Plate crown control device
The plate crown control device addresses inaccuracies in hot rolling by using a learning mechanism to adjust work roll settings based on actual measurements, improving accuracy and stability in achieving target plate crown.
Patent Information
- Application Number
- JP2024515977
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2022-04-20
- Publication Date
- 2025-09-09
- Estimated Expiration
- 2042-04-20
AI Technical Summary
Existing methods for controlling strip crown in hot rolling are limited by inaccuracies in setting calculations due to roll wear, thermal expansion, and complex alloy behaviors, leading to errors in predicting and achieving target plate crown, which can result in poor flatness and operational limits.
A plate crown control device that uses a learning mechanism to correct setting values by integrating actual measurements from a plate crown meter and flatness gauge, adjusting bending force and shift positions of work rolls based on learning tables and weighting coefficients to improve accuracy.
The device enhances the accuracy of work roll bending and curved roll shift, ensuring stable strip threading and improved yield by correcting setting values based on actual measurements, thereby achieving the target plate crown.
Smart Images

Figure 0007736177000031 
Figure 0007736177000032 
Figure 0007736177000033
Abstract
Description
[Technical Field]
[0001] The present invention relates to a plate crown control device, and more particularly to a learning and control method for a plate crown control device in hot rolling. [Background technology]
[0002] The requirements for thickness accuracy in the width direction of rolled materials are becoming stricter every year. In rolling operations that use a prediction model for the crown of a strip, it is becoming extremely important to improve the accuracy of such a prediction model.
[0003] To control the strip crown in this way, it is known to use mechanisms such as work roll benders and CVC (Continuous Variable Crown) roll shifters.
[0004] To achieve high-precision strip crown control, it is necessary to properly set the mechanisms of each stand. The set values of these mechanisms are calculated by a strip crown control device. Various strip crown prediction methods and set value calculation methods have been proposed, and efforts have been made to improve their accuracy.
[0005] The applicant is aware of the following documents as being relevant to the present invention: [Prior art documents] [Patent documents]
[0006] [Patent Document 1] Japanese Patent No. 3253013 [Patent Document 2] Japan Special Publication No. 3-72364 [Patent Document 3] Japanese Patent Publication No. 59-215205 [Non-patent literature]
[0007] [Non-Patent Document 1] Shusuke Yanagi, Masanori Ikeda, Hiroshi Kunii, "Development of a Sheet Crown Set-up Model for Aluminum Hot Rolling," Kobe Steel Technical Report, Vol. 58, No. 3 (Dec. 2008), p. 29 [Non-patent document 2] KNShohet, NATownsend “Roll bending methods of crown control in four-high plate mills”, J. The Iron and Steel Institute, (Nov. 1968) p1088 [Non-patent document 3] KNShohet, NATownsend “Flatness control in plate rolling”, J. The Iron and Steel Institute, (Oct. 1971) p769 Summary of the Invention [Problem to be solved by the invention]
[0008] In recent years, crown gauges that measure the thickness distribution in the width direction of the rolled material and flatness gauges that measure the shape of the rolled material have been introduced at the delivery side of the rolling mill. Dynamic feedback control that actively utilizes these has been adopted to manufacture high-quality rolled material.
[0009] However, dynamic feedback control using a plate crown meter or flatness meter cannot be applied to the leading edge of the rolled material, so the plate crown of the product at the leading edge depends on the accuracy of the settings of the mechanism of each stand by the plate crown control device at the start of rolling.
[0010] Therefore, the accuracy of setting the mechanism of each stand by the plate crown control device is improved by learning using a plate crown meter and a flatness meter. A widely used learning method is to correct the internal target value in the setting calculation for the next piece based on the actual measurement value of the plate crown meter.
[0011] For example, if the measured value of the plate crown is 10 microns larger than the target value of 40 microns, the internal target value for the next material setting calculation is 30 microns, which is the target value minus 10 microns.
[0012] However, this method involves calculating targets that differ from the original targets for later stands, which results in larger errors in the various prediction models and limits the accuracy of the calculations.
[0013] Furthermore, in Patent Document 1, the deviation between the actual crown value measured at the finishing delivery side and the predicted value is allocated to the mechanical sheet crown predicted value of each stand, thereby correcting the model error.
[0014] However, the actual crown prediction error at the finishing delivery side is simply divided equally among all stands or allocated according to the plate thickness ratio. Therefore, when the learning results are reflected in the setting calculation, there is a possibility that problems such as deterioration of the flatness at the delivery side of each stand may occur, or the bending force or shift position of each stand may reach its mechanical or operational limit, making it impossible to achieve the target plate crown.
[0015] Furthermore, Non-Patent Document 1 discloses a method of correcting each influence coefficient of a mechanical sheet crown model by taking the actual value of the bender corrected by the operator as the positive value. This method is based on the premise that there is almost no roll wear, as in aluminum hot rolling, and the behavior of the thermal expansion of the roll (thermal crown) is stable and can be predicted with a small error. Correction of the estimated error of the roll profile is not taken into consideration at all.
[0016] However, in steel rolling, the rolling temperature is high and various alloy components are included, so roll wear typically reaches several hundred microns. Furthermore, because various alloys with different rolling temperatures are rolled in a mixed order, thermal crown exhibits complex behavior. Therefore, even if parameters are adjusted by measuring roll temperatures, it is difficult to sufficiently reduce the prediction error. Therefore, even if the method described in Non-Patent Document 1 is applied to steel rolling, there is a problem in that the influence coefficient cannot be corrected with sufficient accuracy due to the influence of roll profile estimation errors.
[0017] Furthermore, in the methods disclosed in Patent Documents 2 and 3, the strip crown error detected by the strip crown meter is ultimately allocated to each stand using an allocation method determined solely by the genetic coefficient or the shape change coefficient, and then learned. In other words, this allocation does not take into account differences in the mechanical crown adjustment capacity of each stand due to the presence or absence of a CVC or a pair cross mechanism. Therefore, although the margin for inter-stand flatness is maximized, there is a problem in that stands with small mechanical crown adjustment capacity may reach the upper or lower limit of the mechanical crown, making it highly likely that the inter-stand flatness or the final stand delivery crown cannot be achieved.
