JPH01306008A - Method and device for controlling shape of sheet stock - Google Patents

Abstract

PURPOSE:To improve the control accuracy by finding respective differences between separately set shape standards at plural positions on a sheet stock and actual measured values as shape deviations at respective positions and performing shape control considering an influence and a weight coefficients at each position. CONSTITUTION:A flatness meter 5 measuring a flatness at plural positions selected in the lateral direction of a rolled stock 4 is installed and an actuator 7 consisting of a leveling 71, WR bender 72, WR shift 74, and roll coolant 76, etc. is also installed. In that time, the coolant 76 is independently controlled at the same number of positions being equal to the number of measuring positions by the meter 5 in the lateral direction. A flatness controller 6 performs a control for the actuator 7 considering an influence and a weight coefficients at respective measuring positions based on a rolling schedule from a rolling schedule setter 8. The control accuracy is improved because the control repeatedly performs intelligent control calculations in a short time.

Description

Detailed Description of the Invention [Object of the Invention] (Industrial Application Field) The present invention includes leveling, bending, shifting, and coolant as actuators for controlling the shape of a plate material, and these actuators are used to control the rolling mill output. The present invention relates to a method and apparatus for controlling the shape of a plate material for controlling the shape of the plate material on the side.

(Prior art) Control of the flatness of a plate in rolling etc. is aimed at keeping the flatness of the plate at the exit side of the rolling machine (the flatness of the plate in the width direction and the rolling direction) to a desired value. , has been a long-standing issue.

Regarding the flatness control of plates, for example, see "Theory and Practice of Plate Rolling" edited by Japan Iron and Steel Institute (Japanese Edition) (September 10, 1980).
Japan Iron and Steel Institute) pages 308 to 312 (
12.5 Flatness Control). As described in this document, in hot rolling, the pending force of the rolls has conventionally been controlled using equations (1) and (2).

ΔQ −K φΔφ+211ΣΔφ ...(1
)7 However, ΔQ7: Bending force correction amount of final stand Δφ−φ−φ* 1. 2: Proportionality, integral constant φ: Measured flatness value φ*: Target flatness ΔQ-'''' to 1 bi ΔP1+ to 2 pi ΔP1-1+k ・Δ
"(1-1) ...(2) 3 However, ΔQ1: Bending force correction amount ΔP: Rolling load variation amount ΔQ: Bending force variation amount (i-1) k11'k21"31' Rolling dimensions and roll crown Coefficient i determined by: stand number For cold rolling, a flatness control system for a six-high rolling mill is shown in Figure 12.28 on pages 310-311 of the above-mentioned document, and the output of the flatness detector is 3) It is approximated by the fourth-order polynomial in Eq.

φ−λIX+λ2X2+λ3x3+λ4X4・・・・・・
・(3) However, the widthwise ends are X-±11, and the widthwise center is X・-〇.

Based on this equation (3), the symmetric component is determined by the hill-climbing search method and the pending force and intermediate roll position are manipulated, and the asymmetric component is determined by using the manipulated variable to level the rolling position and the flatness is controlled in the same way. was.

(Problems to be Solved by the Invention) In the hot rolling in the above-mentioned conventional technology, flatness control is insufficient because only the pending force is controlled.
In cold rolling, leveling, bending, intermediate roll position, etc. are controlled, but coolant control is not sufficient. Furthermore, the hill-climbing search method takes a long calculation time and is not suitable for, for example, DDC (direct digital control) with a control period of 50 m5.

Furthermore, in the past, there was no system in place to incorporate the experience and knowledge of engineers and operators into flatness control.

[7] Therefore, it was insufficient as a technique for controlling the crown and flatness of the exit side plate material during rolling etc. to desired values.

The present invention has been made in consideration of the above circumstances, and an object of the present invention is to provide a method and apparatus for controlling the shape of an exit plate material to a desired shape in rolling of metal or the like.

[Structure of the invention]

(Means for Solving the Problems) In order to achieve the above object, the present invention provides a shape deviation obtained as a difference between a shape standard set by dividing the shape of a plate material into a plurality of positions and a shape measurement value corresponding thereto. and the square of the difference between the product of the weighting coefficient at the relevant position and the sum of the products of the operation amount, influence coefficient, and weighting coefficient at the relevant position of leveling, bender, and shift, and the sum of the products for each position is the minimum value. Determine the amount of operation for leveling, bender, and shift as follows, and calculate the value obtained by integrating the product of the amount of coolant operation, influence coefficient, and weighting coefficient for each position, and the shape deviation and weighting coefficient for each position. The coolant operation amount for each position is determined from the product minus the sum of the products of the leveling, bender, and shift operations, the influence coefficient, and the weighting coefficient.

