Method for controlling shape of multistage rolling mill

Abstract

To set an initial position of an actuator at the time of passing at an appropriate position.SOLUTION: In a rolling load difference calculation step (S37), a rolling load difference ΔP is calculated by using an actually measured last pass steady part load actual measurement value and a next pass tip part load prediction value. In actuator initial position calculation steps (S41 to S45), an actuator initial position Xx is calculated on the basis of a change in shape of a rolled material W caused by the calculated rolling load difference ΔP and a target shape of a tip part at an n-th rolling pass. In an actuator setting step (S52), before the rolling start of the rolled material W at the n-th rolling pass, a position of a shape-control actuator 53 is set at the calculated actuator initial position Xx.SELECTED DRAWING: Figure 6

Description

[Technical field]

[0001] The present invention relates to a shape control method for a multi-stage rolling mill that rolls a material to be rolled. [Background technology]

[0002] For example, Patent Document 1 describes a multi-stage rolling mill. This multi-stage rolling mill performs reverse rolling, in which the rolling direction of the material being rolled is switched for each rolling pass. Claim 1 of the same document describes moving the intermediate rolls to a preset position at the start of rolling of the next pass. Paragraph 0020 of the same document describes that this preset position is determined in advance by a central control device. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 4-344813 Summary of the Invention [Problem to be solved by the invention]

[0004] The document does not describe what the preset position (initial position) of the intermediate roll is. Here, in a multi-stage rolling mill, the rolls (e.g., intermediate rolls) are moved by an actuator. The document does not describe what the initial position of the actuator is at the start of a pass. It is desirable to be able to set the initial position of the actuator at the start of a pass to an appropriate position.

[0005] Therefore, an object of the present invention is to provide a shape control method for a multi-stage rolling mill, which is capable of setting the initial position of an actuator at the start of a pass to an appropriate position. [Means for solving the problem]

[0006] The shape control method of a multi-stage rolling mill includes a previous pass steady section load actual measurement step, a next pass tip section load predicted value calculation step, a rolling load difference calculation step, an actuator initial position calculation step, and an actuator setting step. The multi-stage rolling mill includes a shape control actuator that controls a shape, which is a distribution of flatness of the rolled material in the plate width direction of the rolled material. The multi-stage rolling mill performs reverse rolling in which the rolled material is rolled while changing the rolling direction of the rolled material for each rolling pass. The previous pass steady section load actual measurement step measures a previous pass steady section load actual measurement value, which is an actual measurement value of the rolling load of the steady section of the rolled material in the (n-1)th rolling pass, where n is an integer of 2 or more. The next pass tip section load predicted value calculation step calculates a next pass tip section load predicted value, which is a predicted value of the rolling load of the tip section of the rolled material in the nth rolling pass. The rolling load difference calculation step calculates a rolling load difference using the actual measured load value of the steady portion of the previous pass and the calculated load prediction value of the tip portion of the next pass. The rolling load difference is the difference between the rolling load of the material to be rolled in the (n-1)th rolling pass and the rolling load of the tip portion in the nth rolling pass. The actuator initial position calculation step calculates an actuator initial position based on a change in the shape of the material to be rolled due to the calculated rolling load difference and a target shape of the tip portion in the nth rolling pass. The actuator initial position is the position of the shape control actuator when the tip portion is rolled in the nth rolling pass. The actuator setting step sets the position of the shape control actuator to the calculated actuator initial position before starting rolling of the material to be rolled in the nth rolling pass. Effect of the Invention

[0007] By the above-described method for controlling shape of a multi-stage rolling mill, the initial position of the actuator at the start of a pass can be set to an appropriate position. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram showing a rolling system 1. [Diagram 2] FIG. 2 is a view of the roll group 20 shown in FIG. 1 as viewed in the axial direction of the rolls. [Diagram 3] 3 is a view of the roll group 20 and the like shown in FIG. 2 as viewed from the rolling direction. [Figure 4] 2 is a graph showing the relationship between various states and time regarding rolling in the rolling system 1 shown in FIG. [Diagram 5] 5 is a graph showing the relationship between the rolling load P and time shown in FIG. 4. [Figure 6] 2 is a flowchart of the operation of the rolling system 1 shown in FIG. [Figure 7] 7 is a graph showing the relationship between the actual measured value and the predicted value of the rolling load P in the steady portion calculated in step S32 and step S33 shown in FIG. 6. [Figure 8] 7 is a graph showing the relationship between the actual measured value and the predicted value of the tip load increase rate λ calculated in step S35 shown in FIG. 6. [Figure 9] 2 is a graph showing the relationship between the shape evaluation function and time when rolling is performed by the rolling system 1 shown in FIG. 1 and a comparative example. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] The rolling system 1 will be described with reference to FIGS.

[0010] 1, the rolling system 1 is a system that rolls a material W to be rolled. The rolling system 1 includes a multi-stage rolling mill 10, a control device 60, an input device 71, and a teaching device 73.

[0011] The multi-stage rolling mill 10 is a rolling mill (facility, device) that rolls a material W to be rolled. The material W to be rolled is something (workpiece) that is rolled by the multi-stage rolling mill 10. The material W to be rolled is plate-shaped. The material W to be rolled is, for example, a metal. The multi-stage rolling mill 10 includes a first reel 11, a second reel 12, a pair of direction change units 15, a pair of shape sensors 17, a roll group 20, and an actuator 50.

[0012] The first reel 11 and the second reel 12 are reels on which the material W to be rolled is wound in a coil shape. One of the first reel 11 and the second reel 12 is an unwinding reel that unwinds the material W to be rolled. The other of the first reel 11 and the second reel 12 (the one different from the unwinding reel) is a take-up reel that winds up the material W to be rolled. As the rolling direction of the material W to be rolled changes, the roles of the unwinding reel and the take-up reel are alternated between the first reel 11 and the second reel 12.

[0013] The direction change unit 15 changes the moving direction (path direction) of the material W to be rolled. The direction change unit 15 is a deflector roll equipped with a deflector. In the example shown in FIG. 1, a pair of direction change units 15 are provided. In detail, the direction change units 15 are provided between the first reel 11 and the roll group 20 (in the path of the material W to be rolled) and between the second reel 12 and the roll group 20. The direction change units 15 are cylindrical or columnar members.

[0014] The shape sensor 17 detects the shape of the rolled material W. The "shape" of the rolled material W is the distribution of the flatness of the rolled material W in the width direction of the rolled material W. The shape sensor 17 is a flatness sensor. The shape sensor 17 detects the shape of the rolled material W (at the exit side) rolled by the work roll 21 (see FIG. 2). As described later, since reverse rolling is performed in the multi-stage rolling mill 10, a pair of shape sensors 17 are provided. More specifically, the shape sensors 17 are provided on the first reel side and the second reel side with respect to the work roll 21 (see FIG. 2). The shape sensor 17 detects the distribution of tension in the rolled material W to detect the distribution of the flatness (i.e., the shape) of the rolled material W. More specifically, the shape sensor 17 has a plurality of elements arranged in the width direction of the rolled material W. Each element detects a force in the thickness direction acting from the rolled material W. Thereby, the shape sensor 17 detects the tension distribution of the rolled material W in the width direction of the rolled material W. As a result, the shape sensor 17 detects the distribution of flatness of the rolled material W in the width direction of the rolled material W (i.e., the shape).

[0015] The roll group 20 rolls the material W to be rolled. As shown in FIG. 2, the roll group 20 includes a large number of rolls (work rolls 21, intermediate rolls 30, and backup rolls 40, which will be described later). The number of rolls in the roll group 20 (the number of stages in the multi-stage rolling mill 10) can be set in various ways. The multi-stage rolling mill 10 is a 20-stage rolling mill in the example shown in FIG. 2, but may also be a 6-stage rolling mill, a 12-stage rolling mill, a 14-stage rolling mill, or the like. The roll group 20 includes work rolls 21, intermediate rolls 30, and backup rolls 40.

[0016] Each of the rolls of the roll group 20 (the work roll 21, the intermediate roll 30, and the backup roll 40) is a substantially cylindrical or columnar member. The rolls are rotatably supported in a housing (not shown) that supports the roll group 20. The rolls are rotatable about central axes extending in the longitudinal direction of the rolls.

[0017] In this roll group 20, the material W to be rolled, the work rolls 21, the intermediate rolls 30 (for example, the first intermediate roll 31 and the second intermediate roll 32), and the backup roll 40 contact each other in this order in a cluster shape (like a bunch of grapes). In this contact order, the side opposite the material W to be rolled is referred to as the "roll back surface side". Specifically, the roll back surface side of the work roll 21 is the intermediate roll 30 side, which is opposite the side that contacts the material W to be rolled, based on the work roll 21. The roll back surface side of the first intermediate roll 31 is the second intermediate roll 32 side, which is opposite the side that contacts the work roll 21, based on the first intermediate roll 31. The back surface side of the second intermediate roll 32 is the backup roll 40 side, which is opposite the side that contacts the first intermediate roll 31, based on the second intermediate roll 32.

[0018] The work rolls 21 come into contact with the material W to be rolled and sandwich the material W, thereby rolling the material W. A pair (two) of work rolls 21 are provided on both sides in the thickness direction of the material W to be rolled. The direction in which the two work rolls 21 sandwich the material W to be rolled (i.e., the thickness direction of the material W to be rolled) is, for example, the vertical direction.

[0019] The intermediate rolls 30 are provided between the work rolls 21 and the backup rolls 40. Depending on the number of stages in the multi-stage rolling mill 10, the intermediate rolls 30 may not be provided. For example, when the multi-stage rolling mill 10 has six stages (six-stage rolling mill), the intermediate rolls 30 may not be provided. The intermediate rolls 30 are provided on both sides in the thickness direction of the material W to be rolled. A plurality of intermediate rolls 30 are provided. The intermediate rolls 30 include a first intermediate roll 31 and a second intermediate roll 32.

[0020] The first intermediate roll 31 rotatably supports the work roll 21 from the roll back side. The outer peripheral surface of the first intermediate roll 31 contacts the roll back side portion of the outer peripheral surface of the work roll 21. Each of the first intermediate rolls 31 rotates while contacting the outer peripheral surface of the work roll 21. A plurality of first intermediate rolls 31 are provided. Each of the plurality of first intermediate rolls 31 rotatably supports the work roll 21 from the roll back side. The number of first intermediate rolls 31 can be set in various ways. In the example shown in FIG. 2, since the multi-stage rolling mill 10 is a 20-stage rolling mill, two first intermediate rolls 31 are provided on each side (for example, in a pair of upper and lower rolls) (a total of four) in the thickness direction of the rolled material W. As shown in FIG. 3, the first intermediate roll 31 has a tapered portion 31t.