[0018] The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a strip crown control device that can improve the setting accuracy of work roll bending, curved roll shift, etc., and achieve stable strip threading and improved yield. [Means for solving the problem]
[0019] The first aspect relates to a plate crown control device applied to a tandem rolling facility. Tandem rolling equipment continuously rolls the material to be rolled using multiple stands. The plate crown control device controls the tandem rolling equipment based on the delivery plate crown setting calculation value of each stand calculated by the setting calculation, the mechanical plate crown setting calculation value of each stand, and the setting values of the bending force and shift position of each stand. The plate crown controller includes at least one processor and a memory. The memory is a plate crown meter installed on the outlet side of the final stand of the plurality of stands for a measurement point determined in the longitudinal direction of the rolled material. to Actual measurement of outlet plate crown measured by Value Remember. The memory stores actual values including the rolling load, bending force, and shift position when the measurement point passes through each stand. The memory stores a learning table that stores a first learning value for each of the stands. The processor is configured to perform the setting calculation, the actual value calculation, the observed value calculation, the first learned value calculation, the first learning, and the setting calculation value correction. The actual value calculation is The thickness of the rolled material at the entry side of each stand, and the thickness of the rolled material at the exit side of each stand, Each of the above-mentioned measured values and, The actual mechanical plate crown calculation value of each stand is calculated based on the above. The observed value calculation calculates the mechanical plate crown observed value of each stand using the first learning weighting coefficient of each stand so that the observed value coincides with the actual measured value of the outlet plate crown of the final stand. The first learned value calculation calculates a first learned current value for each stand based on a difference between the mechanical plate crown observation value for each stand and the mechanical plate crown actual calculation value for each stand. In the first learning, the first learned value of each of the stands stored in the learning table is updated by proportionally dividing it between the first learned current value of each of the stands and a smoothing gain. The setting calculation value correction is performed by calculating the setting values of the bending force and shift position of each stand using the corrected mechanical plate crown setting calculation value of each stand, which is obtained by adding the first learning value of each stand read from the learning table to the mechanical plate crown setting calculation value of each stand in the setting calculation for the next material and thereafter. The learning table stores a second learning value for each of the stands. The calculation of the calculated performance value includes calculating an output sheet crown calculated performance value of each stand based on a mechanical sheet crown calculated performance value of each stand. The processor is configured to further execute a second learning value calculation that calculates a second learned current value of each stand using a difference between the delivery plate crown actual calculation value of the final stand and the delivery plate crown set calculation value of the final stand, a second learning weighting coefficient of each stand, which is calculated without using the delivery plate crown actual measurement value, and a transcription rate of each stand; and a second learning that updates the second learned value of each stand stored in the learning table by proportionally dividing it by the second learned current value of each stand and a smoothing gain. The setting calculation value correction is performed by calculating the setting values of the bending force and shift position of each stand using the corrected mechanical plate crown calculation value of each stand obtained by adding the first learning value of each stand and the second learning value of each stand read from the learning table to the mechanical plate crown setting calculation value of each stand in the setting calculation for the next material and thereafter. The first learning weighting coefficient of each stand is a value obtained by multiplying the smaller of a first plate crown ratio variable range of each stand, which is obtained by multiplying the mechanical plate crown variable range of each stand, calculated based on the bending force of each stand and the upper and lower limits of the shift position, by the transfer rate and dividing the result by the stand outlet plate thickness, and a second plate crown ratio variable range of each stand, which is calculated based on the flatness limit of each stand, by a correction coefficient. The second learning weighting coefficient of each stand is a value obtained by dividing the difference between the actual mechanical strip crown calculation value of each stand and the set mechanical strip crown calculation value of each stand by the delivery strip thickness of each stand.
[0020] The second aspect has the following characteristics in addition to the first aspect. The calculation of the calculated performance value includes calculating an output sheet crown calculated performance value of each stand based on a mechanical sheet crown calculated performance value of each stand. The observed value calculation calculates the observed value of the delivery plate crown of each stand by multiplying the deviation between the actual measured value of the delivery plate crown of the final stand and the calculated actual value of the delivery plate crown of the final stand by the first learning weighting coefficient and the plate thickness ratio of each stand, and adding the calculated actual value of the delivery plate crown of each stand. The observation value calculation calculates the mechanical sheet crown observation value of each stand using the entry sheet crown observation value of each stand, the delivery sheet crown observation value of each stand, a genetic coefficient, a transcription rate, and the ratio of entry-side to delivery-side sheet thickness of each stand.
[0021] The third aspect relates to a plate crown control device applied to a tandem rolling mill. Tandem rolling equipment continuously rolls the material to be rolled using multiple stands. The plate crown control device controls the tandem rolling equipment based on the delivery plate crown setting calculation value of each stand calculated by the setting calculation, the mechanical plate crown setting calculation value of each stand, and the setting values of the bending force and shift position of each stand. The plate crown controller includes at least one processor and a memory. The memory stores actual outlet plate crown measurements taken at measurement points determined in the longitudinal direction of the rolled material by a plate crown meter installed on the outlet side of the final stand of the plurality of stands. The memory stores actual values including the rolling load, bending force, and shift position when the measurement point passes through each stand. The memory stores a learning table that stores a first learning value for each of the stands. The processor is configured to perform the setting calculation, the actual value calculation, the observed value calculation, the first learned value calculation, the first learning, and the setting calculation value correction. The actual value calculation is performed by calculating the actual mechanical plate crown value of each stand based on the plate thickness of the rolled material at the inlet side of each stand, the plate thickness of the rolled material at the outlet side of each stand, and each of the actual measured values. The observed value calculation calculates the mechanical plate crown observed value of each stand using the first learning weighting coefficient of each stand so that the observed value coincides with the actual measured value of the outlet plate crown of the final stand. The first learned value calculation calculates a first learned current value for each stand based on a difference between the mechanical plate crown observation value for each stand and the mechanical plate crown actual calculation value for each stand. In the first learning, the first learned value of each of the stands stored in the learning table is updated by proportionally dividing it between the first learned current value of each of the stands and a smoothing gain. The setting calculation value correction is performed by calculating the setting values of the bending force and shift position of each stand using the corrected mechanical plate crown setting calculation value of each stand, which is obtained by adding the first learning value of each stand read from the learning table to the mechanical plate crown setting calculation value of each stand in the setting calculation for the next material and thereafter. The observation value calculation defines constraints, design variables, and an objective function, and calculates the design variables so as to minimize the objective function. The constraint condition is a sheet crown prediction formula for each stand. The design variables are the mechanical sheet crown observation value of each stand and the delivery sheet crown observation value of each stand except the final stand. The objective function is the sum of absolute values or sum of squares of values obtained by multiplying the deviation between the mechanical plate crown observation value of each stand and the mechanical plate crown actual calculation value of each stand by the first learning weighting coefficient of each stand.
[0022] The fourth aspect relates to a plate crown control device applied to a tandem rolling facility. Tandem rolling equipment continuously rolls the material to be rolled using multiple stands. The plate crown control device controls the tandem rolling equipment based on the delivery plate crown setting calculation value of each stand calculated by the setting calculation, the mechanical plate crown setting calculation value of each stand, and the setting values of the bending force and shift position of each stand. The plate crown controller includes at least one processor and a memory. The memory stores actual outlet plate crown measurements taken at measurement points determined in the longitudinal direction of the rolled material by a plate crown meter installed on the outlet side of the final stand of the plurality of stands. The memory stores actual values including the rolling load, bending force, and shift position when the measurement point passes through each stand. The memory stores a learning table that stores a first learning value for each of the stands. The processor is configured to perform the setting calculation, the actual value calculation, the observed value calculation, the first learned value calculation, the first learning, and the setting calculation value correction. The actual value calculation is performed by calculating the actual mechanical plate crown value of each stand based on the plate thickness of the rolled material at the inlet side of each stand, the plate thickness of the rolled material at the outlet side of each stand, and each of the actual measured values. The observed value calculation calculates the mechanical plate crown observed value of each stand using the first learning weighting coefficient of each stand so that the observed value coincides with the actual measured value of the outlet plate crown of the final stand. The first learned value calculation calculates a first learned current value for each stand based on a difference between the mechanical plate crown observation value for each stand and the mechanical plate crown actual calculation value for each stand. In the first learning, the first learned value of each of the stands stored in the learning table is updated by proportionally dividing it between the first learned current value of each of the stands and a smoothing gain. The setting calculation value correction is performed by calculating the setting values of the bending force and shift position of each stand using the corrected mechanical plate crown setting calculation value of each stand, which is obtained by adding the first learning value of each stand read from the learning table to the mechanical plate crown setting calculation value of each stand in the setting calculation for the next material and thereafter. The observation value calculation further calculates the mechanical plate crown variable range of each stand from the bending force and upper and lower limit values of the shift position of each stand. The observation value calculation further includes multiplying the mechanical strip crown variable range of each stand by a transfer rate and dividing the result by the stand outlet strip thickness to calculate a first strip crown ratio variable range of each stand. The first learning weight coefficient of each stand is determined so that the larger the variable range of the outlet plate crown ratio of each stand obtained by the smaller plate crown ratio variable range of the first plate crown ratio variable range and the second plate crown ratio variable range, the larger the learning weight coefficient.
[0023] The fifth aspect has the following characteristics in addition to the third aspect. The design variables include the plate crown ratio heritability coefficient or transfer rate for each stand. The constraint conditions include an inequality that the plate crown ratio genetic coefficient is larger in the rear stand than in the front stand, or that the transcription rate is smaller in the rear stand than in the front stand.