(Function) According to the present invention, the influence coefficient and weighting coefficient are considered for each position as a shape deviation, which is the difference between the base value and the actual value regarding the shape divided into a plurality of positions, and the operation amount of each actuator is determined. By determining the shape from a comprehensive viewpoint and performing shape control at high speed, it is possible to obtain a plate material with a good shape.

In addition, conventional knowledge and experience regarding shape control can be converted into rules and stored as a knowledge base, and by using that knowledge base to make inferences and perform intelligent shape control,
- Good layer shape control can be achieved.

(Example) Examples of the present invention will be described below with reference to the drawings.

There are various types of rolling mills, such as a two-high rolling mill, a four-high rolling mill, a six-high rolling mill, and a twenty-high rolling mill. Here, a case where flatness control is performed as a specific example of shape control will be described using a six-high rolling mill as an example.

FIG. 1 shows an embodiment of the present invention, in which a rolled material 4 is controlled to a predetermined flatness by a six-high rolling mill having each pair of work rolls 1, intermediate rolls 2, and backup rolls 3. shall be taken as a thing. A flatness meter 5 is provided on the exit side of the rolling mill to measure the flatness of the rolled material by dividing it into a plurality of pieces (M pieces) in the width direction, and the detection result, that is, the actual rolling value, is equal to the flatness standard described later. Flatness control is performed by the flatness control device 6 via the actuator 7 so that the flatness is controlled.

It is assumed that the material to be rolled 4 is sent in the direction of arrow A.

The flatness standard is determined based on the rolling schedule given from the rolling schedule setting section 8.

The actuator tough has leveling 71, work roll (
WR) vendor 72, intermediate roll (IMR) vendor 7
3. Work roll (WR) shift 74, intermediate roll (I
MR) shift 75, and roll coolant 76. The roll coolant 76 can be divided into a plurality of parts (the same number as the number of widthwise flatness detections by the flatness meter 5) in the width direction of the rolled material 4 and can be controlled independently. The rolling schedule set by the rolling schedule setting section 8 takes into consideration, for example, material type, plate thickness, plate width, feed rate, post-processing, and the like.

Based on each input data, the flatness control device 6 calculates, for example, 5
The intelligent control calculation is repeated every 0m5, and the actuator 7
Intelligent control of flatness via.

FIG. 2 is an explanatory diagram of the actuator 7, and as shown in the figure, the actuator 7 includes a leveling 71, a work roll (WR) bender 72, an intermediate roll (IMR) bender 73, a work roll (WR) shift 74, an intermediate roll (
I M R') shift 75, and roll coolant 7
6 (not shown in FIG. 2). Leveling 71 controls the roll gaps on the work side and drive side, respectively. WR vendor 72 is work role 1
Apply bending force to. IMR Pengu 7
3 gives a bending force to 1MR2. WR
A shift 74 shifts the work roll 1 in the width direction of the sheet.

The IMR shift 75 shifts the intermediate roll 2 in the sheet width direction. The roll coolant 76 cools the roll by dividing the roll cooling water into M parts in the board width direction.

Leveling 71, WR vendor 72, IMR vendor 7
3. There are two types of WR shift 74 and IWR shift 75: electric type and hydraulic type, but hydraulic type is usually used.

As shown in FIG. 2, the WR vendor 72, IMR vendor 73, WR shift 74, and IMR shift 75 are
They are provided for each of the upper and lower rolls, respectively.

FIG. 3 is a more detailed block diagram of the embodiment shown in FIG. 1, and will be described in detail below.

Regarding the intelligent control method and apparatus of the present invention, the flatness control itself will be explained first, and then the intelligent control will be explained.

According to the setting result of the rolling schedule setting section 8 (1;,
A flatness standard is determined by the usage standard setting section 60, and a flatness standard is determined based on the setting result of the rolling schedule setting section 8 and the rolling actual value detected by the rolling actual value detecting section 6.
A base correction value is determined by the correction calculation unit 66, and the two are added together by an adder 61 to form a corrected flatness standard, and the difference between this and the flatness determined by the flatness meter 5 by 7 fpJ, that is, the flatness deviation, is added. The flatness control unit 63 operates to set this to zero. Note that the flatness control section 63 stores the influence coefficient and PID control constants are also introduced.

