade group spacing, thickness group spacing, and width group spacing as indexes to store rolling process parameters corresponding to different specifications of products. These rolling process parameters include, but are not limited to, rolling force setpoints, tension setpoints, and reduction allocation values. In existing technologies, the steel grade group spacing and thickness group spacing in this table are both 16 levels, resulting in a relatively coarse grading granularity. This forces multiple products of different steel grades or thicknesses to share the same set of process parameters.

[0022] The friction coefficient model parameter table in this application refers to a data table used in the five-stand rolling mill control system to store parameters related to the friction coefficient model. This table uses roll type, emulsion mode, and stand number as indexes to store friction coefficient parameter values ​​under different operating conditions. The friction coefficient model is used to calculate the friction characteristics of the contact surface between the rolls and the strip during rolling, and this parameter affects the calculation results of rolling force and forward slip value. In the prior art, this table does not include steel grade and thickness dimensions, resulting in products of different steel grades and thicknesses using the same set of friction coefficient parameters.

[0023] The capacity expansion process in this application refers to increasing the number of increments in the steel grade interval, thickness interval, and width interval in the rolling model parameter table. Specifically, the steel grade interval is expanded from the original 16 increments to 96 increments in both the smooth roll mode and the rough roll mode; the thickness interval is expanded from the original 16 increments to 32 increments in both the smooth roll mode and the rough roll mode; and the width interval is reclassified. Through this capacity expansion process, different products that were originally grouped into the same increment are assigned to their respective more appropriate parameter increments.

[0024] The smooth rolls and rough rolls in this application refer to two types of rolls with different surface conditions. Smooth rolls have a smooth surface and are mainly used to produce products with high surface quality requirements; rough rolls have a certain surface roughness and are mainly used to increase friction during the rolling process to meet specific process requirements. Due to the differences in friction characteristics and process parameters between the two roll types, this application clearly distinguishes between the smooth roll mode and the rough roll mode in the model structure optimization and configures independent parameter levels for each.

[0025] The steel grade code in this application refers to a unique identifier constructed based on steel grade information, thickness information, and width information, used to uniquely identify a specific combination of product specifications. In actual production, a steel tapping mark may correspond to multiple different reduction rates or process paths, thus corresponding to multiple steel grade codes, and each steel grade code uniquely corresponds to a steel grade. This application uses the steel grade code as the second search condition for steel grade classification retrieval to achieve parameter matching at the specification level.

[0026] The tapping mark in this application refers to a code used to identify the furnace or batch information. In the original search logic, the tapping mark and yield strength were used together as search conditions. However, since one tapping mark may correspond to multiple different reduction rates or steel grade codes, the specific product specifications cannot be identified by the tapping mark alone. Therefore, this application downgrades the tapping mark to a third search condition, which is only activated when the steel grade code is missing.

[0027] The yield strength in this application refers to the critical stress value when strip steel undergoes plastic deformation, and is a performance indicator reflecting the hardness or softness of the material. In the original search logic, yield strength and steel tapping mark were used together as search conditions; in the search logic of this application, yield strength is used as a component of the third search condition to assist in determining the steel grade classification when the steel grade code is missing.

[0028] The preset load allocation table in this application refers to a newly added data table. This table uses steel grade and thickness as indexes to store pre-set, process-verified load allocation experience values ​​for specific product specifications during the rolling process. The function of this table is to prioritize the preset load allocation values ​​in the table for designated stands when rolling special product specifications, while the remaining stands are still calculated based on the original load allocation model, thereby realizing a hybrid control strategy that combines experience-based priority with model calculation.

[0029] The load distribution value in this application refers to the proportion of reduction or rolling force allocated to each stand during the five-stand rolling process. Reasonable load distribution is a key factor in ensuring rolling stability and strip shape. In this application, the load distribution value comes from two sources: for a specific stand, empirical values ​​are preferentially obtained from a pre-set load distribution table; for the remaining stands, the value is determined by the load distribution model through verification and regression calculations.

[0030] The load distribution model in this application refers to the mathematical model used in the five-stand rolling mill control system to calculate the load distribution values ​​for each stand. The calculation process includes: obtaining the initial load distribution values ​​for each stand; verifying the initial load distribution values, including verification of rolling force limits, rolling force balance between the first and second stands, and power balance between the second to fifth stands; and based on the verification results, performing regression calculations on the initial load distribution values ​​using historical strip rolling data to generate optimized load distribution values. The goal of this model is to output the optimal load distribution scheme while ensuring equipment safety and power balance.

[0031] The designated racks in this application refer to those racks that are preferentially assigned preset load allocation values ​​when load allocation is performed using a preset load allocation table. The number of designated racks is not fixed, but is determined based on the surface condition of the rolls used in current production: when using the smooth roll mode, the number of designated racks is less than or equal to four; when using the rough roll mode, the number of designated racks is less than or equal to three. The quantity constraint is determined based on the respective process characteristics of smooth rolls and rough rolls, as well as production experience, aiming to both intervene in critical racks using empirical values ​​and reserve at least one rack for calculation by the load allocation model to cope with incoming material fluctuations.

[0032] The rolling force limit verification in this application refers to the step of verifying whether the rolling force of each stand obtained by substituting the initial load distribution value into the rolling force calculation formula (such as the classic Hill formula, Bland-Ford formula, or rolling force model based on machine learning) is within the safe range allowed by the mechanical equipment.

[0033] The rolling force balance verification between the first stand and the second stand in this application refers to the step of verifying the ratio of rolling force between the first stand and the second stand. Its purpose is to limit the rolling force of the first stand to be too large to prevent slippage or strip breakage.

[0034] The power balance verification between the second to fifth stands in this application refers to the step of verifying whether the power distribution of the motors between the second to fifth stands is balanced. Its purpose is to fully utilize the efficiency of the motor power of each stand and ensure the rolling speed and production rhythm.

[0035] The historical strip rolling data in this application refers to historical records collected from actual production processes, including parameters such as rolling force, tension, thickness, and speed for strips of different steel grades and specifications. This application utilizes this historical data to perform regression calculations on the initial values ​​of load distribution, enabling the model to be continuously optimized based on actual production results, gradually approaching the optimal load distribution parameters.

[0036] The maximum rolling speed described in this application refers to the highest permissible rolling speed of a five-strand rolling mill when rolling a specific product. In existing technologies, the maximum rolling speed is determined solely based on equipment capacity factors such as power and rolling rhythm, without considering the process characteristics of different steel grades and thicknesses. This leads to mill vibrations easily occurring when thin-gauge products are rolled at high speeds. This application, based on steel grade and thickness information, retrieves a matching speed limit value from a preset speed limit table to automatically limit the maximum rolling speed.