[0025] The sixth aspect relates to a plate crown control device applied to a tandem rolling facility. Tandem rolling equipment continuously rolls the material to be rolled using multiple stands. The plate crown control device controls the tandem rolling equipment based on the delivery plate crown setting calculation value of each stand calculated by the setting calculation, the mechanical plate crown setting calculation value of each stand, and the setting values of the bending force and shift position of each stand. The plate crown controller includes at least one processor and a memory. The memory stores actual outlet plate crown measurements taken at measurement points determined in the longitudinal direction of the rolled material by a plate crown meter installed on the outlet side of the final stand of the plurality of stands. The memory stores actual values including the rolling load, bending force, and shift position when the measurement point passes through each stand. The memory stores a learning table that stores second learning values for the respective stands. The processor is configured to perform the setting calculation, the actual value calculation, the observed value calculation, the first learned value calculation, the first learning, and the setting calculation value correction. The actual value calculation is performed by calculating the calculated mechanical sheet crown actual value and the calculated delivery sheet crown actual value of each stand based on the thickness of the rolled material at the entry side of each stand, the thickness of the rolled material at the exit side of each stand, and each of the actual measured values. The observed value calculation calculates the mechanical plate crown observed value of each stand using the first learning weighting coefficient of each stand so that the observed value coincides with the actual measured value of the outlet plate crown of the final stand. The second learning calculation value is calculated by using the difference between the actual delivery plate crown calculation value of the final stand and the set delivery plate crown calculation value of the final stand, a second learning weighting coefficient for each stand, which is calculated without using the actual delivery plate crown measurement value, and the transcription rate of each stand, to calculate the second learning current value of each stand. In the second learning, the second learned value of each of the stands stored in the learning table is updated by proportionally dividing it between the second learned current value of each of the stands and a smoothing gain. The setting calculation value correction calculates the setting values of the bending force and shift position of each stand using the corrected mechanical plate crown setting calculation value of each stand, which is obtained by adding the second learning value of each stand read from the learning table to the mechanical plate crown setting calculation value of each stand in the setting calculation for the next material and thereafter. The first learning weighting coefficient of each stand is a value obtained by multiplying the smaller of a first plate crown ratio variable range of each stand, which is obtained by multiplying the mechanical plate crown variable range of each stand, calculated based on the bending force of each stand and the upper and lower limit values of the shift position, by the transfer rate and dividing the result by the stand outlet plate thickness, and a second plate crown ratio variable range of each stand, which is calculated based on the flatness limit of each stand, by a correction coefficient. The second learning weighting coefficient of each stand is a value obtained by dividing the difference between the actual mechanical strip crown calculation value of each stand and the set mechanical strip crown calculation value of each stand by the delivery strip thickness of each stand. [Effects of the Invention]
[0026] According to the present invention, the observed mechanical plate crown value is estimated based on the actual measurement value measured by a plate crown meter installed on the delivery side of a tandem rolling mill, and the difference between the observed mechanical plate crown value and the calculated actual mechanical plate crown value can be learned. This learning result can be used to correct the predicted mechanical plate crown value (calculated set value) in the setting calculation for the next strip and thereafter. This can prevent problems such as deterioration of flatness on the delivery side of each stand and failure to achieve the target plate crown due to the bending force or shift position of each stand reaching mechanical or operational limits. As a result, the setting accuracy of work roll bending, curved roll shift, etc. can be improved, allowing the plate crown on the delivery side of the final stand to approach the target value, thereby achieving stable plate threading and improved yield. [Brief explanation of the drawings]
[0027] [Figure 1] 1 is a diagram for explaining a configuration example of rolling equipment according to an embodiment of the present invention; [Figure 2] FIG. 1 is a diagram showing the definition of plate crown and the configuration of a plate crown meter. [Figure 3] FIG. 2 is a diagram for explaining a work roll shift mechanism. [Figure 4] FIG. 2 is a diagram for explaining a work roll bending mechanism. [Figure 5] 1 is a block diagram illustrating an example of an outline of functions of a plate crown control device according to an embodiment of the present invention. [Figure 6] FIG. 10 is a diagram illustrating an example of design variables. [Figure 7] FIG. 10 is a diagram illustrating an example of design variables. [Figure 8] 1 is a graph showing the relationship between aspect ratio (plate width / plate thickness) and crown ratio genetic coefficient. [Figure 9] 1 is a block diagram showing an example of the hardware configuration of a plate crown control device according to an embodiment of the present invention; DETAILED DESCRIPTION OF THE INVENTION
[0028] Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. Note that elements common to the various drawings will be assigned the same reference numerals and redundant explanations will be omitted.
[0029] Embodiment 1. Overview of rolling equipment FIG. 1 is a diagram illustrating an example of the configuration of rolling equipment according to an embodiment of the present invention. The rolling equipment 1 is a tandem rolling equipment that continuously rolls a material 2 to be rolled using N stands (N is a natural number of 2 or more). As an example, the rolling equipment 1 is a finishing rolling mill group having seven stands (N=7), with stand numbers F1 to F7. The rolling equipment 1 is arranged in a hot rolling line.
[0030] At the exit of the final stand (Stand N), there is installed an X-ray thickness gauge 3, an X-ray crown gauge 4 (profile gauge), and an optical flatness gauge 5. However, because these measuring instruments are large and expensive, in many cases only one is installed at the exit of the final stand.
[0031] However, even if the entry thickness and delivery thickness of each stand (i-th stand (i=1, ..., N)) cannot be measured by the thickness meter 3, they can be estimated relatively accurately using a gauge meter method based on the roll gap, rolling load, and mill rigidity at each stand, and a mass flow method that utilizes the fact that the volumetric speed between each stand is the same. By using these methods to control the thickness, the thickness is controlled to match a predetermined target value, so in the present invention, these are treated as known.
[0032] FIG. 2 is a diagram showing the definition of the sheet crown and the configuration of the sheet crown total 4. As shown in FIG. 2, the central sheet thickness in the sheet width direction of the rolled material is defined as h c , the plate thickness at a given distance from the edge (for example, 40 mm) is h e Then, the plate crown is defined as the deviation.
[0033] The rolling mill in each stand in Figure 1 is a roll-shift type rolling mill with a pair of upper and lower work rolls 6, each with a predetermined crown curve for cross-sectional shape control. Each rolling mill also has a work roll shift mechanism 7 and a work roll bending mechanism 8 that move the upper and lower work rolls 6 in the axial direction.
[0034] The plate crown control device 10 is connected to a plate thickness gauge 3, a plate crown gauge 4, a flatness gauge 5, a work roll shift mechanism 7, a work roll bending mechanism 8, etc. The plate crown control device 10 determines the shift amount and bending force of each stand in order to obtain a predetermined cross-sectional shape of the rolled material.
[0035] Figure 3 is a diagram illustrating the work roll shifting mechanism 7. The work rolls 6 are ground to a contour with concaves and convexes expressed by a cubic or higher-order curve and are arranged in opposite directions. The work roll shifting mechanism 7 shifts the upper and lower work rolls 6 axially by the same distance in opposite directions, thereby changing the roll gap distribution and varying the crown of the rolled material. Figure 3 shows the work roll shift state according to three shift positions. The work rolls 6 are called CVC (Continuous Variable Crown) rolls, but are referred to as curved rolls here. The axial movement distance of the work rolls 6 is referred to as the shift amount. The average diameter of the upper and lower work rolls 6 at a certain axial position is referred to as the average roll diameter. The difference between the average roll diameter at the center and end of the roll at a certain shift position is referred to as the equivalent roll crown.
