Online efficient prediction method for rolling force and torque in plate and strip rolling process

Abstract

The invention discloses an online efficient prediction method for rolling force and torque in the plate and strip rolling process. The online efficient prediction method comprises the following steps that firstly, rolling parameters are obtained; secondly, actual rolling parameters of rolled passes are collected; 3, calculating a temperature drop influence correction coefficient; 4, calculating rolling force and torque values of subsequent non-rolling passes; fifthly, rolling force and torque calculation results of all passes are output; and sixthly, the rolling force and torque of the follow-up passes are updated till rolling of all the passes is completed. According to the method, any database related to rolled pieces, rolling mills, friction working conditions and the like does not need to be established in advance, the universality and robustness of the model are high, and the defects that in an existing conventional method, a stable and reliable material database needs to be established in advance for accurately representing the current deformation resistance of materials, and time and labor are consumed are overcome; as the model structure is simple enough, the model structure can be directly written into a PLC (Programmable Logic Controller) without being executed in process automation, and is particularly suitable for a rolling production line with lower automation level and without a secondary control system.

Description

Technical Field

[0001] This invention belongs to the field of hot rolling production technology, specifically relating to an efficient online prediction method for rolling force and torque during strip rolling. Background Technology

[0002] In the strip rolling process, rolling force and torque play a crucial role. The accuracy of rolling force calculation directly impacts the precision of rolling schedule settings, thickness control, and strip shape quality. Rolling torque is a fundamental parameter for mill verification and a vital reference factor for maximizing equipment performance. Current research on rolling force primarily employs methods such as engineering methods, finite element methods, machine learning, and neural networks. However, its prediction accuracy heavily relies on stable and reliable material databases to accurately characterize the current deformation resistance, thermophysical parameters, and various boundary conditions of the rolled piece—a highly time-consuming and labor-intensive task. Furthermore, to obtain various correction coefficients (including chemical composition correction, rolling temperature correction, rolling speed correction, and adaptive coefficients) to maximize model prediction accuracy, rolling force and torque prediction models are generally located in secondary process automation, requiring a high level of automation in the rolling production line.

[0003] Chinese invention patent application number 201811029047.5 discloses a "method for predicting rolling force in the production of variable thickness plate and strip," which predicts the rolling force at any moment during the rolling process of variable thickness plate and strip based on minimizing the total power functional of the rolling deformation zone. Chinese invention patent application number 202310075796.6 discloses a "method for predicting rolling force in the production process of hot-rolled and finished plate and strip," which also utilizes the basic criterion of minimizing the total power functional of the rolling deformation zone, considering the iterative decoupling of rolling force and flattening radius to achieve real-time calculation of rolling force during continuous rolling. Chinese invention patent application number 201510423680.2 discloses a "calculation method for mutual iteration of rolling force and rolling temperature," which is based on a typical engineering rolling force model and uses numerical iteration to treat rolling force and rolling temperature as a whole to simultaneously improve the prediction accuracy of rolling force and rolling temperature. Chinese invention patent application number 202310796036.4 discloses a "rolling force prediction method for finishing mill based on SA-SCSO-TSVR algorithm". This patent provides a machine learning method for predicting rolling force and uses SCSO and SA algorithms to globally optimize model parameters, thereby achieving high-precision and rapid prediction of rolling force.

[0004] As can be seen from the patents above, they primarily revolve around methods based on engineering approaches, finite element simulation, and machine learning. However, the accuracy of the rolling force and torque predictions obtained by these methods is highly dependent on the thermophysical parameters of the rolled piece and the metal rheological constitutive model. The time cost of finite element models is also a significant factor hindering practical application. Furthermore, material databases obtained using various methods (physical experiments, literature reviews, theoretical calculations) may not accurately match the complex operating conditions in actual production processes. In addition, the pursuit of computational accuracy leads to complex model structures and high computational resource consumption, making many prediction models unsuitable for direct application in rolling production lines with low levels of automation. To further expand the application scenarios of rolling force and torque prediction models, especially for rolling production lines with low levels of automation (only basic level one automation), a prediction method with a sufficiently simple model structure, high model versatility, and strong robustness is needed. Summary of the Invention

