JP3065205B2 - Control method of tandem rolling mill - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to control of a tandem rolling mill in continuous rolling of a wire or a steel bar.

[0002]

2. Description of the Related Art Wire rods and steel bars are conventionally rolled by tandem rolling mills. In this type of rolling, a control method disclosed in, for example, Japanese Patent Application Laid-Open No. 62-173018 has been conventionally used to precisely control the dimensions of a rolled material. In this control method, the rolling position and rolling load of each rolling stand and a predetermined gauge
Based on the equation, the estimated plate thickness value at the exit side of each rolling stand is obtained, and the estimated plate thickness value and the inter-stand tension value of the rolled material are input to the optimal regulator control, and the roll thickness for each rolling stand is determined. -The target value of the roll rotation speed and the target value of the gap between the rolls are determined.

[0003]

However, it is very difficult to actually measure the tension of the rolled material between the rolling stands. Therefore, the tension value of the rolled material input to the optimal regulator control was calculated from the rolling torque,
In fact, it is necessary to adopt an estimated value. However, such an estimated value of the tension, if there is a disturbance,
This often involves large errors. for that reason,
Rolling performance is degraded, and large errors tend to occur in the dimensions of the rolled material.

[0004] Accordingly, it is an object of the present invention to prevent rolling performance from deteriorating when there is disturbance.

[0005]

In order to solve the above-mentioned problems, according to the present invention, there is provided a method for controlling a tandem rolling mill for rolling a bar or a wire rod: a cross-sectional area provided on an outlet side of each rolling stand. The cross-sectional area of the material to be rolled is measured by the measuring means (5, 6, 7, 8).
The rotational speed deviation is measured, and the deviation amount (Δ
ai (k) _* ) , the measured roll rotational speed deviation (ΔNi
(k) _* ) and the time derivative of the measured roll rotational speed deviation
Based on ((d / dt) ΔNi (k) _* ) , the operation amount (ΔSrefi (k) of the roll rotation speed of each rolling stand required to bring the cross-sectional area of the rolled material close to its target value. ) _* ) Is generated.

According to a second aspect of the present invention, there is provided a method for controlling a tandem rolling mill for rolling a bar or a wire rod: a cross-sectional area measuring means (5, 5
6, 7, 8), the cross-sectional area of the material to be rolled is measured, and the rolling load deviation (ΔP i (k) _* ) , the inter- roll gap deviation (ΔSi (k) _* ) and the roll of each rolling stand are measured. Rotational speed deviation (ΔNi
(k) _* ) is measured, and the height deviation (Δhi (k) _* ) of the material to be rolled at the exit side of each rolling stand is calculated based on the measured rolling load deviation and the measured gap deviation between the rolls. Deviation of the calculated cross-sectional area (Δai (k) _* ) , the calculated height deviation of the workpiece,
Measured roll spacing deviation, measured roll rotation speed deviation, and time derivative of measured roll rotation speed deviation
Based on ((d / dt) ΔNi (k) _* ) , the amount of operation of the rolling reduction position of each rolling stand required to bring the cross-sectional area and height of the rolled material close to their target values ( ΔSrefi
(k) _* ) and the operation amount of the roll rotation speed (ΔNrefi (k) _* )
Generate

[0007] The description in each parenthesis indicates a corresponding element in an embodiment described later.

[0008]

In any of the first and second aspects of the present invention, the cross-sectional area of the material to be rolled is measured by the cross-sectional area measuring means installed on the exit side of each rolling stand, and control is performed based on the measurement result. Will be implemented. Since the control is performed based on the measured cross-sectional area, there is no need to input a tension value. Therefore, no error occurs due to the estimation of the tension. The cross-sectional area of the material to be rolled can be directly measured, and no significant error occurs even when subjected to disturbance. Therefore, the rolling performance is improved and the dimensional error is reduced.

[0009]

FIG. 1 shows the configuration of a control system according to a first embodiment. This control system controls a four-stand mill used for rolling a wire or a steel bar. Referring to FIG. 1, the rolled material is rolled while being conveyed from left to right as indicated by arrows. In the four rolling stands # 1, # 2, # 3 and # 4, the rolling direction between adjacent stands is 90 degrees.
The vertical stands and the horizontal stands are alternately arranged so as to change degrees. Generally, in this type of control system, the rolling material dimension in the rolling direction is set to a height (or thickness).
, And the dimension of the rolled material in a direction perpendicular thereto is referred to as a width. Therefore, it should be noted that also in this control system, the directions of the height and the width alternately change for each stand.

