A method for automatically controlling the width of a hot-rolled steel coil

Abstract

This invention discloses an automatic control method for the width of hot-rolled steel coils. First, width data from the slab production process is collected and its average value is calculated. Then, the actual steel pulling amount for finishing rolling and coiling is calculated, and an adaptive genetic table for steel pulling amount is created. After the slab is drawn, the adaptive genetic table for steel pulling amount is queried to obtain genetic data. If retrieval fails, the process returns and a default width allowance value is selected. If retrieval is successful, the predicted coiling width allowance value is calculated for the strip to be rolled. Next, a recommended width allowance correction value is calculated. Then, the width allowance of the electrical equipment is set according to the default width allowance value or the recommended width allowance correction value to control the slab rolling process. Finally, data is stored and the adaptive genetic table for steel pulling amount is updated. This invention improves the finished product width accuracy through staged stable control of steel pulling amount, significantly reduces the proportion of manual intervention, significantly increases the width hit rate at the roughing mill exit, and improves the stability of production line width control.

Description

Technical Field

[0001] This invention belongs to the field of hot rolling technology, specifically relating to an automatic control method for the width of hot-rolled steel coils. Background Technology

[0002] In the hot rolling process, width control is mainly achieved by predicting width using a width model, and then adapting the genetic coefficient based on the deviation between the rolled data and the predicted data.

[0003] Existing neural network models for calculating the target width of roughing mills utilize dozens of influencing factors, achieving a high level of accuracy. However, due to the characteristics of such neural network models, learning new layers requires time, inevitably leading to a certain number of blocked coils. Furthermore, for process engineers, maintaining these models is difficult, and adjusting them is challenging. The width calculation model involves a rich set of genetic parameters, which are coupled to some extent. Coupled with operators' limited understanding of width control, relying solely on experience, subjective judgments and inaccuracies are possible, resulting in excessive width intervention, low hit rate, and high blockage rate. Summary of the Invention

[0004] The purpose of this invention is to provide an automatic control method for the width of hot-rolled steel coils. By creating an adaptive model for the hot-rolled width allowance, the method achieves adaptive control of the hot-rolled width allowance. This effectively improves the accuracy of the finished product width, significantly reduces the proportion of manual intervention, greatly increases the width hit rate at the roughing mill exit, and reduces the number of coil width blockages caused by manual intervention.

[0005] To achieve the above-mentioned technical objectives, the present invention adopts the following technical solution: an automatic control method for the width of hot-rolled steel coils, characterized by comprising the following steps.

[0006] S1. Create an adaptive genetic table for steel tension.

[0007] Width data from the slab production process is collected, including actual measurement points of roughing mill exit width, finishing mill exit width, and coiling inlet width, and their mean values ​​are calculated for each. Based on the steel pulling amount calculation formula, the difference between the measured values ​​of finishing mill exit width and roughing mill exit width, and the difference between the measured values ​​of coiling inlet width and roughing mill exit width are calculated to obtain the actual steel pulling amount of finishing mill and actual steel pulling amount of coiling, and an adaptive genetic table of steel pulling amount is created.

[0008] S2, Calculation of width margin prediction value.

[0009] After the slab is drawn, the genetic data is obtained by querying the adaptive genetic table of steel pulling amount. If the acquisition fails, the process returns and the default width margin value is selected, and step S4 is executed directly. If the acquisition is successful, the predicted value of coiling width margin DC_Margin_Pre is calculated for the strip steel to be rolled.

[0010] ,

[0011] Among them, DC_Spread_ACT is the actual amount of steel pulled during coiling, and RM_Margin is the roughing width allowance.

[0012] S3, Recommended margin correction value.

[0013] The recommended value for width margin correction is RM.MGN_RMD.

[0014] ,

[0015] Where △ represents the width design tolerance.

[0016] S4, Width margin setting.

[0017] The width margin is set for the electrical equipment based on the default width margin value or the recommended width margin correction value, and the equipment performs slab rolling control based on the width margin setting data.

[0018] S5. Data storage and updating of the adaptive genetic table for steel pulling quantity.

[0019] After each strip rolling is completed, the data is stored and written into the adaptive genetic table for updating the strip pulling amount.

[0020] Furthermore, the formula for calculating the amount of steel pulled is as follows:

[0021] ,

[0022] ,

[0023] in:

[0024] FM_Spread_ACT represents the actual amount of steel pulled during finishing rolling;

[0025] RM_Width_ACT is the measured width of the roughing mill exit.

[0026] FM_Width_ACT is the measured width of the finishing mill exit.

[0027] DC_Spread_ACT represents the actual amount of steel pulled during coiling;

[0028] DC_Width_ACT is the measured value of the take-up entry width;

[0029] Using steel grade, width layer, thickness layer, and finishing mill exit temperature layer as data division conditions, the initial calculation result of the steel pulling amount is used as the initial solution for the steel pulling amount. When the new strip steel pulling amount data is obtained, it is determined whether the deviation between the new strip steel pulling amount and the initial solution for the steel pulling amount is greater than 10%. If not, the data is summed with the average new strip steel pulling amount accounting for 70% and the initial solution for the steel pulling amount accounting for 30%. The result is used as the current solution and written into the genetic data table for repeated iteration of subsequent width data.

[0030] Furthermore, the mean calculation operation in step S1 is as follows: after removing the first and last 150 data points and outliers with a deviation greater than 30 from the calculated set data from the collected actual data, the mean of the data is used as the initial filtered data.

[0031] Furthermore, the adaptive genetic table for strip stretching is divided into a long-term genetic table and a short-term genetic table. Its expanded items include steel type (SFC), quality number, thickness class, and width class. The data items include actual strip stretching amount (FM_Spread_ACT) for finishing rolling and actual strip stretching amount (DC_Spread_ACT) for coiling. After each strip rolling is completed, the short-term genetic table is updated according to the mean calculated by the data acquisition method in step S1. The long-term genetic table is used for width margin prediction. The short-term genetic table is copied to the long-term genetic table at two times each day, morning and evening, when the shift changes. If the genetic table does not have data for this class, it is added.

[0032] Furthermore, step S5 also includes a screen display. After setting the width margin for the electrical equipment, the recommended values ​​for strip grade and roughing width margin correction are displayed on the screen.

[0033] The beneficial effects of this invention are: by using the method of this invention, the finished product width accuracy is improved by controlling the staged steel pulling amount stably, the proportion of manual intervention is greatly reduced, the width hit rate of the roughing mill exit is greatly improved, the number of steel coil width blockages caused by manual intervention is reduced, the manual labor load is reduced, and the stability of production line width control is improved. Attached Figure Description

[0034] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0035] The present invention will now be described in further detail with reference to the accompanying drawings.

[0036] like Figure 1 As shown, the present invention provides an automatic control method for the width of hot-rolled steel coils, suitable for use on various hot-rolling production lines, which can effectively improve product quality. The specific method includes the following steps:

[0037] S1. Create an adaptive genetic table for steel tension:

[0038] Width data from the slab production process is collected. The width data includes actual measurement points of the roughing mill exit width, finishing mill exit width, and coiling entry width. The first and last 150 data points and outliers with deviations greater than 30 from the calculated set data are removed from the actual data and used as the initial screening data. The mean value is then calculated.

[0039] Based on the steel pulling amount calculation formula, the differences between the measured values ​​of the finishing mill exit width and the roughing mill exit width, and the measured values ​​of the coiling inlet width and the roughing mill exit width are calculated to obtain the actual steel pulling amount of the finishing mill and the actual steel pulling amount of the coiling mill. An adaptive genetic table for steel pulling amount is then created. The steel pulling amount calculation formula is as follows:

[0040] (1),

[0041] (2),

[0042] in:

[0043] FM_Spread_ACT represents the actual amount of steel pulled during finishing rolling;

[0044] RM_Width_ACT is the measured width of the roughing mill exit.

[0045] FM_Width_ACT is the measured width of the finishing mill exit.

[0046] DC_Spread_ACT represents the actual amount of steel pulled during coiling;

[0047] DC_Width_ACT is the measured value of the take-up entry width;

[0048] Using steel grade, width layer, thickness layer, and finishing mill exit temperature layer as data division conditions, the initial calculation result of the steel pulling amount is used as the initial solution for steel pulling amount. When the new strip steel pulling amount data is obtained, it is determined whether the deviation between the new strip steel pulling amount and the initial solution for steel pulling amount is greater than 10%. If not, the data is summed with the average new strip steel pulling amount accounting for 70% and the initial solution for steel pulling amount accounting for 30%. The result is used as the current solution and written into the steel pulling amount adaptive genetic table for repeated iteration of subsequent width data.

[0049] The logic for classifying strip width layers, thickness layers, and finishing mill exit temperature layers is shown in Table 1 below:

[0050] Table 1. Classification of Width, Thickness, and Temperature Layers

[0051] .

[0052] The adaptive genetic table for steel spreading is divided into long-term genetic table and short-term genetic table. Its expanded items include steel type (SFC), quality number, thickness class, and width class. The data items include actual steel spreading amount (FM_Spread_ACT) for finishing rolling and actual steel spreading amount (DC_Spread_ACT) for coiling.

[0053] S2. Calculation of predicted width margin:

[0054] After the slab is drawn, the genetic data is obtained by querying the adaptive genetic table of steel pulling amount. If the acquisition fails, the process returns and the default width margin value is selected, and step S4 is executed directly. If the acquisition is successful, the predicted value of coiling width margin DC_Margin_Pre is calculated for the strip steel to be rolled.

[0055] (3),

[0056] in:

[0057] DC_Margin_Pre: Predicted roll-up width margin;

[0058] DC_Width: Target width for scrolling;

[0059] RM_Width: Target width for rough rolling;

[0060] RM_Margin: Rough rolling width allowance.

[0061] Combining formulas (2) and (3), DC_Width and RM_Width are respectively taken as DC_Width_ACT and RM_Width_ACT, resulting in the final calculation formula for the width margin prediction value:

[0062] (4),

[0063] Among them, DC_Spread_ACT is the actual amount of steel pulled during coiling, and RM_Margin is the roughing width allowance.