[0005] The technical problem this invention aims to solve is to address the shortcomings of the prior art by providing an efficient online prediction method for rolling force and torque in strip rolling processes. This method estimates the rolling force and torque of subsequent unrolled passes based on the measured results of the rolling force and torque of the already rolled passes. It eliminates the need for prior establishment of any database related to the workpiece, mill, friction conditions, etc., and the model exhibits strong versatility and robustness. This overcomes the shortcomings of conventional methods that require the prior establishment of a stable and reliable material database to accurately characterize the current deformation resistance of the material, a task that is itself extremely time-consuming and labor-intensive.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: an efficient online prediction method for rolling force and torque in the strip rolling process, characterized in that the method includes the following steps: Step 1: Obtain the total number of rolling passes according to the rolling schedule. n , as well as the inlet and outlet thicknesses for each pass; Step 2: Allow the already rolled passes to... j =1, using the rolling mill's steel ejection signal to determine the first... j Has the rolling process been completed? If not, we must wait for it to finish. If a steel ejection signal has been received, we must collect and obtain the relevant data through the sensors. j The actual rolling parameters for each pass, including the entry thickness of the workpiece for that pass. Export thickness Actual rolling force and actual rolling torque ; Step 3: Let the intermediate variable k = j Calculate the first k +1 Correction factor for temperature drop effect during the unrolled sheet rolling process (pass 1) ; Step 4: Based on the rolled passes j Actual rolling information and the first k +1 rolling process temperature drop influence coefficient Calculate the first k +1 pass of rolling force not yet rolled and torque Predicted value; Step 5, Order k = k +1, continue calculating the rolling force and torque for subsequent passes until the predicted values ​​for all passes are calculated. k > n Output the calculation results of rolling force and torque for each pass; Step 6: Allow the rolled passes to... j = j +1, continue waiting for the next pass's steel-throwing signal. If the next pass's rolling is not yet complete, wait for the rolling to be completed. If the next pass's steel-throwing signal has been obtained, proceed to steps three to five, updating the rolling force and torque of subsequent passes based on the latest actual rolling data, until all passes are completed.

[0007] The above-mentioned method for online and efficient prediction of rolling force and torque in strip rolling process is characterized in that, in step three... k Correction factor for temperature drop effect in +1 pass of sheet rolling process Calculate using the following formula: ; In the formula, The coefficients to be determined are 0.8775, 0.11755, -0.003, and 1.1, respectively. The entry thickness of the most recently rolled sheet; For the first k +1 pass unrolled entry thickness. Undetermined coefficient in this invention. The specific application scenario of this invention needs to be determined based on the actual working conditions. Take values ​​of 0.8775, 0.11755, -0.003, and 1.1 respectively.

[0008] The above-mentioned method for online and efficient prediction of rolling force and torque in strip rolling process is characterized in that, in step four... k +1 pass rolling force and torque The predicted value is calculated using the following formula: ; ; ; ; In the formula, , , , and The first j The latest rolling pass includes the inlet thickness, outlet thickness, relative reduction, rolling force, and torque. , , , and The first k +1 pass: inlet thickness, outlet thickness, relative reduction, rolling force, and torque for the unrolled portion; and These are the corresponding stress state influence coefficients; The radius of the working roller; ~ The coefficients to be determined are taken as 0.8049, -0.3393, 0.2488, 0.0393, and 0.0732, respectively. These are the coefficients to be determined in this invention. ~ The specific application scenario of this invention needs to be determined based on the actual working conditions. ~ The values ​​were 0.8049, -0.3393, 0.2488, 0.0393, and 0.0732, respectively.

[0009] Compared with the prior art, the present invention has the following advantages: 1. This invention provides an efficient online prediction method for rolling force and torque in the strip rolling process. Its main function is to predict the rolling force and torque of subsequent unrolled passes based on the measured rolling force and torque of the already rolled passes. The biggest advantage of this method is that it does not require the prior establishment of any database related to the workpiece, rolling mill, friction conditions, etc. In other words, this method is independent of the grade and type of the rolled metal during use. The model has strong universality and robustness, which makes up for the shortcomings of existing conventional methods that require the prior establishment of a stable and reliable material database to accurately characterize the current deformation resistance of the material, and this work itself is extremely time-consuming and labor-intensive.