The corrected flatness standard REI output from the adder 61 is y (i-1 to N), and the flatness ME from the flatness meter 5 is
If the AS flatness predetermined value is y, (i-1~N), then the flatness deviation Δy, (i-1
~N) is REP MEAS ...... (1) ΔY'
-Yl'ji.

This flatness deviation Δy is input to the flatness control section 63. Here, the flatness is controlled using the basic equation (2).

coefficients, η1' Kxi, kZi'kui'
kVi'kwl' kcj are weighting coefficients. Also, i-1, 2,...N, and j-1, 2
,...N. Furthermore, ΔX leveling amount Δz: WR pending force Δu: IMR pending force Δv: WR shift 2 Δw: IMR shift amount ΔC8 2 roll coolant 2.

Now, from equations (1) and (2), the flatness control unit 63 calculates ΔX, Δ2°ΔU, which minimizes J.
, ΔV, and ΔW are determined. The solution to equation (3) can be found using a well-known solution method. The solution to equation (4) is ΔX.

Δ2. Let ΔU, ΔV, and ΔW be. This becomes the operation amount of each actuator.

Inserting the solution of equation (4) into equation (2), we obtain (5). Since the value on the right side of equation (5) has been determined, let this be Δa.

Equation (5) is for i-1, 2,...N...
(6) becomes. From equation (6), Δc1. Δc1. ．． ..., Δc
The total coolant, N, can be found.

The actuator operation amount ΔX is determined by the above equations (1) to (6).

Δ2. ΔU, Δ■, ΔW, and ΔCt were found.

The flatness control unit 63 supplies the manipulated variable obtained as described above to the selector 64 through a PID control operation. The selector 64 guides only necessary operation amount signals to the adder 65 according to the configuration of the actuator 7 . The adder 65 performs manual intervention as necessary for each manipulated variable by adding the manual intervention value from the manual intervention section 6B to the initial setting value in the initial setting section 6A using the adder 6C, and calculates the result. is applied to the actuator 7 to control the flatness.

Since the calculations of equations (1) to (6) above can be performed at high speed, it is possible to realize DDC every 50 m5, for example.

The above is the first feature of the present invention, which is capable of high-speed DDC.
This is flatness control in which weights are applied to iml~N.

Next, the second feature of the present invention, intelligent flatness control, will be described.

The rolling schedule setting section 8 in FIG. 3 sets the steel type, plate thickness, plate width, customer, application, roll gap, tension, roll speed, etc. Further, the flatness standard setting section 60 intelligently determines the flatness standard according to the rolling schedule, and similarly, the standard correction calculation section 66 intelligently determines the flatness standard according to the rolling schedule and rolling performance values. The influence coefficient and control constant calculation unit 67 calculates the reference correction value, and further calculates the standard correction value according to the rolling schedule and rolling actual value.
The influence coefficient and PID control constant are determined intelligently.

The flatness control section 63 performs intelligent control of flatness, and the mode selection section 68 intelligently selects the mode of the actuator 7, that is, selects the individual actuators to be operated, according to the rolling schedule and actual rolling values.

Intelligent flatness control will be described in detail below.

First, the rolling schedule set by the rolling schedule setting section 8 is given to the flatness standard setting section 60. The flatness standard setting unit 60 determines the steel type, plate thickness, and (i-1 to N). In this case, the wave-rolled material 4 during rolling usually has a temperature difference in the width direction. Therefore, even if the desired level of flatness is achieved during rolling, when the plate cools down to room temperature after rolling, the flatness of the plate will differ from the desired level. In addition, the temperature difference in the width direction of the wave rolled material 4 during rolling is not determined only by the steel type, thickness, and width, but also by the speed,
It also changes depending on coolant, lubricant, roll temperature distribution, etc. Engineers and operators involved in rolling operations know from experience the relationship between the temperature difference in the width direction of the strip during rolling and the flatness when the temperature reaches room temperature. This relationship can be expressed, for example, by flatness. For example, if there is a temperature difference of 1°C in the board width direction, 0.5I-UNIT (1-UNIT
is 0. This results in a tension difference of 0.07 kg/NA). Therefore, with respect to a given flatness standard in the board width direction, this is corrected by the temperature difference. This becomes the coefficient. Also,
Users may request that the flatness be changed after actually using the product.

In addition, engineers and operators involved in rolling operations have knowledge about rolling machines and the habits of wave-rolled materials, and use this knowledge to determine flatness standards.