Although not shown, four rolling stands # 1,
In # 2, # 3 and # 4, the roll is driven to rotate.
Rotating drive device, and a reduction drive for adjusting the gap between the rolls.
Equipped with a moving device. Rotary drives are fast
Controlled by the temperature controllers 21, 22, 23 and 24.
Each of the rolling-down driving devices includes a rolling-down control device 11,
2, 13 and 14. As shown in FIG.
The four rolling stands # 1, # 2, # 3 and # 4
On each exit side, a total of 5, 6, 7, and 8 cross sections are installed.
ing. These cross sections 5, 6, 7 and 8 are
The cross-sectional area of the rolled materiala ₁ (k), a _Two (k), a _Three (k)as well asa
_Four (k)Is detected. About the configuration and operation of this cross section meter
Will be described later.

In this embodiment, speed control devices 21, 22, 23, and 2 are controlled by using a computer that performs optimal control.
By controlling each control target value (correction amount of drive speed) ΔNrefi (k) _* of No. 4, control is performed so that the cross-sectional area of the rolled material at the exit side of each rolling stand becomes constant. Here, i means a stand number, and in this example, i = 1 to 4
It is.

Hereinafter, the optimum control system in the apparatus shown in FIG. 1 will be described. Here, a roll rotation speed control system (speed control system) was approximated by a secondary system as a dynamic characteristic of a mill drive system when creating a rolling model. The symbols of the control system are shown in Table 1 below, and the reference states are shown in Table 2.

[0013]

[Table 1]

[0014]

[Table 2]

When this non-linear rolling model is linearized and dimensionlessly displayed around the reference states shown in Table 2 and displayed as a state space model, and further discretized with a control period Δt, the following state equation is obtained.

[0016]

X (k + 1) = Ax (k) + Bu (k) + Ew (k) (1) y (k) = Cx (k) + Fw (k) (2) where k Is a sample number (integer), and the state vector x (k) , output y (k) , input u (k) , and disturbance w (k) are as follows.

[0017]

(Equation 2)

The matrices A to C are matrices in a rolling standard state, and Table 2 shows an example of a rolling standard state in a 4-stand mill. The values of E and F are unnecessary for the control system design described below.

By applying the optimal control theory to the present system, a discrete-time optimal control system was designed as follows. That is, disturbance
Assuming that w (k) is step-shaped, the evaluation function J and the control input u (k) that minimizes it are given by the following equations (7) and (8), respectively.

[0020]

(Equation 3)

Here, the control deviation e (k) and the partial state xs (k)
And the first order difference v of the input u (k) are the following (9)
Equation (10) and Equation (11).

[0022]

(Equation 4)

Further, the optimum gain matrix F = [F ₁ | F ₂ | F
₃ ] is obtained from the following Riccati equation.

[0024]

(Equation 5)

Here, the matrices Ф, G and H are based on the error system of the system equation (Equations (1) and (2)) (the following (1)
4-16) is a coefficient matrix obtained from the matrices A to C.

[0026]

(Equation 6)

Controls the speed control devices 21 to 24 shown in FIG.
The details of the processing performed by the computer are as follows. First,
The cross-sectional area detected by the cross-section meter 5 in the calculating unit 31a ₁ (k)Or
DeviationΔa ₁ (k) _* And the calculation unit 32 calculates
Cross section area detected by cross section meter 6a _Two (k)Deviation fromΔa
_Two (k) _* Is calculated, and the cross section meter 7 detects the
Cross sectiona _Three (k)Deviation fromΔa _Three (k) _* And calculate
The cross-sectional area detected by the cross-sectional meter 8 in the part 34a _Four (k)From
The deviationΔa _Four (k) _* Ask for. This is done every predetermined time Δt
repeat.