[0010] 2. The model structure provided by this invention is simple enough to be directly written into a PLC for execution without the need for process automation intervention, making it particularly suitable for rolling production lines with low levels of automation (no secondary process automation control system).

[0011] 3. The method of the present invention has been put into use in a titanium alloy plate rolling production line (with only basic automation level 1). Its prediction results are an important basis for the control of the thickness accuracy and shape quality of the rolled parts, as well as for the verification of rolling mill equipment, effectively ensuring the smooth progress of the rolling process and the stable quality of the rolled products.

[0012] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0013] Figure 1 This is a flowchart of an efficient online prediction method for rolling force and torque in the strip rolling process according to the present invention.

[0014] Figure 2 This is a comparison chart of the calculation results of rolling force and rolling torque in Embodiment 1 of the present invention.

[0015] Figure 3 This is a comparison chart of the calculation results of rolling force and rolling torque in Embodiment 2 of the present invention.

[0016] Figure 4 This is a comparison chart of the calculation results of rolling force and rolling torque in Embodiment 3 of the present invention. Detailed Implementation

[0017] Figure 1 This is a flowchart of an online efficient prediction method for rolling force and torque in the strip rolling process according to the present invention. Figure 1 As can be seen from the above, the present invention provides an efficient online prediction method for rolling force and torque in the strip rolling process, the steps of which are as follows: Step 1: Obtain the total number of rolling passes according to the rolling schedule. n , as well as the inlet and outlet thicknesses for each pass; Step 2: Allow the already rolled passes to... j =1, using the rolling mill's steel ejection signal to determine the first... j Has the rolling process been completed? If not, we must wait for it to finish. If a steel ejection signal has been received, we must collect and obtain the relevant data through the sensors. j The actual rolling parameters for each pass, including the entry thickness of the workpiece for that pass. Export thickness Actual rolling force and actual rolling torque ; Step 3: Let the intermediate variable k = j Calculate the first k +1 Correction factor for temperature drop effect during the unrolled sheet rolling process (pass 1) The first k Correction factor for temperature drop effect in +1 pass of sheet rolling process Calculate using the following formula: ; In the formula, The coefficients are undetermined and are taken as 0.8775, 0.11755, -0.003, and 1.1 respectively. The entry thickness of the most recently rolled sheet; For the first k +1 pass unrolled entry thickness; Step 4: Based on the rolled passes j Actual rolling information and the first k +1 rolling process temperature drop influence coefficient Calculate the first k +1 pass of rolling force not yet rolled and torque Predicted value; the first k +1 pass rolling force and torque The predicted value is calculated using the following formula: ; ; ; ; In the formula, , , , and The first j The latest rolling pass includes the inlet thickness, outlet thickness, relative reduction, rolling force, and torque. , , , and The first k +1 pass: inlet thickness, outlet thickness, relative reduction, rolling force, and torque for the unrolled portion; and These are the corresponding stress state influence coefficients; The radius of the working roller; ~ The coefficients are undetermined and are taken as 0.8049, -0.3393, 0.2488, 0.0393, and 0.0732 respectively. Step 5, Order k = k +1, continue calculating the rolling force and torque for subsequent passes until the predicted values ​​for all passes are calculated. k > nOutput the calculation results of rolling force and torque for each pass; Step 6: Allow the rolled passes to... j = j +1, continue waiting for the next pass's steel-throwing signal. If the next pass's rolling is not yet complete, wait for the rolling to be completed. If the next pass's steel-throwing signal has been obtained, proceed to steps three to five, updating the rolling force and torque of subsequent passes based on the latest actual rolling data, until all passes are completed.