As mentioned above, the knowledge and experience of engineers and operators...
In order to determine N), it is best to express the knowledge and experience of engineers and operators as rules, and then decide N) on a case-by-case basis based on steel type, plate thickness, plate width, rolling conditions, etc.
This is the second feature of the present invention.

This rule expression, as shown in the program example below,
Regarding the rule of flatness reference material temperature difference, if the steel type is electrical steel plate, the plate thickness is 0.29 to 0.31 mm, and the plate width is
1100 to 1300mm, the customer is Company T, the application is motor core, REF is individual, and the material temperature is KA''C, then call subroutine 5ub-1.

As described above, numerical calculations are performed in subroutine 5ub-1 to correct the flatness standard.

FIG. 4 is a functional block diagram that realizes the above intelligent control. 100 engineers and operators (hereinafter simply referred to as operators) have created a knowledge base 1 by converting their knowledge and experience into rules.
Enter 01. A knowledge compiler 102 converts the contents of this knowledge base 101 into a knowledge internal representation 103. This knowledge internal representation 103 is for performing inference at high speed within a computer, and is substantially the same as the knowledge base 101. The inference mechanism 1 receives a request from the input/output unit 105.
．． 04 performs inference using the knowledge internal representation 103 and passes the inference result to the input/output unit 105.

The configuration shown in FIG. 4 includes the flatness standard setting section 60, flatness control section 63, reference correction calculation section 66, influence coefficient and control constant calculation section 67, and mode selection section 68 shown in FIG. 3, respectively.

Below is an example of a program that accomplishes this.

The programming language used here is (1)
In (2) and (3), it is LISP, and in (2) and (3), FORT
It is RAN.

This example is an example of the flatness 1 setting section 60, but the flatness control section 63, reference correction calculation section 66, influence coefficient and control constant calculation section 67, or mode selection section 6, which will be described later,
Each function of 8 can be implemented using exactly the same method.

Program example (1) Description in TDES3; Rule module flatness (ru1cmodulc flatness C flatness standard 1) (flatness standard correction 2) (flatness influence coefficient 3) (flatness control constant 4) (mode selection 5 ) (Flatness control ga6) (Flatness reference material temperature difference 7) ; Rule T Salinity reference material temperature difference (rule A Flatness reference material temperature difference; Conditions for starting the YKA calculation procedure (schcma Flatness CFBK (Steel type Electromagnetic steel plate) (plate thickness ?Th1ckncss (<-0,29&&
>-0JI)) (board width ?width l<-
H[l(1&&>-130(it) (Customer T company) (Application Motor core) (REF Individual) (Material temperature ?Tetap =-KA))-> (call 5ubJ): Calculate Y1KA1 and convert it to array YKA Set child; The following omitted; The following omitted; Procedure declaration; ゛(procedure sub -1, procedure to calculate YKA: entry"5U31": o-rilc"foo, o": Lype 5ubroutine: languagec f77) (2 ) FORTRAN program 5UBROUTI called by the TDES3 expert system (ES)
NE SOB -l C0MM0N /TDESDATA/X (50),
Y (50), KA (50), YKA (5
0) Do too l -1,50 YKA(1) -Y(1) *KA(1)100 C
0NTINUE ND (3) ES call PI in the entire program? OCR
AM overall program 1? EAL X (50), Y (50), KA (50),
YKA (50) COMMON /TDH3DATA/X
, Y, KA, YKACX, Y, KA (
Omitted) Call the ES of CTDES3 to calculate YKA CALL TDES3 Perform the next calculation using the C calculated value YKA (Omitted) TOP ND The above is a program example.

Next, the standard correction calculation section 66 uses the rolling schedule set by the rolling schedule setting section 8 and the rolling actual value detection section 6.
The correction value for the flatness standard set by the flatness standard setting unit 60 is calculated by inputting the actual rolling value detected in step 9, and calculates the correction value for the flatness standard set by the flatness standard setting unit 60, and calculates the correction value for the flatness standard set by the flatness standard setting unit 60, and calculates the correction value for the flatness standard set by the flatness standard setting unit 60, ,Depending on the application, the edge part of the plate is placed under low tension during threading, the flatness standard is changed according to the speed during acceleration/deceleration, and the flatness standard is left as is (correction value zero) during rolling speed.
and so on. This stabilizes the operation and provides plates with good flatness.