Calculated deviation amountΔa ₁ k _* , Δa _Two (k) _* , Δa
_Three (k) _* , Δa _Four (k) _* Is the output y (k), and
The control deviation e (k) is obtained from the equation (9). The arithmetic unit 41
Of the # 1 stand detected by the speed control device 21
Roll rotation speedN ₁ (k)(Rpm), the deviation amountΔN
₁ (k) _* And its differential value, and the arithmetic unit 42 calculates the speed
Of the # 2 stand detected by the degree control device 22
speedN _Two (k)From the deviationΔN _Two (k) _* And its derivative
Calculated and detected by the speed control unit 23 in the arithmetic unit 43
Roll rotation speed of # 3 standN _Three (k)From that
DeviationΔN _Three (k) _* And its differential value,
Of the # 4 stand detected by the speed control device 24
Roll rotation speedN _Four (k)From the deviationΔN _Four (k) _* Toso
Find the derivative of. Roll rotation speed deviation of each standΔ
Ni (k) _* , And its derivative(d / dt) ΔNi (k) _* The partial
State vectorx _S (k)And

The optimum regulator 35 calculates the above equation (8) based on the input control deviation e (k) and the partial state vector x _S (k), and obtains the optimum control input u.
(k), that is, ΔNref1 (k) _* , ΔNref2 (k) _* , ΔNref3
(k) _*, and seek ΔNref4 (k) _*. These ΔNref1
(k) _* , ΔNref2 (k) _* , ΔNref3 (k) _* , and ΔNref4 (k) _*
However, in order to correct the target value Nrefi (k) of the roll rotation speed, the speed control devices 21, 22, 23 and 24 are used, respectively.
Is applied to

That is, based on the cross-sectional area of the rolling material actually detected at the exit side of each rolling stand, the rolling speed of each stand is corrected so as to be optimum. As a result, stable dimensional control that is hardly affected by disturbance is realized.

Next, the control system of the second embodiment shown in FIG. 2 will be described. Also in this embodiment, the control object is the same four-stand mill as in the first embodiment, and the rolling drive devices of each stand are controlled by the draft control devices 11 to 14, and the rotary drive devices of each stand are the speed control devices 21 to 24.

In this embodiment, each control target value (a correction amount of the driving speed) ΔNrefi (k) _{* of} the speed control devices 21, 22, 23, and 24 is calculated by using a computer that performs the optimum control. , 12, 13 and 14 by controlling ΔSrefi (k) _* to control the cross-sectional area of the rolled material at the exit side of each rolling stand to be constant.

An optimal control system in the apparatus shown in FIG. 2 will be described. Here, as a dynamic characteristic of a mill drive system, a roll rotation speed control system (speed control system) was approximated by a secondary system and a roll reduction control system was approximated by a first-order lag as a dynamic characteristic of a mill driving system.
The symbols and reference states of the control system are the same as those in Tables 1 and 2. Also, Si (k) and Srefi (k) are the inter-roll gap of each stand and its target value (i is the stand number).

When this non-linear rolling model is linearized and dimensionlessly displayed around the reference states shown in Table 2 and displayed as a state space model, and further discretized with a control period Δt, the following state equation is obtained.

[0035]

X (k + 1) = Ax (k) + Bu (k) + Ew (k) (17) y (k) = Cx (k) + Fw (k) (18) where k Is a sample number (integer), and the state vector x (k) , output y (k) , input u (k) , and disturbance w (k) are as follows.

[0036]

(Equation 8)

The optimal control theory was applied to this system, and a discrete-time optimal control system was designed as follows. That is, disturbance
Assuming that w (k) is step-shaped, the evaluation function J and the control input u (k) that minimizes it are given by the above-mentioned equations (7) and (8), respectively.

The control deviation e (k) , the partial state xs (k) and the first-order difference v (k) of the input u (k) are respectively calculated by the above (9)
Equation (10) and Equation (11).

Further, the optimum gain matrix F = [F ₁ | F ₂ | F
₃ ] is obtained from the Riccati equation shown in the above equations (12) and (13). Here, the matrices Ф, G and H are the error systems (the (14) and (14), respectively, of the system equations (the equations (17) and (18)).
16) is a coefficient matrix obtained from the matrices A to C.