[0018] Example 1 This embodiment takes the rolling of medium-thick plates in one pass on a single-stand reversible hot rolling mill (roll diameter 801mm) as an example. The raw material TC4 titanium alloy with a thickness of 200mm and a width of 1322mm is hot rolled in 8 passes to obtain a finished plate with a thickness of 34.5mm. The relevant rolling specifications during the hot rolling process are shown in Table 1. Table 1

[0019] This embodiment includes the following steps: Step 1: Obtain the total number of rolling passes according to the rolling schedule. n =8, the thickness of each pass's inlet and outlet is shown in Table 1; Step 2: When the first rolling pass is completed (i.e.) j =1) Perform the first calculation, and collect and obtain the actual rolling parameters for the first pass through relevant sensors, including the entry thickness of the workpiece. Export thickness Actual rolling force Actual average torque ; Step 3: Substitute the values ​​into the formula to calculate the correction coefficient for the temperature drop effect during rolling processes from the 2nd to the 8th passes, i.e. The calculation results are shown in Table 2; Step 4: Calculate the rolling force for passes 2 through 8 according to the formula. and rolling torque The calculation results are shown in Table 2; Table 2

[0020] Step 5: When the second rolling pass is completed (i.e.) j =1) Calculate again, repeating steps 2 to 4 to update the rolling force after the second pass. and rolling torque The calculation results are shown in Table 3; Table 3

[0021] Step 6: Repeat the above process to update the calculation results until all passes have been actually rolled. See details below. Figure 2 .

[0022] From Table 2, Table 3 and Figure 2 The results show that the temperature drop coefficient gradually increases with rolling, compensating for the change in material deformation resistance caused by the temperature drop of the plate. The overall calculation error is within an acceptable range. The rolling force error in the first calculation is within 4%, and the rolling torque error is within 16%. In the second update, the rolling force error is reduced to within 2%, and the rolling torque error is reduced to within 9%. In addition, it can be found that the calculation error increases with the increase of rolling passes. Although the temperature drop factor is considered, uncertainties such as rolling rhythm make the prediction results of passes further away from the actual rolling often have larger errors.

[0023] Example 2 This embodiment takes the two-pass hot rolling of a single-stand reversible hot rolling mill (roll diameter 806mm) as an example. Gr.2 pure titanium with a raw material thickness of 99mm and a width of 1350mm is hot rolled in 8 passes to obtain a finished plate with a thickness of 47.2mm. The relevant rolling parameters in the hot rolling process are shown in Table 4. Table 4

[0024] This embodiment includes the following steps: Step 1: Obtain the total number of rolling passes according to the rolling schedule. n =8, the thickness of each pass's inlet and outlet is shown in Table 4; Step 2: When the first rolling pass is completed (i.e.) j =1) Perform the first calculation, and collect and obtain the actual rolling parameters for the first pass through relevant sensors, including the entry thickness of the workpiece. Export thickness Actual rolling force Actual average torque ; Step 3: Substitute the values ​​into the formula to calculate the correction coefficient for the temperature drop effect during rolling processes from the 2nd to the 8th passes, i.e. The calculation results are shown in Table 5; Step 4: Calculate the rolling force for passes 2 through 8 according to the formula. and rolling torque The calculation results are shown in Table 5; Table 5

[0025] Step 5: The rolling process continues until the fourth pass has been completed (i.e., ...). j=4) Activate recalculation, repeat steps 2 to 4 to update the rolling force after the 4th pass. and rolling torque The calculation results are shown in Table 6; Table 6

[0026] Step 6: Repeat the above process to update the calculation results until all passes have been actually rolled. See details below. Figure 3 .

[0027] From Table 5, Table 6 and Figure 3 The results show that the maximum rolling force error was 17.4% and the maximum rolling torque error was 22.1% in the first calculation. By the fourth update, the maximum rolling force error had been reduced to less than 7.9% and the maximum rolling torque error had been reduced to less than 5.4%.

[0028] Example 3 This embodiment takes the composite plate rolling of a single-stand reversible hot rolling mill (roll diameter 817mm) as an example. The TA2 / Q235B / Q235B / TA2 titanium steel composite plate with a raw material thickness of 58mm and a width of 1050mm is hot rolled in 4 passes to obtain a finished plate with a thickness of 7.9mm. The relevant rolling parameters in the hot rolling process are shown in Table 7. Table 7

[0029] This embodiment includes the following steps: Step 1: Obtain the total number of rolling passes according to the rolling schedule. n =4, the thickness of each pass's inlet and outlet, the specific values ​​are shown in Table 7; Step 2: When the first rolling pass is completed (i.e.) j =1) Perform the first calculation, and collect and obtain the actual rolling parameters for the first pass through relevant sensors, including the entry thickness of the workpiece. Export thickness Actual rolling force Actual average torque ; Step 3: Substitute the values ​​into the formula to calculate the correction coefficient for the temperature drop effect during the rolling process in passes 2 to 4, respectively. The calculation results are shown in Table 8; Step 4: Calculate the rolling force for the 2nd to 4th passes according to the formula. and rolling torque The calculation results are shown in Table 8; Table 8

[0030] Step 5: The rolling process continues, repeating the above steps to update the calculation results until all passes have been rolled. See details below. Figure 4 .

[0031] From Table 7, Table 8 and Figure 4 The results show that the first calculation for rolling titanium-steel composite plates achieved a good prediction effect. The maximum rolling force error occurred in the fourth pass, which was only 7.1%, and the maximum rolling torque error was only 9.8%.

[0032] In summary, the above embodiments, by rolling plates of different materials and specifications such as pure titanium, titanium alloys, and titanium-steel composites, comprehensively evaluated the method proposed in this invention, further confirming the versatility, robustness, and accuracy of the prediction results of the method.

[0033] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention in any way. Any simple modifications, alterations, and equivalent changes made to the above embodiments based on the inventive essence shall still fall within the protection scope of the present invention.

Claims

1. A method for efficient online prediction of rolling force and torque in strip rolling process, characterized in that, The method includes the following steps: Step 1: Obtain the total number of rolling passes according to the rolling schedule. n , as well as the inlet and outlet thicknesses for each pass; Step 2: Allow the already rolled passes to... j =1, using the rolling mill's steel ejection signal to determine the first... j Has the rolling process been completed? If not, we must wait for it to finish. If a steel ejection signal has been received, we must collect and obtain the relevant data through the sensors. j The actual rolling parameters for each pass, including the entry thickness of the workpiece for that pass. Export thickness Actual rolling force and actual rolling torque ; Step 3: Let the intermediate variable k = j Calculate the first k +1 Correction factor for temperature drop effect during the unrolled sheet rolling process (pass 1) ; Step 4: Based on the rolled passes j Actual rolling information and the first k +1 rolling process temperature drop influence coefficient Calculate the first k +1 pass of rolling force not yet rolled and torque Predicted value; Step 5, Order k = k +1, continue calculating the rolling force and torque for subsequent passes until the predicted values ​​for all passes are calculated. k > n Output the calculation results of rolling force and torque for each pass; Step 6: Allow the rolled passes to... j = j +1, continue waiting for the next pass's steel-throwing signal. If the next pass's rolling is not yet complete, wait for the rolling to be completed. If the next pass's steel-throwing signal has been obtained, proceed to steps three to five, updating the rolling force and torque of subsequent passes based on the latest actual rolling data, until all passes are completed.

2. The method for online and efficient prediction of rolling force and torque in strip rolling process according to claim 1, characterized in that, The third step described k Correction factor for temperature drop effect in +1 pass of sheet rolling process Calculate using the following formula: ； In the formula, The coefficients are undetermined and are taken as 0.8775, 0.11755, -0.003, and 1.1 respectively. The entry thickness of the most recently rolled sheet; For the first k +1 pass unrolled entry thickness.

3. The method for online and efficient prediction of rolling force and torque in strip rolling process according to claim 1, characterized in that, The fourth step described k +1 pass rolling force and torque The predicted value is calculated using the following formula: ； ； ； ； In the formula, , , , and The first j The latest rolling pass includes the inlet thickness, outlet thickness, relative reduction, rolling force, and torque. , , , and The first k +1 pass: inlet thickness, outlet thickness, relative reduction, rolling force, and torque for the unrolled portion; and These are the corresponding stress state influence coefficients; The radius of the working roller; ~ The coefficients are undetermined and are taken as 0.8049, -0.3393, 0.2488, 0.0393, and 0.0732 respectively.

Images