[0021] The tapered portion 31t is used to control the shape of the rolled material W. The tapered portion 31t is mainly used to control the shape of the outer portion (end portion) of the rolled material W in the sheet width direction. In the tapered portion 31t, the diameter of the first intermediate roll 31 becomes smaller (tapered) as it approaches the outer end of the first intermediate roll 31 in the longitudinal direction. A certain first intermediate roll 31 (for example, the first intermediate roll 31 above the rolled material W) has a tapered portion 31t at an end on one side (for example, the right) in the longitudinal direction of the first intermediate roll 31. Another first intermediate roll 31 (for example, the first intermediate roll 31 below the rolled material W) has a tapered portion 31t at an end on the other side (the opposite side to the above-mentioned "one side", for example, the left) in the longitudinal direction of the first intermediate roll 31. A lateral control actuator 53c described later moves the first intermediate roll 31 in the longitudinal direction of the first intermediate roll 31, and as a result, the tapered portion 31t moves in the longitudinal direction of the first intermediate roll 31. Note that the shape of this first intermediate roll 31 may be a shape different from the shape having the tapered portion 31t only at one end of the first intermediate roll 31 in the longitudinal direction.

[0022] As shown in FIG. 2, the second intermediate roll 32 rotatably supports the first intermediate roll 31 from the roll back side. The outer peripheral surface of the second intermediate roll 32 contacts the roll back side portion of the outer peripheral surface of the first intermediate roll 31. The second intermediate roll 32 rotates while contacting the outer peripheral surface of the first intermediate roll 31. A plurality of second intermediate rolls 32 are provided. Each of the plurality of second intermediate rolls 32 rotatably supports the first intermediate roll 31 from the roll back side. The number of second intermediate rolls 32 can be set in various ways. In the example shown in FIG. 2, since the multi-stage rolling mill 10 is a 20-stage rolling mill, the second intermediate rolls 32 are provided three on each side (for example, in a pair of upper and lower rolls) (total of six) in the thickness direction of the rolled material W. The second intermediate roll 32 may be a driving roll driven by a driving device (not shown). The driving roll may be a roll other than the second intermediate roll 32.

[0023] The backup roll 40 (support roll) rotatably supports the work roll 21 from the roll back side. More specifically, the backup roll 40 supports the work roll 21 from the roll back side via the intermediate roll 30. In the example shown in FIG. 2, the backup roll 40 rotatably supports the intermediate roll 30 (more specifically, the second intermediate roll 32) from the roll back side. The outer circumferential surface of the backup roll 40 contacts the roll back side portion of the outer circumferential surface of the intermediate roll 30 (more specifically, the second intermediate roll 32). The backup roll 40 rotates while contacting the intermediate roll 30 (more specifically, the second intermediate roll 32). A plurality of backup rolls 40 are provided. Each of the plurality of backup rolls 40 supports the work roll 21 from the roll back side via the intermediate roll 30.

[0024] The number of the backup rolls 40 can be set in various ways. In the 20-stage roll group 20 shown in Fig. 2, the backup rolls 40 are provided on both sides (for example, in an upper and lower pair) in the thickness direction of the rolled material W, four on each side (total of eight). As shown in Fig. 3, the backup roll 40 has a structure divided in the longitudinal direction of the backup roll 40 (width direction of the rolled material W). Specifically, each of the multiple backup rolls 40 includes multiple divided rolls 40r.

[0025] The multiple split rolls 40r are rotatably supported on the backup roll shaft 40s. The multiple split rolls 40r are arranged adjacent to each other in the direction in which the backup roll shaft 40s extends. The number of split rolls 40r constituting one backup roll 40 can be set in various ways, and is six in the example shown in FIG. 3. The split rolls 40r are rolls equipped with bearings (bearing rolls).

[0026] The backup roll shaft 40s is a shaft that rotatably supports the divided roll 40r. The backup roll shaft 40s is connected to a housing (not shown) that supports the roll group 20 via a saddle 53b1. Each roll of the roll group 20 will be described below with reference to FIG.

[0027] 1, the actuator 50 is a device that controls the thickness and shape of the rolled material W. The actuator 50 includes a thickness control actuator 51 and a shape control actuator 53.

[0028] The thickness control actuator 51 is a device that controls the thickness of the material W to be rolled. Specifically, the thickness control actuator 51 includes a wedge type reduction actuator 51a. The wedge type reduction actuator 51a reduces (pressures) the outer peripheral surface of the backup roll 40 by a wedge mechanism. As a result, the wedge type reduction actuator 51a reduces the material W to be rolled via the work roll 21.

[0029] The shape control actuator 53 is a device that controls the shape of the rolled material W. The shape control actuator 53 includes a tilt control actuator 53a, a crown control actuator 53b, a lateral control actuator 53c, and a roll eccentricity control actuator 53d.

[0030] The tilt control actuator 53a (tilting device, tilt reduction device) is a device that tilts (integrally) all of the multiple rolls on one side (e.g., above) in the plate thickness direction of the rolled material W. The tilt control actuator 53a changes the tilt (tilt with respect to the horizontal direction) of the housing that supports the rolls.

[0031] The crown control actuator 53b is a device used to correct changes in the shape of the rolled material W caused by thermal crown when the rolling of the rolled material W is in an accelerating state, as shown in Fig. 3. The crown control actuator 53b pushes out the multiple saddles 53b1 individually, thereby pushing out the divided rolls 40r individually.

[0032] The lateral control actuator 53c is a device used for correcting the shape of the end portion in the sheet width direction of the rolled material W. The lateral control actuator 53c moves a roll having a tapered portion 31t (specifically, the first intermediate roll 31) in the sheet width direction. As a result, the lateral control actuator 53c changes the position of the tapered portion 31t in the sheet width direction.

[0033] The roll eccentricity control actuator 53d is a device for changing the amount of deflection (crown amount) of the entire backup roll 40. Here, the crown control actuator 53b is a device for individually pushing out the saddles 53b1 to individually push out the divided rolls 40r. On the other hand, the roll eccentricity control actuator 53d is a device for deflecting one entire backup roll 40. In detail, each of the multiple saddles 53b1 supporting one backup roll 40 has an eccentric cam (eccentric ring). The amount of eccentricity of the eccentric cam differs for each of the multiple saddles 53b1. The roll eccentricity control actuator 53d rotates the backup roll shaft 40s to rotate this eccentric cam. As a result, the roll eccentricity control actuator 53d moves all the saddles 53b1 supporting one backup roll 40 to deflect the entire backup roll 40.

[0034] As shown in FIG. 1, the control device 60 is a computer that performs input and output of signals, calculation (processing), storage of information, etc. For example, the functions of the control device 60 are realized by executing a program stored in a storage unit of the control device 60 in a calculation unit. The control device 60 may be provided inside the multi-stage rolling mill 10, or may be provided outside the multi-stage rolling mill 10 (for example, a control room, etc.). The control device 60 may be a personal computer or may be provided in a control panel. The connection between the control device 60 and an external device of the control device 60 may be a wired connection (connection via a communication cable, etc.) or a wireless connection (for example, a wireless LAN (Local Area Network), Bluetooth (registered trademark), etc.). The connection between each component of the control device 60 may be a wired connection or a wireless connection. The control device 60 performs various controls of rolling. A pass schedule, which is a rolling plan, is set in the control device 60. A rolling condition (described later) for each pass (described later) is set in the control device 60. The control device 60 includes a plate thickness control device 61 , a shape control device 63 , and a setup calculation device 65 .

[0035] The thickness control device 61 controls the thickness of the material W being rolled during rolling. The thickness control device 61 controls the thickness control actuator 51 and controls the position of the thickness control actuator 51. Specifically, the thickness control device 61 controls the wedge type reduction actuator 51a. The thickness control device 61 controls the rolling load P (see FIG. 4), which is the load acting on the material W being rolled during rolling. For example, the thickness control device 61 transfers the rolling load P to the shape control device 63. For example, the thickness control device 61 records (stores, logs) the relationship between the rolling time and the rolling load P.

[0036] The shape control device 63 controls the shape of the rolled material W during rolling. The shape control device 63 controls the shape control actuators 53, and controls the position of the shape control actuators 53. Specifically, the shape control device 63 controls the tilt control actuator 53a, the crown control actuator 53b (see FIG. 3), the lateral control actuator 53c (see FIG. 3), and the roll eccentricity control actuator 53d (see FIG. 3). Hereinafter, the position of the shape control actuator 53 is also referred to as the "actuator position."

[0037] This shape control device 63 may be an automatic shape control device (AFC; Automatic Flatness Control System) that automatically controls the shape of the rolled material W. In this case, the shape control device 63 acquires (imports, reads) the shape of the rolled material W detected by the shape sensor 17, and controls the shape control actuator 53 based on the acquired shape. In detail, the shape control device 63 automatically controls the actuator position (performs automatic shape control) so that the plate shape of the rolled material W becomes flat. At this time, the shape control device 63 performs, for example, feedback control and multivariable control.

[0038] This shape control device 63 records (stores, logs) the relationship between the rolling time and the actuator position, for example. The shape control device 63 may store an influence coefficient A, which will be described later.

[0039] The setup calculation device 65 calculates (calculates, computes) an actuator initial position Xx (described later, see FIG. 4). The setup calculation device 65 exchanges signals with the shape control device 63, for example.

[0040] The input device 71 is a device for an operator to input information. The input device 71 outputs a signal to the control device 60 in response to an operation by the operator. The input device 71 may include a key, a mouse, a switch, or a touch panel. The input device 71 may include a voice input device or an image input device (e.g., a device for reading codes such as one-dimensional codes and two-dimensional codes). The input device 71 may be part of a personal computer, a smartphone, or a tablet.

[0041] The teaching device 73 teaches information to an operator. The teaching device 73 teaches, for example, information indicating an actuator initial position Xx (described later, see FIG. 4). The teaching device 73 may perform teaching by display. In this case, the teaching device 73 may be equipped with a monitor. The teaching device 73 may perform teaching by voice. In this case, the teaching device 73 may be equipped with a speaker.

[0042] (Activation) The rolling system 1 is configured to operate as follows. A shape control method (flatness control method) of the multi-stage rolling mill 10 is performed as follows. Each of the following operations may be regarded as a "step" of the operation.

[0043] (reverse rolling) The multi-stage rolling mill 10 performs reverse rolling. Reverse rolling is rolling the material W while switching the rolling direction of the material W for each rolling pass. In detail, the multi-stage rolling mill 10 alternates between rolling in the first rolling direction and rolling in the second rolling direction described below. In rolling in the first rolling direction, the material W is unwound from the first reel 11, rolled by the roll group 20, and wound up by the second reel 12. In rolling in the second rolling direction, the material W is unwound from the second reel 12, rolled by the roll group 20, and wound up by the first reel 11.

[0044] The above-mentioned "rolling pass" refers to the period from the start to the end of rolling in one rolling direction. Hereinafter, the rolling pass is also simply referred to as a "pass". Specifically, for example, if the first pass is rolling in the first rolling direction, the second pass is rolling in the second rolling direction, and the third pass is rolling in the first rolling direction. When n is an integer of 2 or more, the n-1th pass is referred to as the "previous pass", and the nth pass is referred to as the "next pass". Here, as described later, the control device 60 calculates the actuator initial position Xx (see FIG. 4) for the next pass. The "previous pass" is the pass immediately before the pass in which the actuator initial position Xx is calculated (i.e., the next pass).

[0045] 4 shows graphs that typically represent the relationship between time and the magnitude of the rolling speed, the rolling load P, the actuator position (lateral position and crown position), and the target shape. In the following, the components of the rolling system 1 (such as the control device 60) will be described mainly with reference to FIG. 1, and the rolls of the roll group 20 will be described mainly with reference to FIG. 2.

[0046] (Rolling speed) The details of the graph showing the relationship between the magnitude of the rolling speed and time (change in the magnitude of the rolling speed over time) shown in FIG. 4 are as follows. At the start of the pass, the rolling speed is 0. After the start of the pass, the rolling speed gradually increases from 0. The rolling state at this time is called an accelerated state. The rolled material W rolled in the accelerated state is called the "tip portion". The part of the rolled material W wound on the first reel 11 or the second reel 12 closest to the reel (the tip portion) and its vicinity are discarded as scrap material. A part of the discarded part of the rolled material W does not need to be included in the "tip portion". In detail, the discarded part of the rolled material W includes the unrolled part (the part not rolled by the work roll 21) closest to the reel and the off-gage part inside the unrolled part (the unwinding side). The off-gage part is rolled by the work roll 21, but the thickness of the off-gage part is outside the target thickness specification, so the off-gage part is discarded. This off-gage portion is included in the "tip portion" of the rolled material W in this embodiment. The rolling system 1 of this embodiment may aim to reduce the portion of the "tip portion" of the rolled material W that becomes the off-gage portion (portion to be discarded) as much as possible.

[0047] When the rolling speed reaches a certain speed (steady-state speed), the rolling speed is maintained at this steady-state speed. The rolling state at this time is called the steady-state. The material W being rolled in the steady-state is called the "steady portion." Note that in the graph shown in Figure 4, the rolling speed in the steady state is completely constant, but the rolling speed actually measured is constantly fluctuating.

[0048] The rolling speed is gradually reduced from the steady speed at a predetermined timing. The rolling state at this time is called a decelerated state. The material W being rolled in the decelerated state is called the "tail end." The "predetermined timing" is, for example, when the amount of the material W wound on the reel (first reel 11 or second reel 12) from which the material W is unwound becomes equal to or less than a predetermined amount.

[0049] The rolling speed is decreased and reaches 0, completing one pass.

[0050] The graph showing the change in the rolling speed over time has a similar shape for the previous pass and the next pass. As the number of passes increases (i.e., as the rolled material W becomes thinner), the rolling speed is gradually increased.

[0051] (Rolling load P) Details of the graph showing the relationship between the rolling load P and time (change in the rolling load P over time) are as follows. Figures 4 and 5 show an example of the relationship between the rolling load P and time in a schematic manner. This graph is intended to make it easier to understand the meaning of each value, which will be described later. In reality, the rolling load P always varies depending on the state of the material W being rolled, and the graph shows a complex curve. The graph showing the relationship between the actuator position (lateral position, crown position) and time, shown in Figure 4, also shows an example in a schematic manner. Furthermore, the minimum value of the vertical axis of these graphs is not 0. The vertical axis of these graphs is an enlarged portion where the numerical value indicated on the vertical axis changes.

[0052] The rolling load P is mainly affected by the coefficient of friction at the contact portion (rolling roll bite) between the work roll 21 and the rolled material W, and secondarily by the thickness of the rolled material W. In detail, the faster the rolling speed, the greater the amount of rolling oil drawn into the rolling roll bite (introduced oil film), and the slower the rolling speed, the less the amount of rolling oil drawn into the rolling roll bite. Therefore, the slower the rolling speed, the greater the coefficient of friction between the work roll 21 and the rolled material W. Therefore, the slower the rolling speed, the greater the rolling load P. Therefore, the rolling load P at the front and tail ends of the rolled material W is greater than the rolling load P at the steady portion. Furthermore, the thicker the rolled material W is relative to the gap between the work rolls 21, the greater the rolling load P. As a result of the friction coefficient at the front and rear ends of the material W being rolled being larger than that at the steady portion, the front and rear ends of the material W are usually likely to be thicker than the steady portion. Therefore, the rolling load P at the front and rear ends of the material W being rolled is larger than the rolling load P at the steady portion. In the example shown in FIG. 4, the graph showing the time change of the rolling load P is as follows. At the start of the pass, the rolling load P increases from the load before the start of the pass to a certain value. In the accelerating state, the rolling load P gradually decreases as the rolling speed increases. In the steady state, the rolling load P becomes almost constant. In the decelerating state, the rolling load P gradually increases as the rolling speed decreases.

[0053] The rolling load P at the tip of the rolled material W is the tip load P T The rolling load P of the steady part of the rolled material W is called the steady part load P S It is also called.

[0054] A graph showing the change in the rolling load P over time has a similar shape between the previous pass and the next pass. Usually, the rolling load P is reduced as the number of passes increases (i.e., as the rolled material W becomes thinner).

[0055] (actuator position) The details of the graphs that show an example of the relationship between the actuator positions (lateral position and crown position) and time are as follows. The actuator positions are determined by automatic shape control (or by the operator's operation of the input device 71) so that the rolled material W becomes flat. The lateral position is the actuator position of the lateral control actuator 53c (see FIG. 3). The lateral position shown in the graph of FIG. 4 shows an example of the position of one of the multiple first intermediate rolls 31 in the strip width direction. The crown position is the actuator position of the crown control actuator 53b (see FIG. 3). The crown position shown in the graph of FIG. 4 shows an example of the position of one of the multiple saddles 53b1 (see FIG. 3). As described above, FIG. 4 shows an example of the relationship between the actuator positions (lateral position and crown position) and time. In FIG. 4, the lateral position and the crown position are shown as constant in most of the stationary part of the rolled material W. However, the actual lateral position and crown position are constantly changed in accordance with the output of the shape control device 63 so as to flatten the sheet shape of the rolled material W. The target shape will be described later.

[0056] The rolling load P will be described below mainly with reference to Fig. 5. Each step (S11 to S52) of the flowchart shown in Fig. 6 will be described with reference to Fig. 6.

[0057] (Summary of setup calculation) The control device 60 (more specifically, the setup calculation device 65) calculates (calculates, computes) the actuator initial position Xx shown in Fig. 4. The calculation of the actuator initial position Xx by the control device 60 is also referred to as a setup calculation.

[0058] The actuator initial position Xx is the actuator position (the position of the shape control actuator 53) at the start of the next pass. The actuator initial position Xx is the position of the shape control actuator 53 when the leading end of the rolled material W is rolled in the next pass. The actuator initial position Xx is specifically represented by a vector (described later).

[0059] The outline of the setup calculation is as follows. The control device 60 calculates an actuator initial position Xx that shortens the shape convergence time of the material W in the next pass as much as possible. Here, the shape of the front end of the material W in the next pass changes depending on the amount of change (rolling load difference ΔP (see FIG. 5)) of the rolling load P at the front end of the material W in the next pass with respect to the rolling load P (rolling load P at the steady portion or tail end) of the previous pass. Therefore, the control device 60 predicts (estimates) the rolling load difference ΔP (steps S31 to S37 in FIG. 6). The control device 60 predicts the amount of change in the shape of the front end of the material W in the next pass with respect to the shape of the material W in the previous pass (shape at the steady portion or tail end) using the rolling load difference ΔP (steps S41 to S43 in FIG. 6). The control device 60 calculates the actuator initial position Xx shown in FIG. 4 based on the predicted amount of change in the shape (steps S44 and S45 in FIG. 6). Before the start of the next pass, the shape control actuator 53 sets (presets, sets up) the actuator position to the actuator initial position Xx (step S52 in FIG. 6).

[0060] (Examples of setup calculations, etc.) A specific example of the operation (mainly the setup calculation) of the rolling system 1 is as follows. A specific example of the shape control method of the multi-stage rolling mill 10 is as follows.

[0061] In steps S11 and S12, the control device 60 reads an influence coefficient A, which will be described later. A (Step S11). The control device 60 reads the load influence coefficient A P (Step S12). The influence coefficient A is stored in advance in, for example, the storage device of the shape control device 63. Actuator influence coefficient A A is specifically expressed as a matrix (see below). Load influence coefficient A P Specifically, is expressed as a vector (see below).

[0062] In step S20, the control device 60 starts a setup calculation. The timing of the start of the setup calculation will be described later.

[0063] In steps S31 to S37, the control device 60 calculates the rolling load difference ΔP used to calculate the actuator initial position Xx.

[0064] The outline of the calculation of the rolling load difference ΔP shown in FIG. 5 by the control device 60 is as follows. The control device 60 calculates the rolling load P of the steady part of the rolled material W in the next pass (next pass steady part load P S(n) The control device 60 calculates (predicts) the next pass steady part load P S(n) The rolling load P at the tip of the rolled material W in the next pass (the load P at the tip of the next pass T(n) The control device 60 calculates (predicts) the tip load increase rate λ, which is the increase rate of the steady portion of the rolling load P of the material W to be rolled in the previous pass (step S35). At this time, the control device 60 can accurately predict the tip load increase rate λ by using the actual measured value of the rolling load P of the steady portion of the material W to be rolled in the previous pass and the predicted value of the rolling load P of the steady portion of the material W to be rolled in the previous pass (steps S31 to S34). The control device 60 calculates (predicts) the tip load increase rate λ, which is the increase rate of the steady portion of the rolling load P of the material W to be rolled in the next pass (steps S35). S(n) Using the predicted value (estimated value) of and the tip load increase rate λ, the tip load P T(n) The control device 60 calculates a predicted value (estimated value) of the rolling load P of the previous pass and the load P of the tip of the next pass. T(n) The rolling load difference ΔP, which is the difference between and is calculated (predicted) (step S37). The details of the calculation of the rolling load difference ΔP by the control device 60 are as follows.

[0065] In step S31, the front pass steady part load P S(n-1)The actual value of the rolling load P is measured (previous pass steady section load actual measurement step). Then, the control device 60 reads the actual value of the previous pass steady section load. The actual value of the rolling load P may be measured based on the load acting on the work roll 21, the load acting on the intermediate roll 30, or the load acting on the backup roll 40, for example. The actual value of the rolling load P may be measured based on the load acting on the actuator 50.

[0066] Front pass steady section load P S(n-1) It is preferable that the flatness of the steady portion of the rolled material W in the previous pass, where the flatness is actually measured, is as good as possible. When the flatness of the steady portion of the rolled material W in the previous pass is good, the calculated value of the actuator initial position Xx becomes more appropriate (specifically, the shape convergence is improved).

[0067] Here, the load on the front pass is P S(n-1) is constant in the graph shown in Fig. 5, but in reality it is always fluctuating. Therefore, the load P S(n-1) is actually measured. The "predetermined timing" may be determined based on, for example, the rolling position (longitudinal length) of the material W to be rolled. Specifically, for example, the "predetermined timing" may be the point in time when rolling of a predetermined percentage (e.g., 90%) of the total length in the longitudinal direction of the material W to be rolled is completed. The "predetermined timing" may also be the point in time when a predetermined time has elapsed since the start of the pass. Note that the pre-pass steady section load P S(n-1) Similarly to the actually measured values ​​of the rolling load P, various predicted values ​​of the rolling load P described later are predicted values ​​of the rolling load P at predetermined timings.

[0068] In step S32, the control device 60 calculates the front path steady section load P S(n-1) The control device 60 calculates a calculated value of the steady section load of the previous pass, which is a predicted value (calculated value) of the steady section load of the next pass P S(n) In order to calculate the predicted value of more accurately, the steady load of the front pass P S(n-1) A predicted value of is calculated (see the explanation of step S34).

[0069] Here, the steady-state load P S is the tip load P T It is possible to predict the steady load P more accurately (easily) than the steady load P S can be predicted based on the rolling conditions of the steady portion of the rolled material W. The "rolling conditions" may include the conditions of the rolling process, and may also include the conditions of the material of the rolled material W. For example, the conditions of the rolling process may include one or more of the following: the difference in plate thickness of the rolled material W before and after rolling with the work rolls 21 (between the entry side and the exit side), the diameter of the work rolls 21, the coefficient of friction between the work rolls 21 and the rolled material W, and the tension of the rolled material W. For example, the conditions of the material of the rolled material W may include one or both of the conditions of the yield stress of the rolled material W and the rolling resistance. The steady portion load P S may be predicted using a load prediction formula based on rolling theory. S may be predicted using an empirically derived load prediction formula. S may be predicted using machine learning.

[0070] Steady-state load P S A case will be described in which the rolling conditions and the steady-state load P are predicted (calculated) using machine learning. S Using the relationship, the steady-state load P S In detail, the rolling conditions when the steady part of the rolled material W is actually rolled and the steady part load P S The actual measured value of and the information of are stored (recorded). Based on this information, the rolling conditions and the steady part load P S The control device 60 calculates the steady-state load P from the rolling conditions (input) during rolling based on the previously extracted relationship. S The predicted value (the calculated value of the steady portion load of the front path in step S32) (output) of the load is predicted.

[0071] Figure 7 shows the load P per unit width WI of the stationary part. S , the predicted value (P S / WI predicted value) and actual value (PS The graph below shows the relationship between the actual measured value of / WI and the predicted value (P S The predicted value of P / WI is a value predicted based on the relationship extracted in advance by machine learning and the rolling conditions. In the example shown in FIG. S The coefficient of determination (R 2 ) was 0.987. This resulted in the steady-state load P S It is shown that it is possible to predict with high accuracy.

[0072] In step S33, the control device 60 calculates the next pass steady part load P S(n) The next pass steady portion load calculation value, which is a predicted value of the load calculated for the steady portion of the rolled material W in the next pass, is calculated (predicted). As with the previous pass steady portion load calculation value (see step S32), the next pass steady portion load calculation value can be calculated (predicted) with high accuracy based on the rolling conditions of the steady portion of the rolled material W in the next pass.

[0073] In step S34, the control device 60 calculates the next pass steady part load P S(n) The next pass steady-state load predicted value is calculated as a next pass steady-state load predicted value (next pass steady-state load predicted value calculation step). This next pass steady-state load predicted value is more accurate than the next pass steady-state load calculated value (step S33) predicted based only on the rolling conditions of the steady-state part of the rolled material W in the next pass. S(n) is the predicted value.

[0074] The reason why it is preferable to calculate the steady-state load prediction value of the next pass is as follows. For example, the steady-state load P S The rolling conditions used to predict include the coefficient of friction between the work roll 21 and the rolled material W, and the deformation resistance of the rolled material W. Here, these rolling conditions are not changed for each coil (rolled material W wound on the first reel 11 and the second reel 12) (chance) (based on constant rolling conditions), and the steady-state load P S However, it is assumed that these rolling conditions vary from coil to coil. Then, the steady-state load P SThere may be a discrepancy between the actual measured value and the predicted value for each coil. Therefore, it is preferable that the control device 60 calculates a next pass steady portion load predicted value taking this discrepancy into account.

[0075] Specifically, the control device 60 calculates a next pass steady section load predicted value using the ratio (deviation ratio, correction coefficient) between the previous pass steady section actual value (step S31) and the previous pass steady section calculated value (step S32). The control device 60 calculates a next pass steady section load predicted value based on this correction coefficient and the next pass steady section load calculated value (step S33). Specifically, for example, the control device 60 calculates the next pass steady section load predicted value using the following formula (1).

[0076]

number

[0077] As a modified example, the control device 60 may use the next pass steady portion load calculated value (step S33) that has not been corrected using the correction coefficient as the next pass steady portion load predicted value.

[0078] In step S35, the control device 60 calculates (predicts) the tip load increase rate λ (tip load increase rate calculation step). The control device 60 calculates the tip load increase rate λ based on the rolling conditions of the next pass. The tip load increase rate λ is calculated by multiplying the next pass steady section load P S(n) Next pass tip load P T(n) is the rate of increase.

[0079] Specifically, for example, the tip load increase rate λ is expressed by the following formula.

[0080] λ=(P T(n) -P S(n) )÷P S(n)

[0081] This formula is an example of the tip load increase rate λ (the same applies to other formulas). For example, the tip load increase rate λ is expressed as λ=P T(n) ÷P S(n), etc. In the following, the tip load increase rate λ is λ = (P T(n) -P S(n) )÷P S(n) , an example represented by the formula:

[0082] (Why is the tip load increase rate λ used?) The reason why the tip load increase rate λ is calculated (predicted) is as follows. T(n) It is difficult to directly predict the tip end load increase rate λ. On the other hand, it is possible to predict the tip end load increase rate λ from the rolling conditions of the next pass. Therefore, the tip end load increase rate λ is calculated.

[0083] The reason why the tip load increase rate λ is calculated (predicted) is as follows. As described above, the tip load P T(n) It is difficult to directly predict the tip load P T This is because it is difficult to construct a theoretical model to predict the tip load P T There are some parts that are somewhat predictable and some parts that are difficult to predict.

[0084] The rolling speed at the tip of the rolled material W is low (specifically, it is lower than the rolling speed at the steady portion) (see FIG. 4). Therefore, the rolling oil is not sufficiently drawn to the portion (roll bite) where the work roll 21 and the rolled material W come into contact with each other. As a result, the coefficient of friction between the work roll 21 and the rolled material W increases, and the rolling load P (tip load P T The change in the rolling load P due to this change in the friction coefficient can be predicted to some extent.

[0085] On the other hand, the following rolling load P is difficult to predict. Some materials for the rolled material W have strain rate dependency of deformation resistance. In materials with strain rate dependency, when the rolling speed is slow, the strain rate is smaller than when the rolling speed is fast, and as a result, the deformation resistance is lower, and the rolling load P is lower. The effect of this deformation resistance on the rolling load P is difficult to predict.

[0086] In addition, the tip load P T varies depending on the entry thickness at the front end of the material W to be rolled (the thickness before it enters the work rolls 21). This "entrance thickness at the front end of the material W to be rolled" is the exit thickness at the tail end of the material W to be rolled in the previous pass (the thickness exiting the work rolls 21). Here, the rolling speed at the tail end of the material W to be rolled is low (0 or nearly 0) (see FIG. 4), and as a result, the accuracy of the exit thickness at the tail end of the material W to be rolled is low (i.e., the actual thickness deviates significantly from the target thickness). Therefore, the exit thickness at the tail end of the material W to be rolled is difficult or impossible to predict. In other words, the entry thickness at the front end of the material W to be rolled in the next pass is difficult or impossible to predict. As a result, the front end load P T is difficult to predict.

[0087] The change in the friction coefficient, the deformation resistance of the material with strain rate dependency, and the precision of the plate thickness at the tip are the factors that affect the tip load P T The effect of the load on the tip P changes depending on the acceleration pattern of the rolling. T It is difficult to predict (constructing a theoretical model for prediction is difficult). Therefore, the load at the tip of the next pass P T(n) It is difficult to directly predict from the rolling conditions of the tip of the workpiece W in the next pass.

[0088] On the other hand, the next pass steady load P S(n) As described above, the load P of the steady portion of the rolled material W in the next pass can be predicted with high accuracy from the rolling conditions of the steady portion of the rolled material W in the next pass (see steps S33 and S34). S(n) Next pass tip load P T(n) The rate of increase of the load at the steady portion of the next pass P S(n) Using the predicted value (see steps S33 and S34) and the tip load increase rate λ, the next pass tip load P T(n) A predicted value of the load at the end of the next pass (predicted value of the load at the end of the next pass) is calculated (step S36 described later).

[0089] (Next pass steady part load P S(n) (Reason for predicting the tip load increase rate λ separately) The control device 60 determines the next pass steady part load P S(n) The reason for separately calculating (predicting) the load increase rate λ of the tip portion (see steps S34 and S35) is as follows. S(n) As described above, it is possible to predict with high accuracy. For example, the steady-state load P S(n) is stable against changes in the operating state. On the other hand, the tip load increase rate λ is predictable, but the accuracy of the prediction of the tip load increase rate λ depends on the next pass steady load P S(n) For example, the tip load increase rate λ is easily changed by the change in the operating state (it depends on the operating state). Therefore, the control device 60 predicts the next pass steady part load P S(n) and the tip load increase rate λ are calculated (predicted) separately. T(n) Compared to the case where the load at the end of the next pass P T(n) The prediction accuracy of the above will improve (the overall prediction accuracy will improve).

[0090] (Specific example of calculation of tip load increase rate λ) A case where the tip load increase rate λ is predicted (calculated) using machine learning will be described. The control device 60 calculates the tip load increase rate λ using the relationship between the rolling conditions and the tip load increase rate λ, which is extracted in advance by machine learning. In detail, the tip load increase rate λ is calculated based on the rolling conditions and the tip load P T The measured value and the steady load P S The actual measured values ​​of and information on are stored (recorded). Based on this information, the relationship between the rolling conditions and the tip load increase rate λ is extracted in advance. Then, the control device 60 predicts the tip load increase rate λ (output) from the rolling conditions (input) during rolling based on the relationship extracted in advance. A specific example of this "rolling condition (input) during rolling" is the same as the rolling condition in the calculation of the previous pass steady section load predicted value (step S32). The rolling conditions used when predicting the tip load increase rate λ may include information on the pass schedule set in advance (before rolling) in the control device 60.

[0091] FIG. 8 shows a graph showing the relationship between the predicted value of the tip load increase rate λ and the actual measured value. The predicted value of the tip load increase rate λ is a value predicted based on the relationship extracted in advance by machine learning and the rolling conditions. In the example shown in FIG. 8, the coefficient of determination (R 2 ) was 0.664. This shows that the prediction accuracy of the tip load increase rate λ is limited compared to the rolling load P in the steady state shown in Figure 7. However, it shows that it is possible to predict the tip load increase rate λ with a certain degree of accuracy.

[0092] In step S36, the control device 60 calculates the next pass end load P T(n) (next pass tip load predicted value) is calculated (predicted) (next pass tip load predicted value calculation step). The control device 60 calculates the next pass tip load predicted value based on the next pass steady section load predicted value (step S34) and the tip load increase rate λ (step S35). Specifically, the control device 60 calculates the next pass tip load predicted value by the following equation (2).

[0093]

number

[0094] In addition, in this formula 2, the explanation of the symbols explained in formula 1 (specifically, the next pass steady part load prediction value) is omitted. The explanation of the symbols explained in the previous formula is omitted in the following formulas as well.

[0095] As a modified example, the control device 60 may calculate the next pass tip load prediction value without using the tip load increase rate λ. For example, the previous pass steady section load P S(n-1) Actual measured value and next pass tip load P T(n) A prediction formula showing the relationship between the load P and the steady portion of the front path is empirically (experimentally) derived. Then, the control device 60 uses the empirically derived prediction formula and the steady portion of the front path load P S(n-1) A next pass end load prediction value may be calculated based on the actual measured value of and.

[0096] In step S37, the control device 60 calculates the rolling load difference ΔP (rolling load difference calculation step). The rolling load difference ΔP is a value used to calculate the actuator initial position Xx (see FIG. 4). The rolling load difference ΔP is the product of the rolling load P of the previous pass and the load P of the front end of the next pass. T(n) The rolling load difference ΔP is the difference between the rolling load P of the previous pass and the load P at the tip of the next pass. T(n) The control device 60 controls the load P in the front pass steady section to the front pass steady section. S(n-1) The actual measured value (actual measured value of the steady part load of the previous pass, step S31) and the load P T(n) The rolling load difference ΔP is calculated by using the predicted value of the next pass tip load P (next pass tip load predicted value, step S34). T(n) The predicted value of and the steady load of the front pass P S(n-1) It is not necessary that the rolling load difference ΔP is the difference between the actual measured value of and Specifically, for example, the control device 60 calculates the rolling load difference ΔP as follows.

[0097] (Calculation of ΔP1 and ΔP2) The control device 60 calculates ΔP1 and ΔP2 used in calculating the rolling load difference ΔP.

[0098] ΔP1 is the load P of the next pass S(n) and the load on the front pass steady section P S(n-1) Specifically, ΔP1 is expressed by the following formula:

[0099] ΔP1=P S(n) -P S(n-1)

[0100] Next pass steady load P used in the calculation of ΔP1 S(n) is preferably the next pass steady section load predicted value (step S34), or may be the next pass steady section load calculated value (step S33). S(n-1) is preferably the actual measured value of the load at the steady portion of the front path (step S31), but may be the calculated value of the load at the steady portion of the front path (step S32).

[0101] ΔP2 is the load at the tip of the next pass P T(n) and the load on the front pass steady part P S(n-1) Specifically, ΔP2 is expressed by the following formula:

[0102] ΔP2=P T(n) -P S(n-1)

[0103] Next pass tip load P used in the calculation of ΔP2 T(n) is the predicted load value of the tip of the next pass (step S36). The steady load P of the previous pass used in the calculation of ΔP2 S(n-1) is preferably the actual measured value of the load at the steady portion of the front path (step S31), but may be the calculated value of the load at the steady portion of the front path (step S32).

[0104] (Calculation of ΔP using adjustment parameter c) It is assumed that there is a rolling load difference ΔP between these ΔP1 and ΔP2. The reason is as follows. As described above, ΔP2 is the difference between the load P T(n) and the load on the front pass steady part P S(n-1) The difference between the load at the tip of the next pass P T(n) can be predicted with a certain degree of accuracy, but the steady-state load P S Therefore, if ΔP2 is used as is to determine the actuator initial position Xx (see Fig. 4), the shape convergence of the rolled material W (described later) may be limited (the control result may be poor). On the other hand, ΔP1 is a function of the load P S(n) and the load on the front pass steady part P S(n-1) It is the difference between ΔP1 and ΔP2, and can be predicted with high accuracy (higher accuracy than ΔP2). Therefore, it is assumed that there is a rolling load difference ΔP between ΔP1 and ΔP2. Specifically, for example, the rolling system 1 is expressed as the following formula (3) using the adjustment parameter c.

[0105] ΔP=c·ΔP1+(1-c)·ΔP2――――(3)

[0106] The adjustment parameter c is a parameter that determines how close (or consistent) the rolling load difference ΔP is to ΔP1 and ΔP2. The adjustment parameter c is, for example, 0 or more and, for example, 1 or less. When the adjustment parameter c is 0, the rolling load difference ΔP is ΔP2. When the adjustment parameter c is 1, the rolling load difference ΔP is ΔP1.

[0107] The adjustment parameter c is adjusted so as to improve the control result of the shape control of the rolled material W. For example, the adjustment parameter c is adjusted to a value that improves the shape convergence (described later) (converges quickly). For example, the adjustment parameter c is adjusted to a value that improves the shape convergence (described later) (converges quickly). T The value of the adjustment parameter c may be adjusted to a value that minimizes the difference between the actual measured value and the predicted value. The value of the adjustment parameter c is adjusted (changed) within a predetermined range (for example, 0 to 1) from a value before adjustment (initial value) (specifically, 0.5, etc.) so as to improve the control result. The adjustment of the adjustment parameter c may be performed by an operator operating the input device 71 (manual operation). The adjustment of the adjustment parameter c may also be performed automatically by the control device 60 based on the control result of the shape control of the rolled material W.

[0108] In steps S41 to S45, the control device 60 calculates the actuator initial position Xx shown in Fig. 4 (actuator initial position calculation step). The actuator initial position Xx is the actuator position at the start of the next pass (the position of the shape control actuator 53 when rolling the front end portion in the nth rolling pass).

[0109] The outline of the calculation of the actuator initial position Xx is as follows. The control device 60 calculates the actuator initial position Xx based on the change in the shape of the material W due to the rolling load difference ΔP (step S37) and the next pass end target shape ft2. In detail, the control device 60 calculates the actuator position change amount ΔX based on the difference between the previous pass target shape ft1 and the next pass end target shape ft2 and the change in the shape of the material W due to the rolling load difference ΔP. SOL (Steps S41 to S44) The actuator position change amount ΔX SOLis the amount of displacement of the actuator position from the actuator position (Bx or Fx) in the previous pass to the actuator initial position Xx at the start of the next pass. SOL The actuator initial position Xx is calculated from the previous pass target shape ft1, the next pass end portion target shape ft2, and the actuator position change amount ΔX SOL Each of the actuator position Bx at the steady portion of the previous path and the actuator position Fx at the tail end of the previous path is represented by a vector, for example (described later). Note that in FIG. 6 and the following formulas (Formulas 3 to 10), symbols representing vectors are underlined. The details of the calculation of the actuator initial position Xx are as follows.

[0110] In step S41, the control device 60 reads the previous pass target shape ft1 and the next pass end target shape ft2 shown in FIG.

[0111] The previous pass target shape ft1 is the target shape of the material W to be rolled in the previous pass. The previous pass target shape ft1 may be set in advance (before step S41) in the control device 60. The actual shape of the stationary part of the material W to be rolled in the previous pass may coincide with the previous pass target shape ft1 due to automatic shape control by the shape control device 63 or the operation of the operator. In this case, the previous pass target shape ft1 may be the shape of the stationary part of the material W to be rolled in the previous pass detected (actually measured) by the shape sensor 17. Specifically, the previous pass target shape ft1 is a vector of values ​​(e.g., tension, etc.) indicating the target shape for each of a plurality of sensor positions in the sheet width direction of the material W to be rolled. The above "sensor position" is the position of each element of the shape sensor 17 in the sheet width direction (the same applies to the "sensor position" below).

[0112] The next pass tip target shape ft2 is the target shape of the tip of the material W in the next pass. The next pass tip target shape ft2 is a vector of values ​​indicating the target shape for each of multiple sensor positions in the plate width direction of the material W. The next pass tip target shape ft2 may be set in advance in the control device 60 (before step S41). The next pass tip target shape ft2 may be calculated (computed) based on the rolling conditions, etc.

[0113] The next pass tip target shape ft2 is set (determined, selected) so as to ensure the sheet running stability. Specifically, for example, the next pass tip target shape ft2 is set so that the difference between the gap (roll gap) of the work rolls 21 and the thickness of the tip of the rolled material W is within a predetermined range. If the rolled material W is too thick for the gap of the work rolls 21, the rolled material W does not easily pass through the gap of the work rolls 21. Also, if the rolled material W is too thin for the gap of the work rolls 21, the rolled material W moves (violates) in the sheet thickness direction in the gap of the work rolls 21. The next pass tip target shape ft2 is set so as to suppress these problems. Also, it is preferable that the next pass tip target shape ft2 is set so as to shorten the shape convergence time. It is preferable that the next pass tip target shape ft2 is set so as to ensure the sheet running stability and to shorten the shape convergence time.

[0114] In step S42, the controller 60 reads the actuator position on the previous pass.

[0115] (Two examples of how to set it up) The actuator position at the start of the next pass (i.e., actuator initial position Xx) is set up (set) based on the actuator position in the previous pass. This actuator position reference can be, for example, the following setup example 1 and setup example 2. Note that in FIG. 6, setup example 1 and setup example 2 are simply described as example 1 and example 2.

[0116] [Setup Example 1] The above reference may be the actuator position at the end of the previous pass (at the tail end of the material W to be rolled in the previous pass). In this case, the control device 60 calculates the actuator initial position Xx (the actuator position at the start of the next pass) based on the actuator position at the end of the previous pass.

[0117] [Setup Example 2] The above reference may be the actuator position in the steady state of the previous pass (at the steady portion of the rolled material W in the previous pass). In this case, the control device 60 calculates the actuator initial position Xx based on the actuator position in the steady state of the previous pass.

[0118] In the case of setup example 1, the control device 60 reads the previous pass steady section actuator position Bx and the previous pass tail end actuator position Fx shown in FIG. 4. The previous pass steady section actuator position Bx is the actuator position at the steady section of the material W to be rolled in the previous pass. Specifically, the previous pass steady section actuator position Bx is a vector of values ​​indicating the actuator position for each of multiple sensor positions in the strip width direction of the material W to be rolled. The previous pass tail end actuator position Fx is the actuator position at the tail end of the material W to be rolled in the previous pass (at the end of the previous pass). The previous pass tail end actuator position Fx is a vector of values ​​indicating the actuator position for each of multiple sensor positions in the strip width direction of the material W to be rolled.

[0119] In the case of setup example 2, the controller 60 reads the front pass tail end actuator position Fx. In the case of setup example 2, the controller 60 does not need to read the front pass steady portion actuator position Bx.

[0120] In step S43, the control device 60 calculates the shape change V1 or the shape change V2. Each of the shape change V1 and the shape change V2 is represented by, for example, a vector (described later).

[0121] (Shape change V1 in setup example 1) In the case of the above setup example 1, the control device 60 calculates the shape change V1. The shape change V1 is the amount of change in the shape of the rolled material W from the tail end (end) of the rolled material W in the previous pass to the front end (start) of the rolled material W in the next pass. The shape change V1 is the amount of change in the shape of the rolled material W taking into account the change in actuator position, the rolling load difference ΔP, and the influence coefficient A. Specifically, the shape change V1 is a vector of values ​​indicating the change in shape for each of multiple sensor positions in the plate width direction of the rolled material W.

[0122] For example, as shown in equation (5) (see equation 4) described later, the shape change V1 is expressed by the actuator position change amount ΔX SOL1 The actuator position change amount ΔX is expressed as SOL1 is the amount of change in the actuator position that should be changed from the end of the previous pass to the start of the next pass (the amount of actuator increase that should be set up). Actuator position change amount ΔX SOL1 is the amount of change (increase) in the actuator position from the actuator position at the end of the previous pass to the actuator initial position Xx.

[0123] For example, as shown in equation (4) (see equation 3) below, the actuator position change amount ΔX SOL1 can be expressed using the target shape difference (ft2-ft1), the actuator position difference (Bx-Fx), the rolling load difference ΔP, and the influence coefficient A. Here, the target shape difference (ft2-ft1) is the difference between the previous pass target shape ft1 and the next pass leading edge target shape ft2. The actuator position difference (Bx-Fx) is the difference between the previous pass steady section actuator position Bx and the previous pass tail end actuator position Fx. The actuator position difference (Bx-Fx) is the amount by which the shape control actuator 53 is moved by automatic shape control or by operator operation when a load change occurs as the rolled material W being rolled in the previous pass moves from the steady section to the tail end.

[0124] (Influence coefficient A) The shape change V1 is (actuator position change ΔX SOL1is expressed using the influence coefficient A (see equations (4) and (5) described later). As a result, the actuator initial position Xx is calculated taking into account the influence coefficient A. The influence coefficient A includes the actuator influence coefficient A A (See step S11) and the load influence coefficient A P (See step S12).

[0125] Actuator influence coefficient A A is a coefficient indicating the influence of the position of the shape control actuator 53 on the shape of the rolled material W. Actuator influence coefficient A A is a coefficient indicating the amount of change in the shape of the rolled material W when the shape control actuator 53 moves by a unit amount. Specifically, the actuator influence coefficient A A is a matrix (arranged vertically and horizontally) showing coefficients for each of a plurality of sensor positions in the sheet width direction of the rolled material W for each of a plurality of types of shape control actuators 53.

[0126] Load influence coefficient A P is a coefficient that indicates the effect of the rolling load P of the rolled material W on the shape of the rolled material W. Load effect coefficient A P is a coefficient that indicates the amount of change in the shape of the rolled material W when the rolling load P changes by a unit amount. Specifically, the load effect coefficient A P is a vector indicating coefficients for each of multiple sensor positions in the plate width direction of the rolled material W.

[0127] Specifically, for example, the actuator position change amount ΔX SOL1 is expressed as the following equation (4).

[0128]

number

[0129] The above equation (4) can be expressed as the following equations (5) and (6).

[0130]

number

[0131] The above equation (5) expresses the actuator position change amount ΔX SOL1 , the actuator influence coefficient A A This is an equation that represents the shape change V1 from the end of the previous pass to the start of the next pass.

[0132] In the automatic shape control, an equation of the same form as the above equation (5) may be used. In this case, the actuator initial position Xx can be calculated by a calculation method similar to that used in the automatic shape control. In addition, it is preferable that the influence coefficient A used in the automatic shape control can be used as the influence coefficient A used for the actuator initial position Xx.

[0133] (Shape change V2 in setup example 2) In the case of the above setup example 2, the control device 60 calculates the shape change V2. The shape change V2 is the amount of change in the shape of the rolled material W from the steady part (steady state) of the rolled material W in the previous pass to the leading end (start time) of the rolled material W in the next pass. The shape change V2 is the amount of change in the shape of the rolled material W taking into account the change in actuator position, the rolling load difference ΔP, and the influence coefficient A. Specifically, the shape change V2 is a vector of values ​​indicating the change in shape for each of multiple sensor positions in the plate width direction of the rolled material W.

[0134] For example, as shown in equation (8) (see equation 6) described later, the shape change V2 is expressed by the actuator position change amount ΔX SOL2 The actuator position change amount ΔX is expressed as SOL2 is the amount of change in the actuator position (the amount of actuator increase to be set up) that should be changed from the steady part of the rolled material W in the previous pass to the tip part (start time) of the rolled material W in the next pass. Actuator position change amount ΔX SOL2 is the amount of change (increase) in the actuator position from the actuator position of the steady part of the rolled material W in the previous pass to the actuator initial position Xx.

[0135] For example, as shown in equation (7) below, the actuator position change amount ΔX SOL2 can be expressed using the target shape difference (ft2-ft1), the rolling load difference ΔP, and the influence coefficient A. Here, in setup example 1, the actuator position change amount ΔX SOL1 was expressed using the actuator position difference (Bx-Fx) (see equation (4) above). On the other hand, in setup example 2, the actuator position change amount ΔX SOL2 In this case, the difference in actuator position (Bx-Fx) is not required.

[0136] Specifically, for example, the actuator position change amount ΔX SOL2 is expressed as the following equation (7).

[0137]

number

[0138] The above equation (7) can be expressed as the following equations (8) and (9).

[0139]

number

[0140] The formula (5) of the setup example 1 and the formula (8) of the setup example 2 are isomorphic. Therefore, these formulas are expressed ignoring the subscripts "1" and "2". Specifically, the shape change V from the previous pass to the tip of the next pass is expressed as the actuator position change ΔX from the previous pass to the tip of the next pass. SOL and the actuator influence coefficient A A Using these, it can be expressed as the following equation (10).

[0141]

number

[0142] In step S44, the control device 60 calculates the actuator position change amount ΔX SOL The control device 60 calculates A A ΔX SOL The actuator position change amount ΔX at which the magnitude of -V is minimum SOL (The least squares method is applied to the above formula (10)). Specifically, the control device 60 calculates the following formula (11).

[0143]

number

[0144] In addition, A A T A A is the transpose of (A A T A A ) -1 is (A A T A A ) is an inverse matrix of. In addition, in the above formula (11), there is no constraint coefficient K (described later) in step S44 shown in FIG.

[0145] (Calculation in case where there is a shape control actuator 53 not used for shape control) The rolling system 1 may include multiple types of shape control actuators 53 (for example, the crown control actuator 53b and the lateral control actuator 53c shown in FIG. 3). In this case, only some types of shape control actuators 53 among the multiple types of shape control actuators 53 included in the rolling system 1 may be selected and used for shape control (called actuator selection mode). When the actuator selection mode is performed, the control device 60 performs calculations, for example, as follows.

[0146] Actuator influence coefficient A when the actuator selection mode is performed and when it is not performed A may be changed.

[0147] Actuator influence coefficient A when the actuator selection mode is performed and when it is not performed A may be used without changing the constraint coefficient K. The constraint coefficient K is a matrix having values ​​only in the diagonal components. In the matrix of the constraint coefficient K, the components corresponding to the types of shape control actuators 53 not used for shape control are set to values ​​that are sufficiently large compared to the components corresponding to the types of shape control actuators 53 used for shape control. As a result, the amount of movement of the types of shape control actuators 53 not used for shape control (the effect on the shape of the rolled material W) is set to approximately 0. Specifically, for example, the components of the types of shape control actuators 53 used for shape control are set to 1, and the components of the types of shape control actuators 53 not used for shape control are set to the huge number 10 10 And so on.

[0148] When this constraint coefficient K is used, the control device 60 calculates the actuator position change amount ΔX by the following equation (12) instead of the above equation (11): SOL Calculate the following.

[0149]

number

[0150] (Calculation in the case where there is a shape control actuator 53 that has a small effect on shape control) Among the multiple types of shape control actuators 53 included in the rolling system 1, the influence of some types of shape control actuators 53 on shape control (how they affect the control result) may be made smaller than the influence of other types of shape control actuators 53 on shape control. This is called an influence reduction mode. Below, the influence reduction mode will be described, mainly focusing on the differences from the actuator selection mode.

[0151] Actuator influence coefficient A with and without the influence reduction mode A may be changed.

[0152] Actuator influence coefficient A with and without the influence reduction mode A may be left unchanged and the constraint coefficient K may be used (similar to the actuator selection mode). When the influence reduction mode is performed, in the matrix of the constraint coefficient K, the components corresponding to the type of shape control actuators 53 whose influence is reduced are set to sufficiently large values ​​compared to the components corresponding to the type of shape control actuators 53 whose influence is not reduced. By adjusting the magnitude of the components of the type of shape control actuators 53 whose influence is reduced, it is possible to adjust the influence that this type of shape control actuator 53 has on the shape of the rolled material W. Specifically, for example, the components of the type of shape control actuators 53 whose influence is not reduced are set to 1, and the components of the type of shape control actuators 53 whose influence is reduced are set to 10. 2 By setting the magnitude of the component of the type of shape control actuator 53 whose influence is to be reduced to this level, this type of shape control actuator 53 can have some influence on shape control (a smaller influence than other types of shape control actuators 53).

[0153] In step S45, the control device 60 calculates the actuator initial position Xx.

[0154] In the case of the above setup example 1, the actuator initial position Xx is the actuator position change amount ΔX SOL1 The actuator initial position Xx is a position obtained by changing the actuator position by only Xx. In this case, the actuator initial position Xx is expressed by the following equation (13) (see equation 10). The actuator initial position Xx is a vector of values ​​indicating the actuator position for each of a plurality of sensor positions in the plate width direction of the rolled material W.

[0155] In the case of the above setup example 2, the actuator initial position Xx is the actuator position change amount ΔX SOL2In this case, the actuator initial position Xx is expressed by the following equation (14).

[0156]

number

[0157] In step S51, the control device 60 outputs the actuator initial position Xx.

[0158] In step S52, before the start of rolling the material W in the next pass, the position of the shape control actuator 53 is set (set up) to the actuator initial position Xx (actuator setting step). This setting may be manually set by an operator or automatically set by the control device 60.

[0159] A specific example of outputting the actuator initial position Xx (step S51) and setting the shape control actuator 53 (step S52) is as follows.

[0160] (Example where the actuator initial position Xx is manually set) The control device 60 outputs the calculated actuator initial position Xx to the teaching device 73 (see FIG. 1) (step S51). The teaching device 73 teaches the operator information indicating the actuator initial position Xx (performs operator guidance). For example, the teaching device 73 displays information indicating the actuator initial position Xx. The operator operates the input device 71 (see FIG. 1) based on the information taught by the teaching device 73. At this time, the operator operates the input device 71 so that the actuator position becomes the actuator initial position Xx. As a result, the actuator position is set to the actuator initial position Xx (step S52).

[0161] (An example in which the actuator initial position Xx is automatically set) The control device 60 may automatically control the actuator position based on the calculated actuator initial position Xx (step S51). As a result, the actuator position may be set to the actuator initial position Xx (step S52).

[0162] (Type of shape control actuator 53 set at actuator initial position Xx) The types of shape control actuators 53 include a tilt control actuator 53a (see FIG. 1), a crown control actuator 53b (see FIG. 3), a lateral control actuator 53c (see FIG. 3), and a roll eccentricity control actuator 53d (see FIG. 3). The type of shape control actuator 53 set at the actuator initial position Xx calculated by the control device 60 may be one type or multiple types. For example, the shape control actuator 53 set at the actuator initial position Xx may be either one or both of the lateral control actuator 53c (see FIG. 3) and the crown control actuator 53b (see FIG. 3). The shape control actuator 53 set at the actuator initial position Xx preferably includes the crown control actuator 53b (see FIG. 3) which is easy to move even when rolling is stopped.

[0163] (Start timing of setup calculation) The timing at which calculation of the actuator initial position Xx starts (step S20) (called the start timing of the setup calculation) is at least before the start of the next pass.

[0164] In the above setup example 1, the control device 60 calculates the actuator initial position Xx based on the actuator position Fx at the tail end of the previous pass shown in Fig. 4. Therefore, in setup example 1, the start timing of the setup calculation needs to be after the end of the previous pass.

[0165] In the above setup example 2, the control device 60 calculates the actuator initial position Xx based on the previous pass steady section actuator position Bx. Therefore, the control device 60 can calculate the actuator initial position Xx without using information at the end of the previous pass. Therefore, the control device 60 can start the setup calculation before the end of the previous pass. As a result, in the setup example 2, the actuator initial position Xx can be calculated earlier than in the setup example 1. As a result, the time from the end of the previous pass to the start of the next pass can be shortened, and the productivity of the rolled material W in the rolling system 1 can be improved. Since the actuator initial position Xx is calculated early, when the operator manually sets the actuator position, the operator's operation is likely to be completed earlier.

[0166] Specifically, in setup example 2, the start timing of the setup calculation must be after the end of rolling at a reference position of the steady part of the rolled material W in the previous pass. For example, the start timing of the setup calculation may be during deceleration of the rolled material W before the end of the previous pass. In this case, the control device 60 may start the setup calculation when it is determined that the rolling speed is decelerating. Note that the start timing of the setup calculation in setup example 2 may be during rolling of the steady part (steady state) as long as it is after the end of rolling at a reference position of the steady part of the rolled material W in the previous pass. Also, although the effect of being able to advance the start timing of the setup calculation is not obtained, the setup calculation in setup example 2 may be performed after the end of the previous pass.

[0167] (comparison) As shown in FIG. 9, the shape convergence time of the material W to be rolled was compared between an example (example E) of this embodiment and a comparative example. Example E is an example in which the actuator position at the start of the next pass is set to the actuator initial position Xx calculated by the control device 60. The comparative example is an example in which the actuator position at the start of the next pass is set by the operator's sense (by the operator's know-how) rather than the actuator initial position Xx calculated by the control device 60. In addition, in example E and the comparative example, the shape control actuator 53 that controls the shape of the material W to be rolled is the lateral control actuator 53c (see FIG. 3) and the crown control actuator 53b (see FIG. 3). The horizontal axis of the graph shown in FIG. 9 is the time from the start of rolling of the pass (the next pass in example E). The vertical axis of the graph is the shape evaluation function. The shape evaluation function is the deviation between the target shape of the material W to be rolled and the measured shape.

[0168] As shown in Fig. 9, in the comparative example, the shape evaluation function is disturbed (fluctuates greatly) after rolling begins. On the other hand, in example E, the shape evaluation function decreases monotonically immediately after rolling begins and stabilizes in less than half the time of the comparative example (see the arrow in Fig. 9). That is, it can be seen that in example E, the shape of the rolled material W reaches a stable shape (converged shape) in less than half the time of the comparative example. It can be seen that the shape convergence is improved in example E compared to the comparative example.

[0169] (Consider) The shape control device 63 may control the shape of the rolled material W by automatic shape control. In this case, the shape convergence varies depending on the shape (initial shape) of the rolled material W at the start of a pass. This initial shape is determined by the initial position of the shape control actuator 53 before the start of the next pass. Conventionally, this initial position was determined by the know-how of the operator. Therefore, the shape convergence was dependent on the skill of the operator. Therefore, the shape convergence varied depending on the skill of the operator. In addition, the shape convergence varied for each coil, and the shape convergence varied for each pass. As a result of the variation in the shape convergence, there was a variation in the shape accuracy of the tip of the rolled material W. In addition, as a result of the variation in the shape convergence, there was a variation in the acceleration pattern (speed increase timing) of rolling. Therefore, the variation in the shape convergence hindered the product quality of the rolled material W and the productivity of the rolled material W in the multi-stage rolling mill 10.

[0170] On the other hand, in this embodiment, the control device 60 calculates the actuator initial position Xx as described above. Then, the actuator position at the start of the next pass is set to the calculated actuator initial position Xx. As a result, it is possible to suppress the variation in shape convergence due to the skill of the operator, and the variation in shape convergence for each coil and pass.

[0171] (Effects of the first invention) The effects of the shape control method of the multi-stage rolling mill 10 shown in Fig. 1 are as follows. The multi-stage rolling mill 10 is equipped with a shape control actuator 53. The shape control actuator 53 controls the shape, which is the distribution of flatness of the rolled material W in the plate width direction of the rolled material W. The multi-stage rolling mill 10 performs reverse rolling in which the rolling direction of the rolled material W is switched for each rolling pass. The shape control method of the multi-stage rolling mill 10 includes a previous pass steady section load actual measurement step, a next pass tip section load predicted value calculation step, a rolling load difference calculation step, an actuator initial position calculation step, and an actuator setting step.

[0172] [Configuration 1] In the previous pass steady section load measurement step (see step S31), the previous pass steady section load measurement value is measured. The previous pass steady section load measurement value is the actual rolling load P of the steady section of the rolled material W in the (n-1)th rolling pass (previous pass steady section load P S(n-1) The next pass tip load predicted value calculation step (see step S36) calculates the next pass tip load predicted value. The next pass tip load predicted value is a predicted value of the rolling load P at the tip of the rolled material W in the nth rolling pass (next pass tip load P T(n) (see FIG. 5)). The rolling load difference calculation step (see step S37) calculates the rolling load difference ΔP (see FIG. 5) using the actual measured load value of the steady portion of the previous pass (see step S31) and the calculated predicted load value of the tip of the next pass (see step S36). The rolling load difference ΔP is calculated as the product of the rolling load P of the rolled material W in the (n-1)th rolling pass (rolling load P of the previous pass) and the rolling load P of the tip of the nth rolling pass (tip load P of the next pass T(n) ) and . The actuator initial position calculation step (see steps S41 to S45) calculates the actuator initial position Xx based on the change in shape of the rolled material W due to the calculated rolling load difference ΔP and the target shape of the front end in the nth rolling pass (next pass front end target shape ft2). The actuator initial position Xx (see FIG. 4) is the position of the shape control actuator 53 when rolling the front end in the nth rolling pass. The actuator setting step (see step S52) sets the position of the shape control actuator 53 to the calculated actuator initial position Xx before starting rolling of the rolled material W in the nth rolling pass (next pass).

[0173] By the above [Configuration 1], the position of the shape control actuator 53 at the start of the next pass can be set to the actuator initial position Xx determined by calculation, not to the initial position of the actuator position that depends only on the operator's feeling. Therefore, the initial position of the shape control actuator 53 at the start of a pass (next pass) can be set to an appropriate position.

[0174] Here, the shape (initial shape) of the rolled material W at the start of the next pass changes depending on the initial position of the shape control actuator 53 at the start of the next pass, and the shape convergence after the start of the next pass changes. In the above [Configuration 1], the position of the shape control actuator 53 at the start of the next pass can be set to the actuator initial position Xx determined by calculation. Therefore, it is possible to suppress the variation in the shape convergence after the start of the next pass, compared to when the initial position is set depending only on the operator's sense.

[0175] (Effects of the second invention) [Configuration 2] The shape control method for the multi-stage rolling mill 10 includes a next pass steady section load predicted value calculation step and a tip section load increase rate calculation step. The next pass steady section load predicted value calculation step (see step S34) calculates a next pass steady section load predicted value. The next pass steady section load predicted value is a predicted value of the rolling load P of the steady section in the nth rolling pass (next pass steady section load P S(n) The tip load increase rate λ (see FIG. 5) is calculated based on the rolling conditions in the nth rolling pass in the tip load increase rate calculation step (see step S35). The tip load increase rate λ is calculated based on the rolling conditions in the nth rolling pass in the S(n) ) for the rolling load P at the tip in the nth rolling pass (next pass tip load P T(n) The next pass tip load predicted value calculation step (see step S36) calculates the next pass tip load predicted value based on the calculated next pass steady section load predicted value (see step S34) and the calculated tip load increase rate λ (see step S35).

[0176] The above [Configuration 2] has the following effects. The rolling load P (steady-state load P S ) can be predicted with high accuracy. On the other hand, the rolling load P at the tip of the rolled material W (tip load P T ) is the steady-state load P S The accuracy of prediction of the tip load increase rate λ is S The accuracy of the prediction is lower than that of the tip load PT Therefore, in the above [Configuration 2], a next pass steady section load predicted value that can be predicted with high accuracy (see step S34) and a predictable tip load increase rate λ (see step S35) are calculated. Then, based on the next pass steady section load predicted value and the tip load increase rate λ, a next pass tip load predicted value is calculated (see step S36). Therefore, for example, the next pass tip load P T(n) Compared to the case where the load at the tip of the next pass P T(n) As a result, the load at the end of the next pass P T(n) It is possible to make the actuator initial position Xx (see FIG. 4) calculated using the above more appropriate value. Specifically, it is possible to further improve the shape convergence of the material W to be rolled after the start of the next pass.

[0177] (Effects of the third invention) [Configuration 3] Rolling conditions and tip load P T The measured value and the steady load P S The relationship between the rolling conditions and the tip portion load increase rate λ is extracted in advance from the actual measured values ​​of and (see the explanation of step S35). In the tip portion load increase rate calculation step (see step S35), the tip portion load increase rate λ is calculated from the rolling conditions in the nth rolling pass based on the relationship extracted in advance.

[0178] By the above [Configuration 3], the tip end load increase rate λ in the next pass can be appropriately predicted based on the rolling conditions of the next pass. As a result, the actuator initial position Xx calculated using the tip end load increase rate λ can be made a more appropriate value.

[0179] (Effect of the fourth invention) [Configuration 4] The actuator initial position calculation step (see steps S41 to S45) calculates the actuator influence coefficient A A and load influence coefficient A P Based on this, the actuator initial position Xx is calculated (see step S43). Ais a coefficient indicating the effect that the position of the shape control actuator 53 has on the shape of the rolled material W. P is a coefficient indicating the effect that the rolling load P of the rolled material W has on the shape of the rolled material W.

[0180] By the above [Configuration 4], it is possible to calculate the actuator initial position Xx taking into consideration the influence of the position of the shape control actuator 53 on the shape of the rolled material W and the influence of the rolling load P of the rolled material W on the shape of the rolled material W. Therefore, the actuator influence coefficient A A and load influence coefficient A P In comparison with the case where the actuator initial position Xx is calculated without using at least one of the above, the actuator initial position Xx can be set to a more appropriate value.

[0181] (Effect of the fifth aspect of the invention) [Configuration 5] The multi-stage rolling mill 10 performs automatic shape control to automatically control the shape of the material W to be rolled. Actuator influence coefficient A A and load influence coefficient A P One or both of the coefficients are used for automatic shape control.

[0182] According to the above [Configuration 5], the actuator influence coefficient A used in the actuator initial position calculation step (see steps S41 to S45) A and load influence coefficient A P The coefficients used in the automatic shape control can be used as one or both of the coefficients in the above. Therefore, it is not necessary to set the influence coefficient A for calculating the actuator initial position Xx separately from the coefficients used in the automatic shape control.

[0183] (Effect of the sixth aspect of the invention) [Configuration 6] In the actuator initial position calculation step (see steps S41 to S45), an actuator initial position Xx is calculated while the material W to be rolled is decelerating before the end of the (n-1)th rolling pass.

[0184] According to the above [Configuration 6], the actuator initial position Xx can be calculated earlier than when the calculation of the actuator initial position Xx is started after the previous pass is completed. Therefore, the next pass can be started earlier. Therefore, the overall work time of the reverse rolling performed by the multi-stage rolling mill 10 can be shortened, and the productivity of the material W to be rolled by the multi-stage rolling mill 10 can be improved.

[0185] (Effects of the seventh aspect of the invention) [Configuration 7] The shape control actuator 53 shown in FIG. 3 is either one or both of a lateral control actuator 53c and a crown control actuator 53b.

[0186] The above [Configuration 7] provides the following effects. Each of the lateral control actuator 53c and the crown control actuator 53b can reliably control the shape of the rolled material W. These shape control actuators 53 are set to the actuator initial position Xx at the start of the next pass (step S52). Therefore, the shape (initial shape) of the rolled material W at the start of the next pass is appropriately controlled. As a result, the shape convergence of the rolled material W after the start of the next pass can be improved.

[0187] (Modification) The above embodiment may be modified in various ways. For example, the modified examples in the above embodiment may be combined in various ways. For example, the number of components (including modified examples) in the above embodiment may be changed, or some of the components may not be provided. For example, the connection of each component of the control device 60 shown in FIG. 1 may be changed. For example, the arrangement of the components may be changed. For example, the inclusion relationship of the components may be changed in various ways. For example, a component described as a lower component included in a higher-level component may not be included in this higher-level component, but may be included in another component. For example, a component described as a plurality of different members or parts may be treated as a single member or part. For example, a component described as a single member or part may be provided as a plurality of different members or parts. For example, the order of the steps in the flowchart shown in FIG. 6 may be changed, or some of the steps may not be performed. For example, various parameters (such as adjustment parameter c) used for controlling the control device 60 may not be the same as the various parameters, and may be parameters that can be converted into various parameters. For example, the control device 60 may perform substantially the same processing (calculation, judgment, etc.) as the processing (calculation, judgment, etc.) in the above embodiment (including modified examples). The various processes may be combined in various ways. For example, the calculation by the control device 60 may be performed in various ways. The formulas used in the calculation and the calculation procedures may be modified in various ways. The values ​​used in the calculation and the values ​​calculated by the calculation may be changed in various ways. Specifically, for example, instead of performing a calculation using a certain value, the calculation may be performed using a value that can be converted to this value. Also, instead of computing (calculating) a certain value, a value that can be converted to this value may be computed. For example, each component may have only a part of each feature (action function, arrangement, shape, operation, etc.). [Explanation of symbols]

[0188] 10 Multi-stage rolling mill 53 Shape Control Actuator 53b Crown Control Actuator 53c Lateral Control Actuator A A Actuator Influence Coefficient A P Load Influence Factor P Rolling load P S Static load (Static rolling load P) P S(n) Next pass steady part load (rolling load P of the steady part of the rolled material W in the nth rolling pass) P S(n-1) Previous pass steady part load (rolling load P of the steady part of the rolled material W in the n-1th rolling pass) P T Tip load (tip rolling load P) P T(n) Next pass tip load (rolling load P at the tip of the rolled material W in the nth rolling pass) W Rolled material Xx Actuator initial position ΔP Rolling load difference λ Tip load increase rate

Claims

1. A shape control method for a multi-stage rolling mill, comprising the steps of: The multi-stage rolling mill includes a shape control actuator that controls a shape, which is a distribution of flatness of the rolled material in the plate width direction of the rolled material, and performs reverse rolling in which the rolled material is rolled while changing the rolling direction of the rolled material for each rolling pass. a previous pass steady portion load measurement step of measuring a previous pass steady portion load value, which is an actual measurement value of the rolling load of the steady portion of the rolled material in the (n-1)th rolling pass, where n is an integer of 2 or more; a next pass tip load predicted value calculation step of calculating a next pass tip load predicted value which is a predicted value of the rolling load at the tip of the rolled material in the n-th rolling pass; a rolling load difference calculation step of calculating a rolling load difference, which is a difference between the rolling load of the rolled material in the (n-1)th rolling pass and the rolling load of the tip portion in the (n)th rolling pass, using the actual measured value of the load at the steady portion of the previous pass and the calculated predicted value of the load at the tip portion of the next pass; an actuator initial position calculation step of calculating an actuator initial position, which is a position of the shape control actuator when the front end portion is rolled in the nth rolling pass, based on the change in shape of the rolled material caused by the calculated rolling load difference and a target shape of the front end portion in the nth rolling pass; an actuator setting step of setting the position of the shape control actuator to the calculated actuator initial position before starting rolling of the rolled material in the n-th rolling pass; A shape control method for a multi-stage rolling mill comprising the steps of:

2. A shape control method for a multi-stage rolling mill according to claim 1, a next pass steady portion load predicted value calculation step of calculating a next pass steady portion load predicted value which is a predicted value of the rolling load of the steady portion in the n-th rolling pass; a tip load increase rate calculation step of calculating a tip load increase rate, which is an increase rate of the rolling load of the tip portion in the nth rolling pass relative to the rolling load of the steady portion in the nth rolling pass, based on the rolling conditions in the nth rolling pass; Equipped with the next pass end load predicted value calculation step calculates the next pass end load predicted value based on the calculated next pass steady portion load predicted value and the calculated end load increase rate; A shape control method for a multi-stage rolling mill.

3. A shape control method for a multi-stage rolling mill according to claim 2, A relationship between the rolling conditions and the tip load increase rate is extracted in advance from the rolling conditions, the actual measured value of the rolling load of the tip portion, and the actual measured value of the rolling load of the steady portion, The tip portion load increase rate calculation step calculates the tip portion load increase rate from the rolling conditions in an n-th rolling pass based on the relationship extracted in advance. A shape control method for a multi-stage rolling mill.

4. A shape control method for a multi-stage rolling mill according to claim 1, The actuator initial position calculation step calculates the actuator initial position based on an actuator influence coefficient indicating an influence of a position of the shape control actuator on a shape of the rolled material, and a load influence coefficient indicating an influence of a rolling load of the rolled material on the shape of the rolled material. A shape control method for a multi-stage rolling mill.

5. A shape control method for a multi-stage rolling mill according to claim 4, The multi-stage rolling mill performs automatic shape control for automatically controlling the shape of the rolled material, One or both of the actuator influence coefficient and the load influence coefficient are coefficients used in the automatic shape control. A shape control method for a multi-stage rolling mill.

6. A shape control method for a multi-stage rolling mill according to claim 1, The actuator initial position calculation step calculates the actuator initial position during deceleration of the rolled material before the end of the (n-1)th rolling pass. A shape control method for a multi-stage rolling mill.

7. A shape control method for a multi-stage rolling mill according to any one of claims 1 to 6, The shape control actuator is one or both of a lateral control actuator and a crown control actuator. A shape control method for a multi-stage rolling mill.

Images