[0036] Figure 4 is a diagram illustrating the work roll bending mechanism 8. The work roll bending mechanism 8 applies vertical bending forces to the ends of the work rolls 6, thereby bending the work rolls 6 and changing the contact stress distribution between the work rolls 6 and the rolled material 2, thereby changing the crown of the rolled material 2. The bending force is generated by a hydraulic cylinder installed between the upper and lower work rolls 6 and acts on the work rolls 6 via axle boxes attached to the ends of the work rolls and their bearings. The hydraulic cylinders are located at both ends of the work rolls 6, and the sum of the forces generated by the two cylinders is called the bending force. As shown in Figure 4, increasing the bending force decreases the crown of the work rolls, while decreasing the bending force increases the crown. Generally, increasing the rolling load causes the rolls to deflect and the crown of the work rolls to increase. Therefore, the bending force must be increased appropriately as the rolling load increases.
[0037] 2. Plate crown control device 5 is a block diagram illustrating an outline of the functions of the plate crown control device 10. The plate crown control device 10 includes a setting calculation unit 51, an actual result calculation value calculation unit 52, an observation value calculation unit 53, a first learning value calculation unit 54, a first learning unit 55, a second learning value calculation unit 56, and a second learning unit 57.
[0038] 2-1.Setting calculation section The setting calculation unit 51 performs setting calculations before the rolled material 2 enters the finishing mill, calculating the shift position setting values and bending force setting values for the upper and lower work rolls 6 of each stand so that the desired sheet crown can be obtained at the finish exit and the rolled material will be flat between each stand. Generally, setting calculations involve numerically calculating rolling phenomena using a mathematical model to determine the rolling schedule. Specifically, this involves calculating the target dimensions and temperatures for each process, from the removal of the slab (base material) from the heating furnace to the completion of winding the hot-rolled coil (product), and then calculating the initial settings of each actuator (roll gap and initial shift position) to achieve these targets.
[0039] The plate crown control device 10 operates the work roll shift mechanism 7 to move the shift position of the work rolls 6 to the shift position setting value. Furthermore, after the rolled material 2 is bitten, the work roll bending mechanism 8 is operated to match the bending force to the bending force setting value. From the perspective of machine protection, the shift position of each stand is generally fixed during rolling.
[0040] Here, we will show how to calculate the set values in the setting calculation. It is known that the change in strip crown at the inlet and outlet of each stand is expressed by the following formula (1). The formula for predicting strip crown at the inlet and outlet of each stand shown in formula (1) is disclosed in, for example, Patent Document 2.
[0041]
number
[0042] i is the stand number, for example, 1 to 7 in the case of a 7-stand rolling mill. i is the plate thickness at the entry side of each stand. h i is the plate thickness at the delivery side. C Hi is the sheet crown of the rolled material 2 at the entry side (entry sheet crown). hi is the strip crown at the delivery side (delivery strip crown). C Mi is a mechanical plate crown.
[0043] Mechanical plate crown, also known as roll gap crown, is a plate crown determined by the elastic deformation of the rolling mill. Mechanical plate crown expresses the gap distribution between the upper and lower work rolls 6, i.e., the plate thickness distribution, when it is assumed that the widthwise distribution of the rolling load acting between the rolled material 2 and the work rolls 6 is uniform, as the difference between the plate thickness at the center position and the plate thickness at the edge in the width direction, i.e., plate crown.
[0044] The mechanical sheet crown can be theoretically determined by calculating the elastic deformation of the rolling mill based on rolling conditions including the roll dimensions and roll crown of the rolling mill, the sheet width, entry thickness, delivery thickness, rolling load, shift position, and bending force of the material 2 to be rolled. For example, Patent Document 2 discloses an analytically obtained approximation formula. Patent Document 3 discloses a solution method in which the roll axial direction is divided into many small regions.
[0045] From these, the mechanical plate crown is expressed by the following formula (2): Note that the diameter, length, elastic modulus, Poisson's ratio, etc. of the roll do not change, so they are omitted here.
[0046]
number
[0047] Here, P is the rolling load, F is the bending force, H is the entry thickness, and h is the delivery thickness. w is the equivalent roll crown of work roll 6. C b is the equivalent roll crown of the backup roll. The equivalent roll crown is affected by the shift position as well as thermal expansion and wear, and is expressed as the following function:
[0048]
number
[0049] where L wrsi indicates the shift position of the curve roll. C thrmi represents the crown change due to the thermal expansion of the work roll. weari represents the crown change due to work roll wear. bthrmi represents the crown change due to the thermal expansion of the backup roll. bweari represents the crown change due to backup roll wear.
[0050] (Plate crown ratio genetic coefficient) Also, η in Eq. (1) i indicates the plate crown ratio genetic coefficient.
[0051] Here, the plate crown ratio heredity coefficient is the ratio of the plate crown ratio at the entry side of each stand (entry plate crown ratio) (C hi / H i When only the strip crown ratio at the stand delivery side (delivery strip crown ratio) (C hi / h i )
[0052] It is known that the strip crown ratio heredity coefficient is expressed as a function value with parameters including geometric conditions such as strip width / thickness ratio and roll diameter (for example, Patent Document 2).
[0053] The reason why the exit sheet crown ratio is affected by the entry sheet crown ratio in this way is that a change in the entry sheet crown causes a slight difference in the widthwise distribution of elongation during rolling in that stand, which in turn changes the widthwise distribution of tension in the roll bite, which in turn changes the contact stress distribution between the roll and the rolled material, and therefore changes the elastic deformation of the roll.
[0054] (Transfer rate) In addition, ζ in the formula (1) represents the transfer rate.
[0055] Here, the transcription rate is the value (C Mi / h i When only the strip crown ratio at the stand delivery side (delivery strip crown ratio) (C hi / h i )
[0056] When the entrance and exit crown ratios of the sheet do not differ significantly, it is known that there is the following relationship between the sheet crown ratio heredity coefficient and the transcription rate (for example, Patent Document 2).
[0057]
number
[0058] (flatness limit) Furthermore, at the exit side of each stand, variations in the crown of the strip result in differences in longitudinal elongation strain depending on the position in the width direction. If the crown ratio (value of crown divided by thickness) at the entry side of the stand and the crown ratio at the exit side of the stand are the same, no difference in elongation strain occurs, and the rolled material 2 remains flat. Conversely, if the difference between the crown ratio at the entry side of the stand and the crown ratio at the exit side of the stand exceeds the allowable value (called the flatness limit), the rolled material buckles, resulting in edge waves or poor flatness due to intermediate elongation.
[0059] The flatness limit has been experimentally investigated, and for example, the following formula (5) is known (Non-Patent Document 3).
[0060]
number
[0061] where B i is the strip width at the exit of each stand, and in thin plate hot rolling, the results are almost the same even if the value converted from the product width to the hot dimension (dimension taking thermal expansion into account) is used.
[0062] The sheet crown ratio at the entrance of the first stand of the finishing mill is determined by the conditions at the roughing mill side, and is different from the target sheet crown ratio at the exit of the final stand of the finishing mill. For this reason, it is necessary to change the sheet crown ratio at some stands of the finishing mill so that the sheet crown at the exit of the final stand matches the target sheet crown. Therefore, based on equation (5), the sheet crown ratio is mainly changed in the front-stage stands, where the flatness limit is large, and the change in sheet crown ratio is made smaller in the rear-stage stands.
[0063] (Control based on setting calculation) In the setting calculation, from this point of view, the outlet plate crown C of each stand is hi Calculate the mechanical plate crown C of each stand by substituting it into equation (1). Mi Then, by substituting the calculated value into equations (2) and (3) and solving them inversely, the bending force and the shift position are determined.
[0064] There is a priority order for the operation of the bending force and the shift position. First, the bending force is set to an arbitrary value, and the shift position is operated. When this value reaches the mechanical or operational limit of movement, the shift position is fixed at the limit of movement and the bending force is operated.
[0065] However, when a large number of rolled materials 2 are continuously rolled, the work rolls 6 undergo thermal expansion and wear, making it difficult to maintain good accuracy in the setting calculations.