Continuously rolled materials have welding points on the plate, and in this case as well, it is necessary to change the flatness standard. In addition, there is a coiler on the exit side of the rolling mill that winds up the rolled plate, but if the thickness of the plate in the width direction changes, the tension distribution in the width direction of the plate changes on the exit side of the rolling mill, so the coiler is used to wind up the rolled plate.

Change the flatness standard according to the coil diameter. In this case, as in the case of the flatness standard setting unit 60, a correction value for the flatness standard is determined using the knowledge base and the inference mechanism, and the correction values are added by the adder 61.

In the influence coefficient and control constant calculating section 67, influence coefficients and PID control constants are calculated. The influence coefficient is a coefficient expressed as a fraction on the right-hand side of equation (2), and is calculated based on the well-known rolling theory. Bender 73, WR Shift 74, IMR
Separately determine the influence of shift 75 and roll coolant 76,
It shall be summarized in the form of a table. This will be called an influence coefficient table. This influence coefficient table is stored in the influence coefficient and control constant calculation unit 67 according to the steel type, plate thickness, plate width, etc.

This influence coefficient table is calculated by the flatness standard setting unit 6 according to the steel type, plate thickness, plate width, roll diameter, lubricating oil, tension between stands, etc.
A knowledge base is created in the same manner as in the case of 0, inference is made using an inference mechanism, and the result is manually input to the flatness control unit 63.

The influence coefficient and control constant calculation unit 67 stores PID control constants for each actuator according to the steel type, plate thickness, plate width, etc., and provides the PID control constants to the flatness control unit 63.

The PrD control constant is based on the time constant and gain of each actuator of the actuator tough, the time constant and gain of the flatness meter 5, the influence coefficient stored in the calculation unit 67, the flatness control period of the flatness control unit 63, etc. Optimal control can be achieved by changing accordingly. Therefore, according to these values, rules for determining PID control constants are stored in the calculation unit 67 in the same manner as in the case of the flatness standard setting unit 60, and the inference mechanism performs inference. The result is given to the flatness control section 63.

The mode selection unit 68 determines whether the sheet is in a passing state, an acceleration/deceleration state, or a poor rolling condition based on the rolling schedule from the rolling schedule setting unit 8 and the rolling actual values from the six rolling actual value detection units, The operation is performed as follows according to the judgment result.

For example, turn off the leveling 71 when threading.

WR vendor 72 is ON, IMR vendor 73 is 0NS
Turn off WR shift 74, turn off IMR shift 75,
Selector 64 to turn on roll coolant 76
Issue operation commands to.

During acceleration/deceleration and rolling, the ON/OFF state of each actuator is changed depending on the steel type, plate thickness, plate width, speed, etc. In addition, other sub-control functions (not shown) of the rolling mill, such as
Tension control, running distance change control, shear control (not shown),
The ON/OFF state of each actuator is changed depending on the state of coiler control, automatic plate thickness control, etc. As in the case of the flatness standard setting unit 60, these are made into rules, inferred by an inference mechanism, and a command regarding the ON/OFF state of each actuator is given to the selector 64 according to the inference result.

The flatness control unit 63 performs calculations of equations (1) to (6) and PID calculations, and the flatness control unit 63 performs calculations of equations (1) to (6) and PID calculations, depending on the ON/OFF state of each actuator commanded by the mode selector 68.
6) Modify and use the formula. This is also the flatness standard setting section 60
A rule is prepared in the same way as above, inference is made according to the contents of the command from the mode selection unit 68, and a flatness control calculation is performed according to the result of this inference.

As described above, the present invention rules the experience and knowledge of engineers, operators, and plate users according to rolling schedules and actual rolling values, and implements flatness control inferred by an inference mechanism, so that changes in equipment and operations are not affected. Optimal flatness control can be achieved accordingly. This is the second feature of the present invention.

(Effect of the embodiment) In the embodiment described above, as described in the first feature, the flatness control with weighting coefficients expressed by equations (1) to (6) can be performed at high speed by using the influence coefficient. By doing so, a plate with good flatness can be obtained.

In addition, as mentioned in the second feature, the knowledge and experience of engineers, operators, users of rolled plates, etc. is used in rules, inference is made by an inference mechanism, and flatness control is performed based on the results. , it is possible to obtain an image with better flatness.

(Other Embodiments) In the embodiments, a 6-high rolling mill was described, but the present invention can be applied to rolling mills with other numbers of stages, such as 2-high, 4-high, 8-high, 12-high, 20-high, etc. Can be applied. In that case, the number of actuators increases or decreases depending on the number of stages. Further, in equations (1) to (6), the shift can be applied in exactly the same way to flatness control where the shift is used only for setting.