The contents of the processing of the computer which controls the speed control devices 21 to 24 and the pressure reduction control devices 11 to 14 shown in FIG. 2 are as follows. First, in the calculating unit 31, the deviation Δa ₁ (k) _* from the cross-sectional area a ₁ (k) detected by the cross-section meter 5
The deviation Δa ₂ (k) _* is calculated from the cross-sectional area a ₂ (k) detected by the cross-section meter 6 in the arithmetic unit 32, and the arithmetic unit 33
, The deviation Δa ₃ (k) _* is obtained from the cross-sectional area a ₃ (k) detected by the cross-sectional meter 7, and the arithmetic unit 34 calculates
The deviation Δa ₄ (k) _* is obtained from the detected cross-sectional area a ₄ (k) . This is repeated every predetermined time Δt.

In the calculating section 37, the inter-roll gap value ΔS output by the rolling reduction devices 11 to 14 of the respective rolling stands.
i (k) _* and rolling load value ΔPi (k) _* are input, and based on a known gage meter formula, the deviation amount of the cross-section height of the rolled material on the exit side of each stand.

[0042]

(Equation 9) Ask for.

The obtained deviation amounts Δa ₁ (k) _* , Δa ₂ (k) _* , Δa
₃ (k) _* , Δa ₄ (k) _* and

[0044]

(Equation 10) Is the output y (k), and the control deviation e from (9)
Find (k).

In the calculation section 41, the speed control device 21
Roll rotation speed N ₁ (k) (r ) of the # 1 stand detected at
pm), the deviation ΔN ₁ (k) _* and its differential value are obtained, and the arithmetic unit 42 calculates the deviation amount ΔN ₁ (k) _* from the roll rotation speed N ₂ (k) of the # 2 stand detected by the speed control device 22. the deviation amount ΔN ₂ (k) _* and the determined differential value, the calculator 43, the # 3 stands detected by the speed control device 23 b - Le rotational speed N ₃ (k), the deviation amount .DELTA.N ₃ (k) _* and its differential value are calculated, and the arithmetic unit 44 calculates the speed control device 2
From the roll rotation speed N ₄ (k) of the # 4 stand detected in step 4, the deviation ΔN ₄ (k) _* and its differential value are obtained. The roll rotation speed deviation ΔNi (k) _{* of} each stand and its differential value (d / dt) ΔNi (k) _* are defined as a partial state vector x _S (k) .

The optimum regulator 36 calculates the above equation (8) based on the input control deviation e (k) and partial state vector x _S (k), and obtains the optimum control input u.
(k), that is, ΔNref1 (k) _* , ΔNref2 (k) _* , ΔNref3
(k) _* , ΔNref4 (k) _* , ΔSref1 (k) _* , ΔS
ref2 (k) _* , ΔSref3 (k) _* , and ΔSref4 (k) _* are obtained.
ΔNref1 (k) _* , ΔNref2 (k) _* , ΔNref3 (k) _* , and ΔN
ref4 (k) _* is used to correct the target value Nrefi (k) of the roll rotation speed, so that the speed controllers 21, 22, 23 are respectively provided.
, Sref1 (k) _* , ΔSref2 (k) _* , ΔSref1 (k) _*
Sref3 (k) _* and ΔSref4 (k) _* are applied to the reduction controllers 11, 12, 13 and 14, respectively, in order to correct the target value Srefi (k) of the inter-roll gap.

That is, the cross-sectional area deviation (Δai (k) _* ) of the rolled material actually detected at the exit side of each rolling stand and the gauge
Rolled material cross-section height deviation (Δhi (k) _* )
Is modified so that the roll rotation speed of each stand is optimized. As a result, stable dimensional control that is hardly affected by disturbance is realized.

Next, the cross sections 5 to 5 used in each of the above embodiments were used.
8, the structure and operation will be described. FIG. 3 shows the configuration of one cross section meter. In FIG. 3, 51 is an oscillator, 52
Is a power amplifier that amplifies the output of the oscillator 51 and has a constant current function, 53 is a DC power supply, M is a test material, S is a detection unit,
Reference numeral 56 denotes a magnetizing coil for completely saturating the test material M.