[0066] Therefore, in this embodiment, after the start of finish rolling, the plate crown control device 10 collects actual plate crown and flatness measurements at the final stand, measured by the plate crown meter 4 and flatness meter 5, at measurement points determined in the longitudinal direction of the rolled material 2, as well as actual values including the rolling load, bending force, and shift position when the measurement points passed through each stand. Based on these actual measurements and actual values, the plate crown control device 10 learns a correction amount (learned value) for the setting calculation, and corrects the calculated setting value with the learned correction value in the setting calculation for the next rolled material (next material). This learning improves the accuracy of the setting calculation for the next material and thereafter, making it possible to improve the plate crown accuracy of the rolled material.
[0067] The learning method of the present invention is shown below.
[0068] 2-2. Actual Calculation Value Calculation Section First, the actual result calculation unit 52 calculates the actual result of the mechanical plate crown of each stand.
[0069] The actual calculated value is a value calculated by applying the actual values collected during rolling as parameters to the same model formula used in the setting calculation. Below, they are distinguished by adding the subscript ACAL. From formula (2) and formula (3), the actual calculated value of mechanical sheet crown is calculated as follows:
[0070]
number
[0071] where P i ACT is the measured rolling load. F i ACT is the actual bending force measurement value. L wrsi ACT is the actual shift position value. C thrmi ACAL is the amount of change in roll crown due to thermal expansion of the work roll 6. C weari ACAL is the amount of change in roll crown due to wear of the work roll 6. C bthrmi ACAL is the change in roll crown due to thermal expansion of the backup roll. bweari ACAL is the amount of change in roll crown due to wear of the backup roll.
[0072] Substituting these into equation (1) and performing calculations sequentially from the upstream stand to the downstream stand, the calculated actual delivery strip crown value for each stand can be obtained.
[0073]
number
[0074] In this way, the actual calculation value calculation unit 52 calculates the calculated mechanical plate crown actual value C of each stand based on the above-mentioned actual measurement values and actual values. MiACAL and the calculated actual value of the exit plate crown for each stand, C hi ACAL Calculate.
[0075] 2-3. Observation value calculation section The observed value calculation unit 53 calculates the observed mechanical sheet crown value of each stand using the first learning weighting coefficient of each stand so that the observed value matches the actual measured value of the sheet crown at the outlet of the final stand. H1 SUP Starting from the calculated value (setup value) before rolling starts, adjust the exit mechanical crown C of each stand except the final stand so that the exit crown of the final stand matches the actual measured value. Mi , Exit plate crown C hi These are estimated values, and will be distinguished below by adding the subscript OBS. Observation C Mi OBS There are several ways to estimate this, and two calculation methods are shown here.
[0076] 2-3-1. First calculation method of observed values The first calculation method of the observed value (corresponding to claim 3) is a method based on the idea of maintaining the distribution of the outlet sheet crown of each stand as much as possible, since in the setting calculation, the outlet sheet crown of each stand is calculated so that the flatness between each stand is within the flatness limit after taking into account the above-mentioned constraints.
[0077] Therefore, the strip crown error measured at the delivery side of the final stand is allocated to each stand using the strip thickness ratio and the learning weight coefficient. As shown in the following equation (10), the observed value calculation unit 53 calculates the actual measured strip crown value C hN ACT and the calculated final stand exit plate crown value C hN ACAL The first learning weight coefficient w of each stand is used to calculate the deviation from i and thickness ratio (h i / H i ) and calculate the actual value of the exit plate crown for each stand, C hiACAL Adding this, the observed value of the outlet plate crown of each stand, C hi OBS Calculate.
[0078]
number
[0079] In addition, the first learning weight coefficient w i is the learning weighting coefficient for adjusting the distribution ratio of the strip crown ratio error measured at the final stand delivery side to the learning value of each stand. i How to determine this will be described later.
[0080] Equation (1) is transformed into equation (10) to obtain the observed inlet plate crown value C of each stand. Hi OBS (=C hi-1 OBS ), and the observed value of the outlet plate crown of each stand C hi OBS , and the mechanical plate crown observation value C of each stand Mi OBS Substituting this, we obtain the following equation (12).
[0081]
number
[0082] As shown in equation (12), the observation value calculation unit 53 calculates the inlet plate crown observation value C Hi OBS and the observed value of the outlet plate crown of each stand, C hi OBS and the heritability coefficient η i and the transcription rate ζ i and the ratio of the thickness of the entry side to the exit side of each stand (h i / H i ) and the mechanical plate crown observation value C Mi OBS Calculate.
[0083] 2-3-2. Second calculation method of observed values The second calculation method of the observed value (corresponding to claims 4 and 6) is based on the idea of trying to maintain as much of the distribution as possible, since the optimal combination of outlet plate crown and mechanical plate crown for each stand is calculated in the setting calculation taking into account the above-mentioned constraints.
[0084] In this calculation method, the observation value calculation unit 53 calculates the mechanical plate crown observation value C of each stand. Mi OBS , and the observed value of the exit plate crown of each stand except the last stand, C hi OBS In calculating the above, the constraints, design variables, and objective function are defined as follows, and the design variables are calculated so as to minimize the objective function.
[0085] The constraint is the prediction formula for the strip crown at the inlet and outlet of each stand (i = 1, ..., N) shown in Equation (1). The strip crown between each stand in Equation (1) is the observed value C hi OBS The observed value C Mi OBS Equation (13A) is the constraint for the first stand, equation (13B) is the constraint for a stand that is neither the first nor the last stand, and equation (13C) is the constraint for the last stand. As shown in Figure 6, the mechanical blade crown observation value C Mi OBS and the observed crown value C of the delivery plate at each stand except the last stand hi OBS are used as design variables. That is, the number of design variables is (N×2-1), where N is the number of stands. The entry plate crown of the first stand is set to the calculated value C H1 SUP , the final stand's exit plate crown is measured C hN ACT is used.
[0086]
number
[0087] In addition to the plate crown and mechanical plate crown of each stand, as shown in Figure 7, the crown ratio genetic coefficient η i or transcription rate ζ i When including in the design variables, either the inequality of formula (13D) or formula (13E) is added to the constraints (corresponding to claim 6). Formula (13D) shows that the rear stand has a higher plate crown ratio heredity coefficient η i Equation (13E) shows that the transfer rate ζ is larger in the rear stand than in the front stand. i This shows that is small.
[0088]
number
[0089] This is because the crown ratio genetic coefficient η i is known to be a function value with the horizontal axis representing the aspect ratio (strip width / strip thickness), and it increases as the strip thickness becomes thinner in the later stands.
[0090] The objective function Φ is either Equation (14) or Equation (15) shown below. That is, the mechanical plate crown observation value C Mi OBS and the calculated mechanical plate crown performance value C of each stand Mi ACAL The deviation from the first learning weight coefficient w i The sum of the absolute values or the sum of squares of the values multiplied by is used as the objective function.
[0091]
number
[0092] In addition, the first learning weight coefficient w i is a learning weighting coefficient for adjusting the allocation ratio of the strip crown ratio error actually measured at the delivery side of the final stand to the learning value of each stand, and may be set to an arbitrary value or calculated using the method described below.
[0093] The observation value calculation unit 53 calculates each design variable so as to minimize the objective function Φ. As a result, the mechanical plate crown observation value C Mi OBS Equations (13A) to (13C) are obtained by Mi OBS Since this is a linear equation, linear programming such as the Simplex method can be used for this calculation when the objective function is equation (11). Also, quadratic programming can be used when the objective function is equation (12). According to equation (11), the calculation load is low and calculation can be performed in a short time. On the other hand, according to equation (12), the mechanical plate crown observation value C Mi OBS It may be possible to calculate a more reasonable allocation for each stand.
[0094] 2-4. First learning value calculation unit and first learning unit The mechanical plate crown observation value C of each stand obtained by the first calculation method (2-3-1) or the second calculation method (2-3-2) of the observation value described above Mi OBS The first learning value calculation unit 54 calculates the mechanical plate crown observation value C of each stand using the Mi OBS and the calculated mechanical plate crown performance value C of each stand Mi ACAL Based on the difference between the first learning current value Z MOi CUR Calculate (Current value: CUR).
[0095]
number
[0096] The first learning unit 55 calculates the first learning value Z of each stand stored in the learning table (steel type × plate thickness × plate width). MOi OLD (OLD value) and the first learning current value Z of each stand MOi CUR The new first learned value Z is updated as shown in the following equation (17). MOi NEW (NEW value) is written back to the learning table.
[0097]
number
[0098] where α MOi is the smoothing gain (0≦α MOi ≦1).
[0099] First learning value Z MOi mainly includes correction amounts for the estimation error of the mechanical plate crown change due to the rolling load and bending force, and the estimation error of the roll thermal expansion and wear.
[0100] 2-5. How to determine the first learning weighting coefficient The mechanical plate crown observation value C of each stand is calculated using the first calculation method (2-3-1) or the second calculation method (2-3-2) of the observation value described above. Mi OBS In obtaining this, the first learning weight coefficient w is used to adjust the allocation of the plate crown ratio error measured at the final stand to the first learning value of each stand. i There are different ways to decide this:
[0101] The first is the first learning weight coefficient w for all stands (i=1,...,N). i to the same value (e.g., w iThis method of determination makes it possible to distribute the strip crown ratio error measured at the final stand to the learned values of each stand nearly evenly, making it easier for the operator to understand the changes in bending force and shift position due to learning, and therefore easier for them to intervene.
[0102] The second is that the larger the variable range of the outlet plate crown ratio of each stand, the larger the first learning weight coefficient w i The bending force and shift position of each stand are set to upper and lower limits due to machine specifications and operational reasons. Substituting these upper and lower limits into formulas (2) and (3), the upper limit C of the mechanical plate crown is obtained. Mi MAX and lower limit C Mi MIN By substituting these upper and lower limit values into equation (1) to obtain the difference between them (mechanical strip crown variable range), multiplying it by the transfer rate, and dividing it by the stand exit strip thickness, the first strip crown ratio variable range Δγ hi CHG1 Equation (18) is obtained.
[0103]
number
[0104] On the other hand, the maximum and minimum values of the crown ratio change that are within the flatness limit can be obtained by equation (5), and the difference between these values gives the second plate crown ratio variable range Δγ hi CHG2 Equation (19) is obtained.
[0105]
number
[0106] The smaller of the variable ranges of the strip crown ratios in equation (18) and equation (19) is used to obtain the variable range of the strip crown ratio at the exit of each stand. The larger the variable range of the strip crown ratio at the exit of each stand, the larger the learning weight coefficient w i The variable range of the plate crown ratio is calculated under typical operating conditions, and the first learning weight coefficient w i can be decided.
[0107] In addition, the first learning weighting coefficient w i If the first plate crown ratio variable range Δγ is determined by calculation, a more appropriate value can be obtained. hi CHG1 and the second plate crown ratio variable range Δγ hi CHG2 The smaller of the two is the correction coefficient k i Multiplying by the first learning weight coefficient w i Calculate.
[0108]
number
[0109] Here, the stand-specific correction coefficient k i and c i is an arbitrary constant, usually k i is the same value for all stands, c i is set to zero, but may be fine-tuned depending on the operating conditions.
[0110] In addition, the first learning weight coefficient w i As shown in equations (10) and (11), w TOTAL It is normalized by
[0111] According to this method, the sheet crown ratio error measured at the final stand is allocated to each stand's learning value in accordance with the sheet crown ratio variable range of each stand, which has the advantage that problems such as the sheet crown ratio variable range of one stand becoming full, the shift position and bending force of that stand reaching their upper and lower limits, or the flatness of the stand's outlet side deteriorating are unlikely to occur.
[0112] 2-6. Second learning value calculation unit and second learning unit Next, the second learning (corresponding to claim 2) will be explained. In the second learning, the actual calculated value C hN ACAL and the calculated value C hN SUP The difference between the thickness ratio (h i / h N ) and the second learning weight coefficient u i The amount of mechanical plate crown correction required to correct the plate crown error at the exit of each stand is expressed by the following equation (21A) using the transfer rate, and this is the second learned current value (Current value: CUR) for the mechanical plate crown setting value.
[0113]
number
[0114] As shown in equation (21A), the second learning value calculation unit 56 calculates the delivery plate crown actual value C hN ACAL and the final stand's calculated crown setting value C hN SUP and the second learning weight coefficient u i and the transcription rate ζ of each stand i Using this, the second learning current value Z of each stand MAi CUR Calculate.
[0115] In addition, the second learning weight coefficient u i In the second learning, the actual calculated value of the final stand's exit plate crown C hN ACAL and the calculated value C hN SUP is the learning weighting coefficient for adjusting the distribution ratio of the difference between the learning value of each stand and the learning value of each stand. i How to determine this will be described later.
[0116] The second learning unit 57 calculates the second learning value Z of each stand stored in the learning table (steel type × plate thickness × plate width). MAi OLD (OLD value) and the second learning current value Z of each stand MAi CUR The new second learned value Z MAi NEW (NEW value) is written back to the learning table.
[0117]
number
[0118] where α MAi is the smoothing gain (0≦α MAi ≦1).
[0119] Second learning value Z MAi The second learning value Z contains the difference between the shift position and bending force at the time of setting calculation and learning, and mainly includes the amount of correction due to manual intervention by the operator. Since it is highly likely that the operator manually intervenes after visually checking the condition of each stand, MAi It is not appropriate to distribute this to other stands, but it is appropriate to distribute it mainly to the learning value of the stand in question.
[0120] Therefore, the second learning weight coefficient u for each stand i is the calculated value of the mechanical plate crown of each stand, C Mi ACAL and the calculated value C Mi SUPIt is preferable to determine it based on the difference between
[0121]
number
[0122] That is, the second learning weight coefficient u of the i-th stand i is the calculated mechanical plate crown value C of the i-th stand Mi ACAL and the calculated mechanical plate crown setting value C of the i-th stand Mi SUP The difference between this and the delivery thickness h i This corresponds to claim 7. i The value of is allocated largely to the area where the difference between the actual calculated value and the set calculated value is large.
[0123] 2-7. Setting calculation section (setting calculation for subsequent materials) Next, the correction in the setting calculation for the next or subsequent slab will be described. As shown in the formula (24), the setting calculation unit 51 corrects the mechanical plate crown setting calculation value C Mi SUP The first learning value Z of each stand read from the learning table MOi NEW and the second learning value Z of each stand MAi NEW The calculated setting value correction 51a is executed to calculate the corrected calculated mechanical plate crown setting value of each stand to which the above-mentioned value is added.
[0124]
number
[0125] As a result, the calculated mechanical sheet crown setting values for each stand are corrected based on the learning results, and by substituting these values into equations (2) and (3) and solving them inversely, the setting values for the bending force and shift position are corrected, allowing the finishing delivery sheet crown to approach the target value.
[0126] 3.Effects As described above, the sheet crown control device 10 according to this embodiment estimates the mechanical sheet crown observation value using the crown variable range or quadratic programming based on the actual measurement value measured by the sheet crown meter 4 located on the exit side of the final stand, and can learn the difference between the mechanical sheet crown observation value and the actual calculated mechanical sheet crown value. Using this learning result (first learned value), the predicted mechanical sheet crown value (calculated setting value) can be corrected in the setting calculation for the next sheet or later.
[0127] This can prevent problems such as deterioration of flatness on the delivery side of each stand and failure to achieve the target sheet crown due to the bending force or shift position of each stand reaching mechanical or operational limits. As a result, the accuracy of setting work roll bending, curved roll shift, etc. can be improved, allowing the sheet crown on the delivery side of the final stand to approach the target value, realizing stable sheet threading and improved yield.