[0064] S3, Recommended margin correction value.

[0065] Considering product planning tolerances, taking half the tolerance as the ideal target, calculate the recommended value for width margin correction, RM.MGN_RMD.

[0066] ,

[0067] Where △ represents the width design tolerance.

[0068] Considering the stability of learning and the ease of operation of screen parameters, a value of 0 is recommended for minor corrections.

[0069] S4. Width allowance setting: The width allowance of the electrical equipment is set according to the default width allowance value or the recommended width allowance correction value. The equipment performs slab rolling control according to the width allowance setting data.

[0070] S5. Data Storage and Update of Adaptive Genetic Table for Strip Rolling: After each strip rolling cycle, the data is stored and written to the updated adaptive genetic table for strip rolling. Data storage includes all planned and actual data from previous steps for retrospective analysis. Additionally, relevant data is displayed on a screen. After setting the width allowance for the electrical equipment, the recommended values ​​for strip grade and roughing width allowance correction are displayed on the screen.

[0071] After each strip rolling is completed, the short-term genetic table is updated according to the mean calculated by the data collection method in step S1; the long-term genetic table is used when predicting the width margin. The short-term genetic table is copied to the long-term genetic table at two times every day when the shift changes in the morning and evening; if the genetic table does not have data for this layer, it is added.

[0072] The above content is only used to illustrate the technical solution of the present invention. Simple modifications or equivalent substitutions made by those skilled in the art to the technical solution of the present invention do not depart from the essence and scope of the technical solution of the present invention.

Claims

1. A method for automatically controlling the width of hot-rolled steel coils, characterized in that, Includes the following steps: S1. Create an adaptive genetic table for steel tension: Width data from the slab production process is collected, including actual measurement points of roughing mill exit width, finishing mill exit width, and coiling inlet width, and their mean values ​​are calculated for each. Based on the steel pulling amount calculation formula, the difference between the measured values ​​of finishing mill exit width and roughing mill exit width, and the difference between the measured values ​​of coiling inlet width and roughing mill exit width are calculated to obtain the actual steel pulling amount of finishing mill and coiling, and an adaptive genetic table of steel pulling amount is created. The formula for calculating the amount of steel pulled is as follows: ， ， in: FM_Spread_ACT represents the actual amount of steel pulled during finishing rolling; RM_Width_ACT is the measured width of the roughing mill exit. FM_Width_ACT is the measured width of the finishing mill exit. DC_Spread_ACT represents the actual amount of steel pulled during coiling; DC_Width_ACT is the measured value of the take-up entry width; Using steel grade, width layer, thickness layer, and finishing mill exit temperature layer as data division conditions, the initial calculation result of the steel pulling amount is used as the initial solution for the steel pulling amount. When the new strip steel pulling amount data is obtained, it is determined whether the deviation between the new strip steel pulling amount and the initial solution for the steel pulling amount is greater than 10%. If not, the data is summed with the average new strip steel pulling amount accounting for 70% and the initial solution for the steel pulling amount accounting for 30%. The result is used as the current solution and written into the steel pulling amount adaptive genetic table for repeated iteration of subsequent width data. S2. Calculation of predicted width margin: After the slab is drawn, the genetic data is obtained by querying the adaptive genetic table of steel pulling amount. If the acquisition fails, the process returns and the default width margin value is selected, and step S4 is executed directly. If the acquisition is successful, the predicted value of coiling width margin DC_Margin_Pre is calculated for the strip steel to be rolled. ， Among them, DC_Spread_ACT is the actual amount of steel pulled during coiling, and RM_Margin is the roughing width allowance; S3. Calculation of recommended value for margin correction: Calculate the recommended value for width margin correction, RM.MGN_RMD. ， Where △ represents the width design tolerance; S4. Width allowance setting: The width allowance of the electrical equipment is set according to the default width allowance value or the recommended width allowance correction value. The equipment performs slab rolling control according to the width allowance setting data. S5. Data storage and updating of the adaptive genetic table for strip rolling: After each strip rolling is completed, the data is stored and written to update the adaptive genetic table for strip rolling. Step S5 also includes a screen display. After setting the width margin for the electrical equipment, the recommended values ​​for strip grade and roughing width margin correction are displayed on the screen.

2. The automatic width control method for hot-rolled steel coils according to claim 1, characterized in that: The mean calculation operation in step S1 is as follows: after removing the first and last 150 data points and outliers with a deviation greater than 30 from the calculated set data from the collected actual data, the mean is calculated as the initial filtered data.

3. The automatic width control method for hot-rolled steel coils according to claim 1, characterized in that: The adaptive genetic table for steel pulling amount is divided into a long-term genetic table and a short-term genetic table. Its expanded items include steel type SFC, quality number, thickness class, and width class. The data items include actual steel pulling amount FM_Spread_ACT for finishing rolling and actual steel pulling amount DC_Spread_ACT for coiling. After each strip rolling is completed, the short-term genetic table is updated according to the mean calculated by the data collection method in step S1; the long-term genetic table is used when predicting the width margin. The short-term genetic table is copied to the long-term genetic table at two times a day, morning and evening when the shift changes.

Images