FIG. 4 shows a cross section of the detecting section S. Detector S
Is a penetrating coil with a coil wound on a concentric axis,
The coil configuration of the detection unit S through which the test material M passes through the concentric shaft portion is such that reception coils 55a and 55b are provided side by side with the transmission coil 54 interposed therebetween, and a magnetizing coil 56 is further provided on the outer circumference thereof. It is wound. Receiving coil 55a, 5
The output signal of 5b is amplified by the amplifiers 57a and 57b, input to the synchronous detectors 58a and 58b together with the detection reference signal f, and is converted into a DC signal. 59a, 59b
Is a differential amplifier, and c is a bias adjuster. The difference amplifiers 59a and 59b subtract a predetermined bias voltage set by the bias adjuster c from the output signals of the synchronous detectors 58a and 58b. The bias voltage input to the difference amplifier 59a is equivalent to the synchronous detection output obtained when the air core is empty (the state where the test material M is not inserted).
a becomes zero output. The bias voltage input to the differential amplifier 59b is equivalent to the output of a synchronous detector obtained when the center of the detection unit S and the test sample M are matched at the time of being centered. Similarly, the differential amplifier 59b has zero output, and 6
Numeral 0 denotes a DC amplifier, which amplifies a deviation signal generated due to a positional shift of the test material M under a preset gain. This deviation signal becomes an error correction amount and is input to the difference calculator 61 in the next stage. The difference calculator 61 includes a DC amplifier 60
Of the output signal of the differential amplifier 59a. The calculation result is linearized by the linearizer 62 and becomes a final output.

By the above-described difference calculation by the difference calculator 61, the error due to the positional deviation of the test material M caused by the non-linearity of the magnetic flux density distribution has been eliminated, and highly accurate cross-section detection has been realized. The aforementioned bias adjuster c or DC amplifier 60
Is set in advance according to the detection target (test material M) by preparing a calibration table for each standard sample.

Next, an embodiment in which the transmitting and receiving coils of the detecting section S are installed under the above-described arrangement conditions will be described. As shown in FIG. 4, the receiving coils 55a, 55a,
5b was arranged. Coil specifications and their intervals L1, L
2 is as follows.

Transmission coil 54: Inside diameter 55mm Length 10mm Number of turns 10 turns Receiving coil 55a, 55b: Same specification Inner diameter 55mm Length 10mm Number of turns 10 turns Coil distance: L1 3mm L2 13mm Between coils satisfying the above-mentioned arrangement conditions Distance L1, L2
Are 3 mm and 13 mm, respectively, in this example. The alternating current applied to the transmitting coil 54 has a frequency of 200 kHz.
Is a sine wave.

FIG. 5 shows the receiving coils 55a and 55 when the sample axis (the center line of the steel bar to be measured) in the above embodiment is aligned with the center axis of the detection coil (the amount of sample movement is 0 mm).
The output change (deviation from the reference) due to the radial movement of the test material (steel bar) with reference to the output signal of b is shown. E in the figure
1 and f1 are the sample (bar) diameter 35 mm and 2
Is the deviation output obtained from the receiving coil 55a at 5 mm, and g and h are the same as those of the receiving coil 55b.
Is the deviation output obtained from As is apparent from FIG. 5, if only the receiving coil 55a is used, even if the metal conductor has the same dimensions, the deviation output increases as the coil conductor approaches the coil edge, and the detection accuracy decreases. On the other hand, as can be seen from the characteristics g and h of the correction receiving coil 55b, the same deviation output is obtained in opposite phases. However, the equivalent is
This is the result of adjusting the output level by the DC amplifier 60. If the deviation output of g and h is subtracted from the output of the difference amplifier 59a of the receiving coil 55a, it is possible to correct the error due to the radial displacement of the steel bar.

FIG. 6 shows error curves e2, f2 and e0, f0 before and after correcting the positional deviation of the sample (steel bar). The X axis indicates the amount of movement of the sample, and the Y axis indicates the cross-sectional area error. e0, e2 and f0, f2 are characteristic curves for sample diameters of 35 mm and 25 mm, respectively. In FIG. 6, the characteristics e2 and f2 before correction have a sample position deviation of 5 mm.
Indicates that an error of about 2% has occurred. On the other hand, after the correction, the error was suppressed to within about 0.15%. Especially, the error is 0% at each point of the sample moving distance of 5 mm.
The reason for this is that the gain of the DC amplifier 60 is set at this point.

FIG. 5 shows the corrected detection output of each sample. The X-axis is the cross-sectional area, and the Y-axis is the output after linear correction by the linearizer 62. It can be seen that a good detection output corresponding to the cross-sectional area is obtained. As described above, the correction effect shown in FIG. 6 appears, and high-precision detection is realized with a simple device configuration. In addition, real-time measurement is possible. Since the variation of the bar and the like manufactured in the actual operation line has a small variation in one lot, if a detection coil is prepared for each bar or the like to be manufactured, more accurate detection can be expected. Needless to say, a series of signal processing can be easily processed by an electronic computer.

In the above-described embodiment, the optimal control system of the discrete time system has been derived.
If equations (1) and (2) are given in a continuous time system, an optimal control system of a continuous time system can be constructed in a similar manner.

[0057]

According to any of the first and second aspects of the present invention, the cross-sectional area of the material to be rolled is measured by the cross-sectional area measuring means installed on the exit side of each rolling stand, and based on the measurement results. Control is performed. Since the control is performed based on the measured cross-sectional area, there is no need to input a tension value. Therefore, no error occurs due to the estimation of the tension. The cross-sectional area of the material to be rolled can be directly measured, and no significant error occurs even when subjected to disturbance. Therefore, the rolling performance is improved and the dimensional error is reduced. In the invention of claim 1, the cross-sectional area of the rolled material on the exit side of each rolling stand is controlled to be constant, and in the invention of claim 2, the cross-sectional height of the rolled material is further controlled to be constant.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a control system according to a first embodiment.

FIG. 2 is a block diagram illustrating a configuration of a control system according to a second embodiment.

FIG. 3 is a block diagram showing a configuration of one cross section meter.

FIG. 4 is a longitudinal sectional view of a detection unit S shown in FIG.

FIG. 5 is a graph showing detection outputs of two receiving coils in one embodiment.

FIG. 6 is a graph showing a measurement error of a cross-sectional meter due to a positional shift of a test material.

FIG. 7 is a graph showing a sectional area detection output.

[Explanation of symbols]

# 1, # 2, # 3, # 4: Rolling stand 5, 6, 7, 8: Cross section meter 11, 12, 13, 14: Reduction control device 21, 22, 23, 24: Speed control device 31, 32, 33, 34, 37: calculation units 35, 36: optimal regulator units 41, 42, 43, 44: calculation unit S: detection units 55a, 55b: reception coil M: test material 56: magnetization coil 51: oscillator 57a , 57b: Amplifier 52: Power amplifier 58a, 58b: Synchronous detector 53: DC power supply 59a, 59b: Difference amplifier 54: Transmission coil 60: DC amplifier

Continuation of the front page (56) References JP-A-62-244509 (JP, A) JP-A-4-322809 (JP, A) JP-A-6-285521 (JP, A) (58) Fields studied (Int .Cl. ⁷ , DB name) B21B 37/16-37/20

Claims (2)

(57) [Claims] In a control method of a tandem rolling mill for rolling a bar or a wire rod, a cross-sectional area of an object to be rolled is measured by a cross-sectional area measuring means provided at an outlet side of each rolling stand, and a roll of each rolling stand is reduced. The roll rotation speed deviation is measured, the measured cross-sectional area deviation amount, the measured roll rotation speed deviation,
And, based on the time differential value of the measured roll rotational speed deviation, generating a manipulated variable of the roll rotational speed of each rolling stand required to bring the cross-sectional area of the rolled material closer to its target value. And a method for controlling a tandem rolling mill. 2. A method for controlling a tandem rolling mill for rolling a bar or a wire rod: measuring a cross-sectional area of an object to be rolled by a cross-sectional area measuring means provided at an outlet side of each rolling stand, and rolling load of each rolling stand. The deviation, the inter-roll gap deviation and the roll rotation speed deviation are measured, and based on the measured rolling load deviation and inter-roll gap deviation, the height deviation of the material to be rolled at the exit side of each rolling stand is calculated. The measured deviation of the cross-sectional area, the calculated height deviation of the workpiece, the measured deviation between the rolls, the measured roll rotation speed deviation, and the time derivative of the measured roll rotation speed deviation On the basis of the above, to generate the operation amount of the rolling reduction position of each rolling stand and the operation amount of the roll rotation speed required to bring the cross-sectional area and the height of the rolling object close to their target values. Tandem rolling mill control Your way.