[0128] Furthermore, according to the strip crown control device 10 of this embodiment, as expressed in equation (20), it is possible to learn the strip crown model of each stand so as to ensure a margin by taking into account both the mechanical strip crown variable range and the shape dead zone of each stand. This makes it less likely that the upper or lower limit of the mechanical crown will be reached, and provides the advantage of increasing the possibility that the inter-stand flatness and the final stand delivery crown can be made to match the target values.
[0129] Furthermore, the strip crown control device 10 according to this embodiment can learn the difference between the calculated actual value of the delivery strip crown and the calculated target value of the delivery strip crown at the final stand (second learned value). This also makes it possible to learn the results of corrections to the bending force and shift position by the operator.
[0130] 4. Variations In the above-described embodiment, when there is little manual intervention by the operator, the second learned value Z MAiThe calculation can be omitted (corresponding to claim 1). MOi It is also possible to omit this calculation (corresponding to claim 8).
[0131] In addition, when a strip crown meter is installed on the delivery side of a stand other than the final stand, the present invention can be applied by reading the final stand as the stand where the strip thickness meter is installed and each stand as each stand upstream of the strip thickness meter.
[0132] Furthermore, when a plate crown control device 10 other than the work roll shift mechanism 7, such as a pair cross device, a VC roll device, or an intermediate roll shift device of a six-high rolling mill, is used, the present invention can be applied by replacing the shift position with the operation amount of the actuator of each device.
[0133] In addition, in the case of a reversing mill that rolls multiple times (passes) while changing the rolling direction in one stand, the present invention can be applied by reading "stand" as "pass."
[0134] 5. Hardware configuration example 9 is a conceptual diagram showing an example of the hardware configuration of the processing circuit of the plate crown control device 10 in the above-described embodiment. The functions of the plate crown control device 10 are realized by the processing circuit. In one aspect, the processing circuit includes at least one processor 91 and at least one memory 92. In another aspect, the processing circuit includes at least one dedicated hardware 93.
[0135] When the processing circuit includes a processor 91 and a memory 92, each function is realized by software, firmware, or a combination of software and firmware. At least one of the software and firmware is written as a program. At least one of the software and firmware is stored in the memory 92. The processor 91 realizes each function by reading and executing the program and various data stored in the memory 92. The various data stored in the memory 92 include actual measurement values measured by the plate crown meter 4 and flatness meter 5 described above. The data also includes actual values such as the rolling load, bending force, and shift position of each stand. The memory 92 also includes a learning table that stores the first learning value and the second learning value of each stand.
[0136] If the processing circuitry comprises dedicated hardware 93, the processing circuitry may be, for example, a single circuit, multiple circuits, a programmed processor, or a combination thereof. Each function is implemented by the processing circuitry.
[0137] Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments and can be implemented in various modifications without departing from the spirit of the present invention. When the numbers, quantities, amounts, ranges, etc. of each element are mentioned in the above-described embodiments, the present invention is not limited to the mentioned numbers unless otherwise specified or clearly specified in principle. Furthermore, the structures, etc. described in the above-described embodiments are not necessarily essential to the present invention unless otherwise specified or clearly specified in principle. [Explanation of symbols]
[0138] 1. Rolling equipment 2. Rolled material 3 Thickness gauge 4-plate crown total 5 Flatness meter 6 Work Rolls 7 Work roll shift mechanism 8 Work roll bending mechanism 10 Plate crown control device 51 Setting calculation section 51a Setting calculation value correction 52 Actual calculation value calculation section 53 Observation value calculation section 54 First learning value calculation unit 55 First Study Section 56 Second learning value calculation unit 57 Second Study Section 91 processors 92 memory 93 Hardware
Claims
1. A plate crown control device that controls a tandem rolling facility that continuously rolls a rolled material in a plurality of stands, based on a delivery plate crown setting calculation value of each stand calculated by setting calculation, a mechanical plate crown setting calculation value of each stand, and setting values of a bending force and a shift position of each stand, at least one processor and a memory; The memory includes: An actual delivery crown value measured by a plate crown meter installed on the delivery side of the final stand of the plurality of stands at a measurement point determined in the longitudinal direction of the rolled material; Actual values including rolling load, bending force, and shift position when the measurement point passes through each stand; a learning table that stores the first learning value of each stand; The processor: an actual value calculation for calculating an actual mechanical sheet crown calculated value of each stand based on the thickness of the rolled material at the inlet side of each stand, the thickness of the rolled material at the outlet side of each stand, and each of the actual values; an observation value calculation for calculating an observation value of a mechanical sheet crown of each stand using a first learning weighting coefficient of each stand so that the observation value of a mechanical sheet crown of each stand coincides with the actual measurement value of the delivery sheet crown of the last stand; a first learning value calculation for calculating a first learning current value of each stand based on a difference between the mechanical plate crown observation value of each stand and the mechanical plate crown actual calculation value of each stand; a first learning for updating the first learned value of each of the stands stored in the learning table by dividing the first learned current value of each of the stands proportionally by a smoothing gain; and in the setting calculation for the next material and thereafter, a setting calculation value correction is performed to calculate the setting values of the bending force and the shift position of each stand using the corrected mechanical plate crown setting calculation value of each stand obtained by adding the first learning value of each stand read out from the learning table to the mechanical plate crown setting calculation value of each stand, the learning table stores a second learning value for each of the stands; the calculating of the calculated performance values includes calculating an output sheet crown calculated performance value of each stand based on a mechanical sheet crown calculated performance value of each stand, The processor: a second learning value calculation for calculating a second learned current value for each stand using a difference between the delivery sheet crown actual calculation value of the final stand and the delivery sheet crown set calculation value of the final stand, a second learning weighting coefficient for each stand, the second learning weighting coefficient being calculated without using the delivery sheet crown actual measurement value, and a transcription rate for each stand; a second learning process for updating the second learned value of each of the stands stored in the learning table by proportionally dividing the second learned current value of each of the stands by a smoothing gain, The setting calculation value correction is to calculate setting values of the bending force and the shift position of each stand using a corrected mechanical plate crown calculation value of each stand obtained by adding the first learning value of each stand and the second learning value of each stand read out from the learning table to the mechanical plate crown setting calculation value of each stand in the setting calculation for the next material and thereafter, the first learning weighting coefficient of each stand is a value obtained by multiplying the smaller of a first plate crown ratio variable range of each stand, which is obtained by multiplying the mechanical plate crown variable range of each stand, calculated based on the bending force of each stand and the upper and lower limit values of the shift position, by the transfer rate and dividing the result by the stand outlet plate thickness, and a second plate crown ratio variable range of each stand, which is calculated based on the flatness limit of each stand, by a correction coefficient; the second learning weighting coefficient of each stand is a value obtained by dividing the difference between the actual mechanical strip crown calculation value of each stand and the set mechanical strip crown calculation value of each stand by the delivery strip thickness of each stand; A plate crown control device characterized by:
2. the calculating of the calculated performance values includes calculating an output sheet crown calculated performance value of each stand based on a mechanical sheet crown calculated performance value of each stand, The observation value calculation is calculating an observed value of the delivery plate crown of each stand by multiplying the deviation between the actual measurement value of the delivery plate crown of the final stand and the calculated actual value of the delivery plate crown of the final stand by the first learning weighting coefficient and the plate thickness ratio of each stand and adding the result to the calculated actual value of the delivery plate crown of each stand; calculating the mechanical sheet crown observation value of each stand using the entry sheet crown observation value of each stand, the delivery sheet crown observation value of each stand, a genetic coefficient, a transcription rate, and a ratio of entry-side to delivery-side sheet thickness of each stand; 2. The plate crown control device according to claim 1, wherein:
3. A plate crown control device that controls a tandem rolling facility that continuously rolls a rolled material in a plurality of stands, based on a delivery plate crown setting calculation value of each stand calculated by setting calculation, a mechanical plate crown setting calculation value of each stand, and setting values of a bending force and a shift position of each stand, at least one processor and a memory; The memory includes: An actual delivery crown value measured by a plate crown meter installed on the delivery side of the final stand of the plurality of stands at a measurement point determined in the longitudinal direction of the rolled material; Actual values including rolling load, bending force, and shift position when the measurement point passes through each stand; a learning table that stores the first learning value of each stand; The processor: an actual value calculation for calculating an actual mechanical sheet crown calculated value of each stand based on the thickness of the rolled material at the inlet side of each stand, the thickness of the rolled material at the outlet side of each stand, and each of the actual values; an observation value calculation for calculating an observation value of a mechanical sheet crown of each stand using a first learning weighting coefficient of each stand so that the observation value of a mechanical sheet crown of each stand coincides with the actual measurement value of the delivery sheet crown of the last stand; a first learning value calculation for calculating a first learning current value of each stand based on a difference between the mechanical plate crown observation value of each stand and the mechanical plate crown actual calculation value of each stand; a first learning for updating the first learned value of each of the stands stored in the learning table by dividing the first learned current value of each of the stands proportionally by a smoothing gain; and in the setting calculation for the next material and thereafter, a setting calculation value correction is performed to calculate the setting values of the bending force and the shift position of each stand using the corrected mechanical plate crown setting calculation value of each stand obtained by adding the first learning value of each stand read out from the learning table to the mechanical plate crown setting calculation value of each stand, The observation value calculation is The sheet crown prediction formula for each stand is used as a constraint condition, the mechanical sheet crown observation value of each stand and the delivery sheet crown observation value of each stand excluding the final stand are set as design variables, The sum of absolute values or sum of squares of values obtained by multiplying the deviation between the mechanical plate crown observation value of each stand and the mechanical plate crown actual calculation value of each stand by the first learning weighting coefficient of each stand is used as an objective function, calculating each of the design variables so as to minimize the objective function; the first learning weighting coefficient of each stand is a value obtained by multiplying the smaller of a first plate crown ratio variable range of each stand, which is obtained by multiplying a mechanical plate crown variable range of each stand, calculated based on the bending force of each stand and the upper and lower limit values of the shift position, by a transfer rate and dividing the result by the stand outlet plate thickness, and a second plate crown ratio variable range of each stand, which is calculated based on the flatness limit of each stand, by a correction coefficient; A plate crown control device characterized by:
4. A plate crown control device that controls a tandem rolling facility that continuously rolls a rolled material in a plurality of stands, based on a delivery plate crown setting calculation value of each stand calculated by setting calculation, a mechanical plate crown setting calculation value of each stand, and setting values of a bending force and a shift position of each stand, at least one processor and a memory; The memory includes: An actual delivery crown value measured by a plate crown meter installed on the delivery side of the final stand of the plurality of stands at a measurement point determined in the longitudinal direction of the rolled material; Actual values including rolling load, bending force, and shift position when the measurement point passes through each stand; a learning table that stores the first learning value of each stand; The processor: an actual value calculation for calculating an actual mechanical sheet crown calculated value of each stand based on the thickness of the rolled material at the inlet side of each stand, the thickness of the rolled material at the outlet side of each stand, and each of the actual values; an observation value calculation for calculating an observation value of a mechanical sheet crown of each stand using a first learning weighting coefficient of each stand so that the observation value of a mechanical sheet crown of each stand coincides with the actual measurement value of the delivery sheet crown of the last stand; a first learning value calculation for calculating a first learning current value of each stand based on a difference between the mechanical plate crown observation value of each stand and the mechanical plate crown actual calculation value of each stand; a first learning for updating the first learned value of each of the stands stored in the learning table by dividing the first learned current value of each of the stands proportionally by a smoothing gain; and in the setting calculation for the next material and thereafter, a setting calculation value correction is performed to calculate the setting values of the bending force and the shift position of each stand using the corrected mechanical plate crown setting calculation value of each stand obtained by adding the first learning value of each stand read out from the learning table to the mechanical plate crown setting calculation value of each stand, The processor further includes, in the observation value calculation, calculating a variable range of the mechanical plate crown of each stand from the bending force of each stand and the upper and lower limits of the shift position; calculating a first sheet crown ratio variable range for each stand by multiplying the mechanical sheet crown variable range for each stand by a transfer rate and dividing the result by the stand outlet sheet thickness; Calculate a second plate crown ratio variable range for each stand based on the flatness limit of each stand; the first learning weight coefficient of each stand is determined so that the learning weight coefficient increases as the variable range of the delivery sheet crown ratio of each stand obtained by the smaller sheet crown ratio variable range of the first sheet crown ratio variable range and the second sheet crown ratio variable range increases; A plate crown control device characterized by:
5. The design variables include a plate crown ratio heritability coefficient or transcription rate of each stand, The constraint condition includes an inequality that the plate crown ratio genetic coefficient is larger in the rear stand than in the front stand, or the transcription rate is smaller in the rear stand than in the front stand.
4. The plate crown control device according to claim 3, wherein:
6. A plate crown control device that controls a tandem rolling facility that continuously rolls a rolled material in a plurality of stands, based on a delivery plate crown setting calculation value of each stand calculated by setting calculation, a mechanical plate crown setting calculation value of each stand, and setting values of a bending force and a shift position of each stand, at least one processor and a memory; The memory includes: An actual delivery crown value measured by a plate crown meter installed on the delivery side of the final stand of the plurality of stands at a measurement point determined in the longitudinal direction of the rolled material; Actual values including rolling load, bending force, and shift position when the measurement point passes through each stand; a learning table that stores the second learning value of each stand; The processor: an actual value calculation for calculating an actual mechanical sheet crown calculated value and an actual delivery sheet crown calculated value of each stand based on the thickness of the rolled material at the entry side of each stand, the thickness of the rolled material at the delivery side of each stand, and each of the actual values; an observation value calculation for calculating an observation value of a mechanical sheet crown of each stand using a first learning weighting coefficient of each stand so that the observation value of a mechanical sheet crown of each stand coincides with the actual measurement value of the delivery sheet crown of the last stand; a second learning value calculation for calculating a second learned current value for each stand using a difference between the delivery sheet crown actual calculation value of the final stand and the delivery sheet crown set calculation value of the final stand, a second learning weighting coefficient for each stand, the second learning weighting coefficient being calculated without using the delivery sheet crown actual measurement value, and a transcription rate for each stand; a second learning for updating the second learned value of each of the stands stored in the learning table by proportionally dividing the second learned current value of each of the stands and a smoothing gain; and in the setting calculation for the next material and thereafter, a setting calculation value correction is performed to calculate the setting values of the bending force and the shift position of each stand using the corrected mechanical plate crown setting calculation value of each stand obtained by adding the second learning value of each stand read out from the learning table to the mechanical plate crown setting calculation value of each stand, the first learning weighting coefficient of each stand is a value obtained by multiplying the smaller of a first plate crown ratio variable range of each stand, which is obtained by multiplying the mechanical plate crown variable range of each stand, calculated based on the bending force of each stand and the upper and lower limit values of the shift position, by the transcription rate and dividing the result by the stand outlet plate thickness, and a second plate crown ratio variable range of each stand, which is calculated based on the flatness limit of each stand, by a correction coefficient; the second learning weighting coefficient of each stand is a value obtained by dividing the difference between the actual mechanical strip crown calculation value of each stand and the set mechanical strip crown calculation value of each stand by the delivery strip thickness of each stand; A plate crown control device characterized by:
Citation Information
Patent Citations
Method for controlling and setting crown and shape in sheet rolling
JP1984215205A
Production of screen process printing stencil
JP1991072364A
Learning control method in process lens
JP1992367901A
Method for measuring plate crown of continuous casting machine, and its control method
JP1997234506A
Plate crown / Shape controlling method in hot rolling
JP1999104721A