[0037] The preset speed limit table in this application refers to a newly added data table. This table, indexed by steel grade information and thickness range, stores the preset maximum rolling speed safety limits for different product specifications. The speed limits in this table are determined based on analysis of historical production data. Specifically, for thin-gauge tin-plated substrate products prone to mill vibration, safe speed values ​​below the equipment's capacity limit are set. This automatically limits the rolling speed while ensuring stable production line operation, avoiding the risk of strip breakage due to untimely or inaccurate manual settings by operators.

[0038] Please see Figure 1 This is a schematic diagram of a model structure optimization method for a five-stand rolling mill provided in an embodiment of this application, which specifically includes: S110. Obtain the rolling model parameter table and friction coefficient model parameter table from the control system of the five-stand rolling mill. The rolling model parameter table is used to store rolling process parameters indexed by steel grade group spacing, thickness group spacing and width group spacing. The friction coefficient model parameter table is used to store friction coefficient parameters indexed by roll type, emulsion mode and stand number. For example, the existing rolling model parameter table and friction coefficient model parameter table in the five-stand rolling mill control system are retrieved. The rolling model parameter table stores rolling process parameters corresponding to different specifications of strip steel, such as rolling force setpoints and tension setpoints, according to an index structure of steel grade group spacing, thickness group spacing, and width group spacing. These parameters are the data source for the model to set the rolling parameters. The friction coefficient model parameter table stores the parameters required for friction coefficient calculation, which are used for calculating rolling force and forward slip value, according to an index structure of roll type, emulsion mode, and stand number.

[0039] S120. Expand the rolling model parameter table to generate the target rolling model parameter table; For example, the obtained rolling model parameter table is expanded by extending the original steel grade group spacing and thickness group spacing in both smooth roll mode and rough roll mode. At the same time, the width group spacing is re-divided so that different steel grades, different thicknesses and different widths that were originally forced to share the same parameter level due to insufficient number of grades are assigned to independent index positions, thereby generating a target rolling model parameter table with a more refined grading structure.

[0040] S130. Add steel type search field and thickness search field to the friction coefficient model parameter table to generate the target friction coefficient model parameter table; For example, the friction coefficient model is configured with corresponding friction coefficient parameters according to different steel grades and thicknesses, changing the original model's single indexing method that only relies on roll type, emulsion mode and stand number, and providing parameter settings that match the product specifications for calculating the rolling force and forward slip value of each stand.

[0041] S140. Construct a preset load allocation table indexed by steel type and thickness. The preset load allocation table is used to prioritize the load allocation setting of the specified stand during the rolling process, and the load allocation value of the remaining stand is determined based on the load allocation model. For example, when constructing the preset load allocation table, empirically validated load allocation values ​​are pre-stored for products of different steel grades and thicknesses. When the five-stand rolling mill begins rolling a specific specification of strip, the table is queried based on the current strip grade and thickness. If a corresponding preset load allocation value is found, this value is preferentially used for the load allocation setting of the designated stand. For the remaining stands not designated, the original load allocation model is used to determine their load allocation values ​​through power balance calculations. This design allows thin-gauge products prone to mill vibration or scratches to use empirical load allocation schemes, avoiding manual adjustment of the reduction amount of each stand by operators during each production run. At the same time, it retains the adaptive calculation capability of the load allocation model for the remaining stands to cope with real-time fluctuations in incoming and outgoing material thicknesses. This improves the automation level of load allocation while ensuring the stability of the rolling process and the quality of the product shape.

[0042] S150. Deploy the target rolling model parameter table, the target friction coefficient model parameter table, and the preset load distribution table to the control system of the five-strand rolling mill as the target model structure of the five-strand rolling mill.

[0043] For example, after constructing the target rolling model parameter table, the target friction coefficient model parameter table, and the preset load allocation table, these three tables are deployed to the control system of the five-stand rolling mill, replacing or supplementing the original model structure to form a target model structure supporting the production of multiple varieties and thin specifications. In the deployed target model structure, the expanded rolling model parameter table provides more refined process parameter indexes for strips of different steel grades, thicknesses, and widths, avoiding setting deviations caused by different specifications of products sharing the same gear; the friction coefficient model parameter table, with added steel grade and thickness search fields, ensures that the configuration of the friction coefficient matches the specific variety of strip to be rolled, improving the calculation accuracy of rolling force and forward slip; the preset load allocation table prioritizes assigning empirical allocation values ​​to designated stands when rolling special specification products, while the load allocation for other stands continues to be calculated by the original load allocation model. Through the coordinated deployment of the above three tables, the control system of the five-stand rolling mill obtains a fully optimized model structure from parameter indexing and friction coefficient matching to load allocation, laying the foundation for control and operation in the actual rolling process.

[0044] In summary, the model structure optimization method for a five-stand rolling mill provided in this application obtains the rolling model parameter table and friction coefficient model parameter table in the five-stand rolling mill control system, which is the basic data organization method of the existing model structure. The rolling model parameter table is expanded to generate a target rolling model parameter table, extending the original steel grade group spacing and thickness group spacing. This allows products of different steel grades and thicknesses, which were previously forced to share the same parameter level due to insufficient grading, to obtain a more refined process parameter index, improving the accuracy of rolling process parameter settings and avoiding deviations in rolling force and tension settings caused by parameter mixing. A steel grade search field and a thickness search field are added to the friction coefficient model parameter table to generate a target friction coefficient model parameter table. This ensures that the friction coefficient model no longer relies solely on roll type, emulsion mode, and stand number, but instead configures model parameters separately according to the actual rolling characteristics of different steel grades and thicknesses, thereby calculating the rolling force and forward slip value of each stand and improving the strip shape and thickness control quality. A preset load allocation table indexed by steel grade and thickness was constructed. During the rolling process, this table prioritizes the load allocation settings for designated stands using the preset load allocation values, while the load allocation values ​​for the remaining stands are still determined based on the original load allocation model. This approach allows process engineers to directly apply their experience-based allocation values ​​for specific steel grades and thicknesses to critical stands (such as stands containing thin-gauge tin-plated substrates prone to vibration), avoiding frequent manual intervention by operators. It also retains the power balance calculation capability for non-designated stands, effectively reducing process risks such as mill vibration and scratches, and improving the stability and automation level of the rolling process. Finally, the generated target rolling model parameter table, target friction coefficient model parameter table, and preset load allocation table were deployed to the control system of the five-stand mill as the target model structure, enabling the above optimizations to be applied to actual production and achieving high-quality and stable rolling of multiple varieties and wide specifications of products.

[0045] In some instances, the rolling model parameter table is expanded to generate a target rolling model parameter table, including: The steel grade interval in the rolling model parameter table is expanded from the original position to the first preset steel grade interval in the smooth roll mode and the second preset steel grade interval in the rough roll mode. The original position is 16. The thickness interval in the rolling model parameter table is expanded from the original level to the first preset thickness level in the smooth roll mode and the second preset thickness level in the rough roll mode. The width group spacing in the rolling model parameter table is re-divided to generate an optimized width group spacing; Based on the amplified steel grade group spacing, the amplified thickness group spacing, and the optimized width group spacing, a target rolling model parameter table is generated.

[0046] For example, please refer to Figure 2 and Figure 3This document presents schematic diagrams of thickness spacing expansion and steel grade spacing expansion in the rolling model parameter table provided in this application embodiment. In the rolling model parameter table of the five-roll mill, the original steel grade spacing has 16 levels. This number of levels can only roughly distinguish steel grades based on yield strength and cannot cover the process differences between the two different roll surface states: smooth rolls and rough rolls. To solve this problem, during the expansion process, the steel grade spacing is first expanded from the original 16 levels to the first preset steel grade level in the smooth roll mode and the second preset steel grade level in the rough roll mode. The first preset steel grade level has 96 levels, and the second preset steel grade level also has 96 levels. The smooth roll mode and the rough roll mode each independently have 96 steel grade levels, allowing steel grades with different yield strengths and different reduction rate requirements that originally shared the same steel grade level to be allocated to different levels according to their actual rolling characteristics, providing a grading basis for indexing rolling process parameters.

[0047] While expanding the steel grade group spacing, the thickness group spacing in the rolling model parameter table was also expanded. The original thickness group spacing had 16 levels, covering a wide range of thicknesses from 0.09mm to 1.0mm, forcing multiple different thicknesses to share the same thickness level. The expansion process increased the thickness group spacing from the original 16 levels to the first preset thickness level in the smooth roll mode and the second preset thickness level in the rough roll mode, with both the first and second preset thickness levels having 32 levels. The smooth roll mode and the rough roll mode each have 32 independent thickness levels, allowing strips of different thicknesses to obtain their thickness index positions on their respective roll surfaces, avoiding the mixing of process parameters due to insufficient thickness grading.

[0048] The width intervals in the rolling model parameter table are redefined to generate optimized width intervals. While the original rolling model parameter table included width intervals, its division method did not match the expanded steel grade and thickness intervals, and it was not optimized for multi-variety production. This application, based on the actual width distribution range in production and the sensitivity of different steel grades and thicknesses to width changes, redefines the width interval division range, creating a coordinated index system between the width intervals and the expanded steel grade and thickness intervals. The redefined width intervals can distinguish the different requirements of narrow and wide strip steel during the rolling process, allocating more accurate rolling process parameters for each combination of steel grade, thickness, and width.

[0049] In summary, this application's embodiments expand and optimize the steel grade group spacing, thickness group spacing, and width group spacing in the rolling model parameter table. Products of different steel grades, thicknesses, and widths, which were previously forced to share the same parameter level due to insufficient grading, now have independent index positions. The smooth roll mode and the rough roll mode are configured with independent 96 steel grade levels and 32 thickness levels, respectively. This ensures that different steel grades, such as plain plates, tin-plated substrates, and secondary cold-rolled materials, as well as various thickness specifications within the 0.09mm to 1.0mm range, are matched with corresponding rolling process parameters, effectively eliminating rolling force and tension setting deviations caused by parameter mixing. Simultaneously, the re-division of the width group spacing further improves the precision of parameter matching. The resulting target rolling model parameter table serves as the data foundation for the five-stand mill control system, improving the accuracy of process parameter settings for different varieties and specifications of products, and providing data support for tension control, rolling force calculation, and thickness accuracy control during the rolling process.

[0050] In some instances, it also includes: Obtain the specification information of the strip steel to be rolled, including steel type, thickness and width. Based on the specification information, the surface condition of the target roll corresponding to the strip to be rolled is determined, wherein the surface condition of the target roll is either a smooth roll condition or a rough roll condition. Based on the surface condition of the target roll, a steel grade retrieval operation is performed from the target rolling model parameter table to determine the target steel grade corresponding to the strip to be rolled.

[0051] For example, the specifications of the strip to be rolled are obtained, including steel grade information, thickness information, and width information. Steel grade information is used to identify the material type of the strip, such as plain steel or tin-plated steel; thickness information is used to determine the target thickness value of the strip; and width information is used to identify the width dimension of the strip. This specifications are input for determining the surface condition of the rolls and performing steel grade retrieval, and are read from the production planning system or material tracking system.

[0052] Based on the acquired specification information, the surface condition of the target roll corresponding to the strip to be rolled is determined. The target roll surface condition is divided into two types: smooth roll condition and rough roll condition. The determination criteria include the steel type, thickness range, and surface quality requirements of the strip to be rolled: for products with high surface quality requirements, such as tin-plated substrates or deep-drawing steel, a smooth roll condition is determined; for products requiring a higher friction coefficient to accommodate large reduction rolling, a rough roll condition is determined. By using the roll surface condition as the first-level retrieval condition, independent parameter index paths are established for the two different production processes of smooth rolls and rough rolls, avoiding deviations in rolling force and tension settings between smooth roll and rough roll products caused by using the same set of parameters.

[0053] Based on the determined surface condition of the target roll, a steel grade retrieval operation is performed from the target rolling model parameter table to determine the target steel grade corresponding to the strip to be rolled. The target steel grade is an identifier in the target rolling model parameter table used to uniquely index a set of rolling process parameters. The retrieval operation first uses the surface condition of the target roll as a filtering condition to determine whether to use the smooth roll mode branch or the rough roll mode branch; then, under the corresponding mode branch, it matches based on steel grade information, thickness information, and width information, ultimately locating the target steel grade uniquely corresponding to that specification combination. This target steel grade is used to retrieve the required rolling force setting value, tension setting value, and other process parameters for that strip from the target rolling model parameter table.

[0054] In summary, this embodiment of the application obtains specification information including steel grade, thickness, and width as input, and uses this information to determine the surface condition of the target roll (either smooth or rough), avoiding potential misjudgments that might occur when operators manually select modes based on experience. Based on this, a steel grade-based search operation is performed using the target roll surface condition as the first search condition, separating the parameter indices for smooth and rough rolls under different production processes. This resolves the problem of interference between smooth and rough roll products caused by sharing the same parameter level. The retrieved target steel grade uniquely corresponds to a set of rolling process parameters. The rolling force, tension, and other set values ​​retrieved from the target rolling model parameter table are matched with the steel grade, thickness, width, and roll surface condition of the strip to be rolled, improving the accuracy of process parameter settings and reducing rolling force deviations and thickness fluctuations caused by ambiguous parameter indices.

[0055] In some instances, based on the surface condition of the target roll, a steel grade retrieval operation is performed from the target rolling model parameter table to determine the target steel grade corresponding to the strip to be rolled, including: Using the surface condition of the target roll as the first search condition, determine the target mode branch in the target rolling model parameter table; Under the target mode branch, determine whether there is a steel grade code corresponding to the strip to be rolled; When a steel grade code exists, the steel grade code is used as the second search condition. The steel grade corresponding to the steel grade code is queried from the target rolling model parameter table, and the queried steel grade is determined as the target steel grade. The steel grade code is a unique identifier constructed based on steel type information, thickness information, and width information, and is used to query the unique corresponding steel grade. When no steel grade code exists, obtain the tap mark and yield strength of the strip to be rolled, use the tap mark and yield strength as the third search condition, query the corresponding steel grade from the target rolling model parameter table, and determine the queried steel grade as the target steel grade; Query the rolling process parameters corresponding to the target steel grade from the target rolling model parameter table.

[0056] For example, the target roll surface condition is used as the first search condition to determine the target mode branch in the target rolling model parameter table. The target roll surface condition is pre-determined based on the specifications of the strip to be rolled, specifically including two values: smooth roll condition and rough roll condition. When the determination result is smooth roll condition, the search operation enters the smooth roll mode branch in the target rolling model parameter table; when the determination result is rough roll condition, the search operation enters the rough roll mode branch. During the expansion process, the target rolling model parameter table has already independently configured its own steel grade group spacing, thickness group spacing, and width group spacing for smooth roll mode and rough roll mode respectively. Therefore, using the target roll surface condition as the starting point of the search ensures that all subsequent parameter indexes are executed under the correct process mode branch, avoiding the mixing of process parameters caused by smooth rolls and rough rolls sharing the same index system. This step is the starting condition for the entire steel grade classification search operation, determining the data table space on which all subsequent search logic depends.

[0057] After determining the target mode branch, the retrieval operation performs a second step under that branch: determining whether a steel grade code corresponding to the strip to be rolled exists. The steel grade code is a unique identifier constructed based on the steel grade, thickness, and width information of the strip to be rolled. Its function is to map a specific combination of steel grade, thickness, and width to a unique code value. In actual production, a single tap mark may correspond to multiple different reduction rates or process paths, thus corresponding to multiple steel grade codes, and each steel grade code uniquely corresponds to a steel grade. When determining whether a steel grade code exists, the retrieval operation queries the control system to see if the corresponding steel grade code record has been pre-maintained for the current strip's specifications. If it exists, it means that the current strip's specification combination has a predefined index identifier, and the retrieval path can be entered; if it does not exist, it means that the current strip's specifications have not yet been assigned a steel grade code, and the operation needs to switch to an alternative retrieval path.

[0058] When a steel grade code corresponding to the strip to be rolled exists, the retrieval operation uses this steel grade code as the second search condition, queries the target rolling model parameter table for the corresponding steel grade code, and identifies the retrieved steel grade as the target steel grade. The steel grade code, as a unique identifier constructed based on steel grade, thickness, and width, has a one-to-one correspondence with the steel grades in the target rolling model parameter table. The retrieval operation uses the steel grade code as the keyword to match within the index structure of the target mode branch. Since the steel grade code uniquely defines the specification combination, the retrieved steel grade is also unique. After determining the target steel grade, the retrieval operation queries the target rolling model parameter table for the rolling process parameters corresponding to that target steel grade, including rolling force settings and tension settings. This path utilizes the precise indexing capability of the steel grade code, ensuring that each combination of steel grade, thickness, and width yields a unique corresponding steel grade, thus obtaining perfectly matching rolling process parameters and avoiding fuzzy parameter matching caused by multiple reduction rates corresponding to the tap mark.

[0059] When no steel grade code exists for the strip to be rolled, the retrieval operation obtains the tapping mark and yield strength of the strip. Using these as the third retrieval criteria, it searches for the corresponding steel grade in the target rolling model parameter table and identifies the found steel grade as the target steel grade. The tapping mark is a code used to identify the heat or batch information, and the yield strength is a performance indicator reflecting the hardness of the strip material. In the case of a missing steel grade code, the retrieval operation downgrades to using the combination of the tapping mark and yield strength as the query criteria. Under the target mode branch, it matches the pre-set yield strength grading to determine the corresponding steel grade. This backup path ensures that even without pre-maintaining steel grade codes for specification information, the retrieval operation can still find a suitable steel grade based on the tapping mark and yield strength, thereby retrieving the corresponding rolling process parameters. This path, as a supplement to the steel grade code path, maintains the integrity of the entire steel grade grading retrieval operation and avoids retrieval failure due to missing steel grade codes.

[0060] After determining the target steel grade, the retrieval operation queries the target rolling model parameter table for the rolling process parameters corresponding to that steel grade. After expansion processing, the target rolling model parameter table has refined the steel grade group spacing, thickness group spacing, and width group spacing, and stores the corresponding rolling process parameters indexed by the steel grade. The query operation uses the target steel grade as the keyword to extract the associated rolling force setting, tension setting, and reduction allocation values, among other process parameters, under the target mode branch. These parameters will be used for the actual rolling control of the five-strand mill. Regardless of whether the target steel grade is determined through the steel grade code path or through the tap mark and yield strength path, the final query result is a complete set of rolling process parameters uniquely associated with that steel grade, ensuring the certainty and traceability of the process parameter settings.

[0061] In summary, the steel grade retrieval operation described in this application separates the parameter indexes for smooth rolls and rough rolls under different production processes by using the surface condition of the target roll as the first retrieval condition. This eliminates the mutual interference caused by smooth rolls and rough rolls sharing the same set of parameter levels in the original model. Within the smooth roll mode or rough roll mode branch, the steel grade code is used as the second retrieval condition to achieve precise matching of specification levels. The steel grade code is constructed from three dimensions: steel grade, thickness, and width. This ensures that each specific specification combination can obtain a unique corresponding steel grade, thereby matching a set of rolling process parameters. This solves the problem of products with different reduction rates sharing the same parameter level due to fuzzy matching relying solely on the tap mark and yield strength in the original model. When the steel grade code is not maintained, the retrieval operation is downgraded to using the tap mark and yield strength as the third retrieval condition, maintaining compatibility with the original retrieval logic and ensuring parameter availability when specification information is incomplete during the production process. Through this hierarchical and progressive retrieval structure, from the surface condition of the rolls to precise specification matching and then to downgraded compatibility, the entire retrieval operation improves the accuracy of parameter setting while maintaining the robustness of the system. Finally, the rolling process parameters retrieved from the target rolling model parameter table are matched with the steel grade, thickness, width and surface condition of the actual strip to be rolled.

[0062] In some instances, steel grade and thickness search fields are added to the friction coefficient model parameter table to generate a target friction coefficient model parameter table, including: Construct a storage structure for friction coefficient parameters with roll type, steel grade, thickness, stand number, and emulsion mode as joint indexes; For each set of index values ​​in the composite index, obtain the corresponding friction coefficient parameter; Write the friction coefficient parameters into the friction coefficient parameter storage structure to generate the target friction coefficient model parameter table.

[0063] For example, by adding steel grade and thickness search fields to the friction coefficient model parameter table, when generating the target friction coefficient model parameter table, a friction coefficient parameter storage structure is first constructed using roll type, steel grade, thickness, stand number, and emulsion mode as a joint index. Specifically, the roll type is used to distinguish between smooth rolls and rough rolls, because the frictional characteristics generated when rolls with different surface conditions come into contact with strip steel are different. Smooth smooth rolls require lower friction coefficient parameters, while rough rolls require higher friction coefficient parameters to provide sufficient rolling friction.

[0064] Steel grade, as the second dimension of the joint index, is used to distinguish different material types such as plain steel, tin-plated steel, and secondary cold-rolled steel. The chemical composition and yield strength of different steel grades determine their different friction behaviors during the rolling deformation process. For example, tin-plated steel requires a separate configuration of friction coefficient parameters to avoid slippage due to its high surface hardness, while plain steel can use more conventional parameter settings.

[0065] Thickness, as the third dimension of the joint index, is used to distinguish various thickness specifications in the range of 0.09mm to 1.0mm. The friction state of thin strip steel differs from that of thick strip steel due to the change in the shape ratio of the deformation zone during the rolling process. Independent parameters need to be configured to ensure the accuracy of the calculation of rolling force and forward slip value.

[0066] The stand number serves as the fourth dimension, used to distinguish the different positions of the first to fifth stands in the five-stand rolling mill. Each stand bears different reduction amounts and strip inlet and outlet thicknesses, and the friction coefficient decreases or increases with the position of the stand. Therefore, each stand requires an independent friction coefficient parameter.

[0067] The emulsion mode, as the fifth dimension, is used to distinguish emulsion operating conditions with different flow rates, concentrations, and injection methods. The lubrication effect of the emulsion directly affects the coefficient of friction. By combining these five dimensions into a joint index, each set of index values ​​in the friction coefficient parameter storage structure corresponds to a uniquely determined operating condition.

[0068] For each set of index values ​​in the composite index, the corresponding friction coefficient parameters are obtained, and then these friction coefficient parameters are written into the constructed friction coefficient parameter storage structure to generate the target friction coefficient model parameter table. The process of obtaining the friction coefficient parameters includes: for each set of working conditions uniquely determined by roll type, steel grade, thickness, stand number, and emulsion mode in the composite index, the measured rolling force and forward sliding values ​​during the actual rolling process under that working condition are extracted from historical production data. The friction coefficient parameters corresponding to that working condition are then deduced using the inverse friction coefficient calculation formula, or the friction coefficient parameters under a specific working condition are determined through on-site experiments. For new steel grades or new thickness specifications for which sufficient historical data has not yet been accumulated, initial friction coefficient parameters can be obtained through interpolation under similar working conditions or calculation using theoretical models. After obtaining the friction coefficient parameters corresponding to each set of index values, the parameter values ​​are written into the previously constructed storage structure according to the dimensional order of the composite index. That is, each record contains a roll type field, steel grade field, thickness field, stand number field, emulsion mode field, and the corresponding friction coefficient parameter value field. Once all index combinations are written, the generated target friction coefficient model parameter table will have the ability to quickly query friction coefficient parameters using a five-dimensional joint index. The smooth roller state and the rough roller state will each store their corresponding parameter values ​​independently without interfering with each other.

[0069] In some instances, load allocation values ​​are determined based on load allocation models, including: Obtain the initial load distribution values ​​for each rack; The initial load distribution value is verified to generate verification results. The verification process includes rolling force limit verification, rolling force balance verification between the first and second stands, and power balance verification between the second and fifth stands. Based on the verification results, regression calculations were performed on the initial load distribution value using historical strip rolling data to generate an optimized load distribution value, which was then used as the load distribution value for each stand.

[0070] For example, when determining the load allocation value based on the load allocation model, the initial load allocation value for each stand is first obtained. The initial load allocation value refers to the initial proportional allocation of the reduction or rolling force borne by each of the first to fifth stands in a five-stand rolling mill. This initial value can be read from a pre-stored load allocation parameter table, or it can be calculated using empirical formulas based on the incoming thickness and exit thickness of the strip to be rolled, as well as the equipment capacity of each stand. The purpose of obtaining the initial value is to provide a basic allocation scheme for verification and optimization. Without the initial value, the load allocation model loses its starting point for adjustment.

[0071] In actual production, the initial load distribution value usually reflects the preliminary roll reduction distribution based on historical data or process experience. However, due to fluctuations in incoming material thickness, changes in steel grade, and differences in equipment condition, directly using the initial value for rolling may cause the rolling force of some stands to exceed the allowable range of the equipment, or the rolling force ratio between the first and second stands to be too large, leading to the risk of slippage or strip breakage, or the uneven distribution of motor power between the second to fifth stands, resulting in a decrease in production efficiency.

[0072] The initial load distribution values ​​are verified to generate verification results. Verification processes include rolling force limit verification, rolling force balance verification between the first and second stands, and power balance verification between the second to fifth stands. The rolling force limit verification involves substituting the initial load distribution values ​​for each stand into rolling force calculation formulas such as the Hill formula or the Bland-Ford formula to calculate the expected rolling force for each stand. The calculated rolling force is then compared to the maximum allowable rolling force threshold of the machinery. If the calculated rolling force for a stand exceeds the threshold, the verification result is a failure, and the load distribution value for that stand needs to be reduced. The rolling force balance verification between the first and second stands is used to limit the rolling force of the first stand from becoming excessive. Specifically, the ratio of the rolling force of the first stand to the rolling force of the second stand is calculated and compared to a preset balance range. If the ratio exceeds the allowable upper limit, it indicates that the first stand is overloaded, and the verification result is a failure, requiring a partial transfer of the load from the first stand to the second stand. The power balance check between the second to fifth racks is used to fully utilize the power efficiency of the motors in these racks. It calculates the deviation between the individual motor power and the average power of each of the four racks. If the deviation exceeds a preset threshold, the check fails, and the load distribution of each rack needs to be adjusted to achieve a more balanced power distribution. Through these three checks, the load distribution model identifies distribution defects in the initial values ​​that could lead to equipment overload, rolling instability, or low efficiency. It outputs a check result indicating whether each check item passed or failed. This result shows whether the initial load distribution values ​​meet safety and efficiency requirements and which racks need adjustment.

[0073] Based on the verification results, regression calculations were performed on the initial load distribution values ​​using historical strip rolling data to generate optimized load distribution values. These optimized load distribution values ​​were then used as the load distribution values ​​for each stand. The specific regression calculation process is as follows: when the verification results show that a certain indicator is unqualified, the load distribution model uses the verification pass as the objective function, takes the load distribution values ​​of each stand as the variables to be optimized, and uses the load distribution parameters of previously successfully rolled strips stored in the historical strip rolling data as training samples. A new set of load distribution values ​​is iteratively calculated using regression algorithms such as least squares or gradient descent. The historical strip rolling data includes the actual load distribution values ​​used in the past production of strips of different steel grades and thicknesses, as well as the corresponding measured rolling force and power values. These data reflect which load distribution simultaneously satisfies the three constraints of rolling force limit, inter-stand rolling force balance, and power balance under real operating conditions. The regression calculation uses the initial value as the starting point for iteration. Based on the deviation direction and amount provided by the verification results, and combined with the optimal allocation mode under similar working conditions in historical data, the load allocation values ​​of each stand are gradually adjusted until the newly calculated rolling force and power all pass the verification. The final optimized load allocation values ​​retain the response characteristics to process requirements in the initial values, and eliminate the risk of exceeding limits in the initial values ​​through regression correction of historical data. This ensures that the reduction undertaken by each stand during actual rolling is within the safe range allowed by the equipment, the rolling force between the first and second stands is maintained in a reasonable ratio, and the motor power output of the second to fifth stands is balanced, improving rolling stability and production efficiency.

[0074] In summary, this application embodiment obtains the initial load distribution values ​​for each stand, performs rolling force limit verification, rolling force balance verification between the first and second stands, and power balance verification between the second to fifth stands on the initial values. It uses historical strip rolling data to perform regression calculations on the initial values ​​to generate optimized values. The rolling force limit verification substitutes the initial load distribution values ​​into the rolling force formula to calculate the expected rolling force for each stand and compares it with the maximum allowable value of the equipment. This identifies distribution defects that may lead to equipment overload, preventing equipment damage or strip breakage accidents caused by rolling forces exceeding the mechanical bearing capacity during actual rolling. The rolling force balance verification between the first and second stands calculates the ratio of the rolling forces of the two stands and compares it with a preset balance range to identify potential problems with excessive load on the first stand. This avoids slippage, scratches, or strip deviation caused by excessive rolling force on the first stand, ensuring the stability of the entry section rolling. The power balance verification between the second to fifth stands analyzes the uniformity of power distribution among the four stand motors, identifying efficiency losses caused by uneven power distribution. This enables the motors in each stand to work collaboratively under near-rated power conditions, improving overall rolling speed and production rhythm. Based on the verification results, regression calculations are performed using historical strip rolling data. The previous optimal allocation mode is used as a learning sample, and the initial values ​​are iteratively adjusted until all verification items pass. The generated optimized load allocation value not only meets the hard constraints of safety and efficiency but also incorporates the process rules from historical production experience. Under the premise of ensuring that the rolling force of each stand does not exceed the limit, the rolling force between stands is balanced, and the power is evenly distributed, the load allocation value is optimized, reducing the frequency of operators manually adjusting the reduction of each stand and improving the level of automatic control of load allocation in the rolling process of multi-variety and thin-gauge products.

[0075] In some instances, the number of specified stands is determined based on the surface condition of the target roll. Specifically, when the target roll surface condition is a smooth roll, the number of specified stands is less than or equal to 4; when the target roll surface condition is a rough roll, the number of specified stands is less than or equal to 3.

[0076] For example, when determining the number of designated stands, the surface condition of the target roll, determined based on the specifications of the strip to be rolled, is used as the criterion. When the target roll surface condition is a smooth roll, the number of designated stands is set to less than or equal to four. This is because the smooth roll surface has a relatively low coefficient of friction between the roll and the strip, making it easier for the strip to slip or deviate during rolling. Therefore, the load distribution value needs to be applied to more stands to ensure rolling stability.

[0077] Specifically, in the five stands of the five-stand rolling mill, in the smooth roll mode, up to four stands are allowed to have their load distribution set using empirical values ​​from a preset load distribution table. At least one stand is left unspecified; its load distribution value is not provided by the preset load distribution table but is determined by the existing load distribution model through power balance calculations. The purpose of retaining at least one stand for load distribution model calculation is to address real-time fluctuations in incoming and exit thicknesses during rolling. Since the incoming thickness in actual production is not constant, and the target exit thickness changes with product specifications, if all five stands use preset, fixed empirical values, the model cannot dynamically adjust the reduction distribution based on actual incoming material fluctuations, potentially leading to rolling force deviations or poor strip shape. Therefore, in the smooth roll mode, the number of specified stands is controlled to be less than or equal to four. This allows for empirical intervention on critical stands prone to rolling force anomalies or vibration risks while retaining the ability for at least one stand to be automatically calculated by the load distribution model, giving the entire system the flexibility for adaptive adjustment.

[0078] When the target roll surface is in a rough roll condition, the number of designated stands is set to less than or equal to three. A rough roll surface has a certain degree of roughness, resulting in a relatively high coefficient of friction between the roll and the strip. The strip is more easily gripped and rolled stably on the rough roll surface, thus requiring relatively less intervention through empirical load distribution.

[0079] Specifically, in the rough roll mode, up to three stands can be designated for load allocation using empirical values ​​from a preset load allocation table, while the load allocation values ​​for the remaining at least two stands are determined by the existing load allocation model through power balance calculations. The reason for a smaller upper limit on the number of stands designated in the rough roll mode compared to the smooth roll mode is that the high friction characteristics of the rough roll itself provide a better stability foundation for the rolling process. Too many designated stands would limit the load allocation model's ability to optimize the overall performance based on actual material fluctuations. Furthermore, in the rough roll mode, the rolled products are typically relatively thick or have a large reduction, necessitating the retention of multiple stands for the load allocation model to handle thickness variations. Controlling the number of designated stands to three or less means that at least two stands have their load allocation values ​​determined by the load allocation model through power balance and regression calculations. These stands can absorb the reduction allocation adjustment requirements caused by fluctuations in incoming material thickness and changes in exit thickness, ensuring the stability of the rolling process and the accuracy of product thickness.

[0080] Whether in smooth roller mode or rough roller mode, the specific selection of the designated frame is not fixed. Operators or process systems can select any frame from the five frames, not exceeding the maximum number, as the designated frame based on the characteristics of the products being produced and the operating status of each frame. The remaining frames that are not selected are calculated by the load distribution model. This selection mechanism allows the use of the preset load distribution table to be applicable to specific frames that are prone to problems, and the designated range can be adjusted according to the actual working conditions.

[0081] In some instances, it also includes: Based on steel grade and thickness information, the maximum rolling speed of the five-strand rolling mill is determined.

[0082] For example, steel grade information is used to identify the material type of the strip, including different varieties such as plain plate, tinplate, and secondary cold-rolled material. Different steel grades have different critical speeds for vibration during high-speed rolling. For example, tinplate, due to its high material hardness and sensitive surface characteristics, is prone to high-frequency vibration in the fourth stand when it is thin, and its safe rolling speed limit is much lower than that of plain plate. Thickness information is used to determine the target thickness value of the strip, especially for extremely thin products with a thickness of less than or equal to 0.2 mm. As the thickness decreases, the tension fluctuation and deformation resistance changes of the strip during rolling become more drastic, and the risk of mill vibration increases significantly. Therefore, it is necessary to set corresponding speed limits based on the actual thickness value.

[0083] After obtaining the steel grade and thickness information, these two pieces of information are used as search criteria to query a pre-constructed preset speed limit table. This table stores the maximum rolling speed safety limits set for different steel grades and thickness ranges. These speed limits are determined based on statistical analysis of historical production data. Specifically, it includes records of critical speeds where mill vibration occurred during the production of various steel grades and thicknesses of strip, as well as the maximum safe operating speed without vibration. A mapping relationship between steel grade, thickness, and speed limit is formed through data fitting and threshold induction. During the query, the steel grade and thickness information of the current strip to be rolled are matched with the index fields in the table to extract the corresponding maximum speed limit value. This value is then used as the maximum rolling speed allowed for the five-strand mill when rolling the current strip.

[0084] After determining the maximum rolling speed, this speed limit value is written into the speed control program of the five-stand mill control system. Specifically, this is written to the acceleration control module and the maximum speed limit module. The original maximum rolling speed output by the speed control program is a theoretical maximum value calculated based on equipment capacity factors such as equipment power and rolling rhythm. This value does not consider the process characteristics of different steel grades and thicknesses. Directly using this theoretical maximum value for high-speed rolling can easily cause mill vibration in the fourth stand for sensitive products such as thin-gauge tin-plated substrates, leading to strip breakage or surface quality defects. After writing the speed limit value, the speed control program first calculates the theoretical maximum speed allowed by the equipment capacity during execution. Then, it reads the safety limit value matching the current steel grade and thickness from the preset speed limit table, and uses the smaller of the two values ​​as the actual maximum rolling speed output. In this way, when the theoretical maximum speed is lower than the safety limit value, the actual maximum speed is limited by the equipment capacity; when the theoretical maximum speed is higher than the safety limit value, the actual maximum speed is limited within the safety limit value, thus avoiding mill vibration caused by excessive speed. Meanwhile, the speed limit value is not fixed. As the steel grade or thickness of the strip to be rolled changes, the control system re-queries the preset speed limit table to obtain the speed limit value that matches the new specification, realizing the speed limit switching between different product specifications without requiring operators to manually modify the value in the maximum speed setting box based on experience.

[0085] Please see Figure 4 The above is a schematic diagram of a model structure optimization device for a five-stand rolling mill provided in an embodiment of this application, including: The parameter acquisition unit 21 is used to acquire the rolling model parameter table and the friction coefficient model parameter table in the control system of the five-roll mill. The rolling model parameter table is used to store rolling process parameters indexed by steel grade group spacing, thickness group spacing and width group spacing. The friction coefficient model parameter table is used to store friction coefficient parameters indexed by roll type, emulsion mode and stand number. The expansion processing unit 22 is used to expand the rolling model parameter table and generate the target rolling model parameter table. The retrieval expansion unit 23 is used to add steel type retrieval field and thickness retrieval field to the friction coefficient model parameter table and generate the target friction coefficient model parameter table. The rolling distribution unit 24 is used to construct a preset load distribution table indexed by steel type and thickness. The preset load distribution table is used to prioritize the load distribution setting of a specified stand during the rolling process, and the load distribution value of the remaining stands is determined based on the load distribution model. The model deployment unit 25 is used to deploy the target rolling model parameter table, the target friction coefficient model parameter table, and the preset load distribution table to the control system of the five-strand rolling mill, as the target model structure of the five-strand rolling mill.

[0086] Please see Figure 5 This application also provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, it implements the steps of the model structure optimization method for a five-strand rolling mill.

[0087] Since the electronic device described in this embodiment is the device used to implement the model structure optimization device for the five-roll mill in this application embodiment, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the method described in this application embodiment. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. Any device used by those skilled in the art to implement the method in this application embodiment is within the scope of protection of this application.

[0088] In practice, when the computer program 311 is executed by the processor, it can implement any of the embodiments corresponding to the first aspect.

[0089] This application also provides a computer program product, which includes computer software instructions that, when executed on a processing device, cause the processing device to perform... Figure 1 The flowchart of a model structure optimization method for a five-stand rolling mill in the corresponding embodiment.

[0090] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to the embodiments of this application are implemented. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium may be any usable medium that a computer can store or a data storage device such as a server or data center that integrates one or more usable media. The usable medium may be a magnetic medium, an optical medium, or a semiconductor medium, etc.

[0091] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

[0092] Although preferred embodiments have been described in this specification, 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 beyond the scope of this specification.

[0093] Obviously, those skilled in the art can make various modifications to this specification without departing from its spirit and scope. Therefore, this specification also intends to include any modifications that fall within the scope of the claims and their equivalents.

Claims

1. A method for optimizing the model structure of a five-stand rolling mill, characterized in that, include: Obtain the rolling model parameter table and friction coefficient model parameter table from the control system of the five-stand rolling mill. The rolling model parameter table is used to store rolling process parameters indexed by steel grade group spacing, thickness group spacing, and width group spacing. The friction coefficient model parameter table is used to store friction coefficient parameters indexed by roll type, emulsion mode, and stand number. The rolling model parameter table is expanded to generate the target rolling model parameter table; Add steel type and thickness search fields to the friction coefficient model parameter table to generate the target friction coefficient model parameter table; Construct a preset load allocation table indexed by steel type and thickness, wherein the preset load allocation table is used to preferentially set the load allocation of a specified stand during the rolling process using the preset load allocation value, and the load allocation value of the remaining stands is determined based on the load allocation model; The target rolling model parameter table, the target friction coefficient model parameter table, and the preset load distribution table are deployed to the control system of the five-strand rolling mill as the target model structure of the five-strand rolling mill.

2. The method according to claim 1, characterized in that, The process of expanding the rolling model parameter table to generate the target rolling model parameter table includes: The steel grade interval in the rolling model parameter table is expanded from the original position to the first preset steel grade in the smooth roll mode and the second preset steel grade in the rough roll mode, where the original position is 16. The thickness interval in the rolling model parameter table is expanded from the original level to the first preset thickness level in the smooth roll mode and the second preset thickness level in the rough roll mode. The width group spacing in the rolling model parameter table is re-divided to generate an optimized width group spacing. Based on the amplified steel grade group spacing, the amplified thickness group spacing, and the optimized width group spacing, the target rolling model parameter table is generated.

3. The method according to claim 2, characterized in that, The first preset steel grade level is 96, the second preset steel grade level is 96, the first preset thickness level is 32, and the second preset thickness level is 32.

4. The method according to claim 1, characterized in that, Also includes: Obtain the specification information of the strip steel to be rolled, wherein the specification information includes steel grade information, thickness information and width information; Based on the specification information, the surface condition of the target roll corresponding to the strip to be rolled is determined, wherein the surface condition of the target roll is a smooth roll condition or a rough roll condition; Based on the surface condition of the target roll, a steel grade retrieval operation is performed from the target rolling model parameter table to determine the target steel grade corresponding to the strip to be rolled.

5. The method according to claim 4, characterized in that, The step of performing a steel grade retrieval operation from the target rolling model parameter table based on the surface condition of the target roll to determine the target steel grade corresponding to the strip to be rolled includes: Using the surface condition of the target roll as the first search condition, the target mode branch in the target rolling model parameter table is determined; Under the target mode branch, determine whether there is a steel grade code corresponding to the strip to be rolled; When the steel grade code exists, the steel grade code is used as the second search condition to query the steel grade corresponding to the steel grade code from the target rolling model parameter table, and the queried steel grade is determined as the target steel grade. The steel grade code is a unique identifier constructed based on the steel type information, the thickness information and the width information, and is used to query the unique corresponding steel grade. When the steel grade code does not exist, the tap mark and yield strength of the strip to be rolled are obtained, and the tap mark and yield strength are used as the third search conditions to query the corresponding steel grade from the target rolling model parameter table, and the queried steel grade is determined as the target steel grade; Query the rolling process parameters corresponding to the target steel grade from the target rolling model parameter table.

6. The method according to claim 1, characterized in that, The step involves adding steel grade and thickness search fields to the friction coefficient model parameter table to generate a target friction coefficient model parameter table, including: Construct a storage structure for friction coefficient parameters with roll type, steel grade, thickness, stand number, and emulsion mode as joint indexes; For each set of index values ​​in the composite index, obtain the corresponding friction coefficient parameter; The friction coefficient parameters are written into the friction coefficient parameter storage structure to generate the target friction coefficient model parameter table.

7. The method according to claim 1, characterized in that, The load allocation values ​​are determined based on the load allocation model, including: Obtain the initial load distribution values ​​for each rack; The initial value of the load distribution is verified to generate a verification result. The verification process includes rolling force limit verification, rolling force balance verification between the first stand and the second stand, and power balance verification between the second stand and the fifth stand. Based on the verification results, regression calculations are performed on the initial load distribution value using historical strip rolling data to generate an optimized load distribution value, and the optimized load distribution value is determined as the load distribution value for each stand.

8. The method according to claim 4, characterized in that, The number of designated stands is determined based on the surface condition of the target roll. When the surface condition of the target roll is a smooth roll, the number of designated stands is less than or equal to 4; when the surface condition of the target roll is a rough roll, the number of designated stands is less than or equal to 3.

9. The method according to claim 4, characterized in that, Also includes: Based on the steel grade information and the thickness information, the maximum rolling speed of the five-strand rolling mill is determined.

10. A model structure optimization device for a five-strand rolling mill, characterized in that, include: The parameter acquisition unit is used to acquire the rolling model parameter table and the friction coefficient model parameter table in the control system of the five-stand mill. The rolling model parameter table is used to store rolling process parameters indexed by steel grade group spacing, thickness group spacing and width group spacing. The friction coefficient model parameter table is used to store friction coefficient parameters indexed by roll type, emulsion mode and stand number. An expansion processing unit is used to expand the rolling model parameter table to generate a target rolling model parameter table. The retrieval expansion unit is used to add steel type retrieval field and thickness retrieval field to the friction coefficient model parameter table to generate the target friction coefficient model parameter table. A rolling distribution unit is used to construct a preset load distribution table indexed by steel type and thickness. The preset load distribution table is used to prioritize the load distribution setting of a specified stand during the rolling process, and the load distribution value of the remaining stands is determined based on the load distribution model. The model deployment unit is used to deploy the target rolling model parameter table, the target friction coefficient model parameter table, and the preset load distribution table to the control system of the five-strand rolling mill, as the target model structure of the five-strand rolling mill.

Images