In that case, the terms of WR shift and IMR shift become zero in equations (1) to (6).

In addition, in the above embodiment, the flatness control of the plate was described, but by changing the flatness meter to a crown meter, the crown (thickness in the width direction of the plate) of the plate can be controlled in exactly the same manner as above. can do.

〔Effect of the invention〕

As detailed above, according to the first feature of the present invention, shape control with weighting constants expressed by equations (1) to (6) is performed at high speed using influence coefficients, thereby achieving a good shape. You can get a good board.

In addition, according to the second feature, the shape is improved by making inferences using the inference mechanism using the knowledge and experience of engineers, operators, users of the rolled plate, etc., and controlling the shape according to the results. You can get a good board.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is an explanatory diagram for explaining the roll configuration of a rolling mill and the action of each actuator, and FIG. 3 is a detailed configuration of an embodiment of the present invention. FIG. 4 is an expert system diagram of each device part related to the intelligent control in FIG. 3. 1... Work roll (WR), 2... Intermediate roll (IMR), 3... Backup roll (BUR),
4... Material to be rolled, 5... Flatness needle, 6... Flatness control device, 60... Flatness standard setting section, 63... Flatness control section, 64... Selector , 66... Reference correction calculation section, 67... Influence coefficient and control constant calculation section, 68
...Mode selection section, 69...Rolling actual Ht1ii! Detection unit, 7... Actuator, 71... Leveling, 72... WR bender, 73... IMR bender, 74... WR shift, 75... IMR shift, 7
6... Roll coolant, 8... Rolling schedule setting unit, 101... Knowledge base, 102... Knowledge compiler, 103... Knowledge internal representation, 104... Inference mechanism, A... Subject Rolled material feeding direction. Applicant's agent Mr. Sato

Claims (1)

[Claims] 1. A method for controlling the shape of a plate, which includes a leveling, a bender, a shift, and a coolant as actuators for controlling the shape of the plate, and uses these actuators to control the shape of the plate on the exit side of a rolling mill, The product of the shape deviation obtained as the difference between the shape reference set by dividing the shape of the plate material into multiple positions and the corresponding shape measurement value, the weighting coefficient at the relevant position, and the leveling, bender, and shift operations at the relevant position. The amount of operation for leveling, bending, and shifting is determined so that the value obtained by squaring the difference between the sum of the products of the amount, influence coefficient, and weighting coefficient and integrating it for each position is minimized,
Leveling from the product of the product of the coolant operation amount, influence coefficient and weighting coefficient for each position, and the shape deviation and weighting coefficient for each position;
A method for controlling the shape of a plate material, characterized in that a coolant operation amount for each position is determined from each operation amount of a bender and a shift, and a value obtained by subtracting the sum of the products of an influence coefficient and a weighting coefficient. 2. The method for controlling the shape of a plate material according to claim 1, characterized in that the output of the calculation in each control process is determined by inference by an inference mechanism based on a knowledge base in which knowledge and experience regarding shape control of the plate material are made into rules. . 3. In a plate shape control device that is equipped with leveling, bender, shift, and coolant as actuators for controlling the shape of the plate, and controls the shape of the plate on the exit side of the rolling mill by these actuators, it is possible to control the shape of the plate at multiple positions. The product of the shape deviation obtained as the difference between the shape standard set by dividing and the corresponding shape measurement value and the weighting coefficient at the relevant position, and the operation amount, influence coefficient, and weight at the relevant position of leveling, bender, and shift a first calculation means for determining each operation amount of leveling, bender, and shift such that the value obtained by integrating the square of the difference between the sum of the products of the coefficients for each position is the minimum; Leveling is performed from the product of the product of the coolant operation amount, influence coefficient, and weighting coefficient for each position, and the product of the shape deviation and weighting coefficient for each position.
A shape of a plate material characterized by comprising: a second calculation means for determining a coolant operation amount for each position from a value obtained by subtracting the sum of the products of the bender and shift operation amounts, an influence coefficient, and a weighting coefficient. Control device. 4. The first calculation means and the second calculation means each include a knowledge base that expresses knowledge and experience regarding shape control of plate materials as rules, and an inference mechanism that infers calculation outputs in each control process based on the knowledge base. 4. The plate shape control device according to claim 3, further comprising: