JPH03254308A - Automatic control method of thickness for rolling mill - Google Patents

Abstract

PURPOSE:To make the automatic thickness control of high accuracy possible by changing the setting of the increasing and decreasing quantity per one control cycle when adjusting the increase and decrease of a prescribed control gain value into smaller as the prescribed control gain value approaches its optimum value. CONSTITUTION:The deviation of thickness on the inlet side of a rolled stock 1 is measured with a thickness gage 5 on the inlet side and outputted to a control computing element 12. In the computing element 12, roll gap is calculated by performing calculation based on the constant that is inputted from a numerical setter 13 and the control gain value from a memory 16 with computing elements 14, 15 and outputted to a output timing circuit 8. After the delay time that is equivalent to the running time of the rolled stock 1, the roll gap is outputted to the hydraulic rolling-down controller 9 of rolling mill with the output timing circuit 8 and the feedforward control of the rolling reduction of roll is executed. Then, because the compensation of zero of the deviation of thickness on the outlet side can't be done with the feedforward control in an open loop, feedback control using a feedback controller 10 is used simultaneously.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method for automatically controlling the thickness of a rolled material in a rolling mill, and in particular, a method for automatically controlling the thickness of a rolled material in a rolling mill, and in particular, a method that uses a feedforward control function and an automatic gain control function. The present invention relates to an automatic plate thickness control method.

[Prior Art] Generally, a feedback control system is employed as a means for automatic plate thickness control in conventional rolling mills. In this feedback control method, it is difficult to perform highly accurate control without delay due to a time delay in the plate thickness detection signal or a time delay in the drive mechanism.

Therefore, conventionally, a feedforward control method has been introduced. Typically, feedforward control of a rolling mill is performed as follows.

First, the plate thickness deviation Δ measured by the plate thickness gauge on the entrance side of the rolling mill
Based on H, the roll gap ΔS is calculated from the following equation (1) or (2).

ΔS=(m/M)・ΔH・(1) ΔS=A・(m7M)・ΔH・(2) Here, m is the plastic constant of the rolled material, M is the mill constant, and A is the correction coefficient (control The plastic constant m and Mill constant M are set in advance as constants for calculation, but the plastic constant m changes depending on the temperature and degree of plastic working even for the same material. Therefore, the control gain value A is set as a correction coefficient, and the roll gap ΔS is calculated by equation (2), which is multiplied by the right side of equation (1), and the control gain value A is Control accuracy is improved by providing a separate adjusting means (described later) that automatically makes small corrections.

The calculated roll gap ΔS is output to the rolling control device at the time when the position of the rolled material where the entrance plate thickness deviation ΔH used in the calculation was measured reaches the rolling roll of the rolling mill, and is calculated based on the roll gap ΔS. The roll reduction amount is controlled and feedforward control is performed.

Since the feedforward control performed in this manner is an open loop and cannot correct the zero point of the outlet thickness deviation Δh, feedback control including an integral element is usually used in combination.

On the other hand, the optimal value of the control gain value A varies depending on the material and dimensions of the rolled material 1, so it is set and inputted prior to rolling as described later, and this control gain value (gain) A is automatically set during rolling. A G C (A ut
automatic gain control) is used.

In this AGC, for example, the plate thickness deviation after rolling (output plate thickness deviation) Δh at the measurement point of the inlet plate thickness deviation ΔH of the rolled material is measured, and the inlet plate thickness deviation ΔH and the outlet plate thickness deviation at the same point are measured. Δ
The control gain value A is increased or decreased based on h. This AGC operation can be summarized as follows.

(2) If the absolute value of the exit plate thickness deviation Δh is less than a predetermined dead band value, the control gain value A is not changed.

■If the absolute value of the exit side plate thickness deviation Δh is greater than the predetermined dead band value,
If the inlet side plate thickness deviation ΔH and the outlet side plate thickness deviation Δh have the same sign, it is determined that the amount of adjustment of the roll gap ΔS is small, and this is to be increased.

The control gain value A is increased by a certain amount (for example, 0.l).

■If the absolute value of the exit side plate thickness deviation Δh is greater than the predetermined dead band value,
If the inlet side plate thickness deviation ΔH and the outlet side plate thickness deviation Δh have different signs, it is determined that the adjustment amount of the roll gap ΔS is large, and in order to reduce this, the control gain value A is set by a certain amount (for example, 0.1 ).

In this way, the excess or deficiency of the control gain value A set and held in the variable area is adjusted to obtain the optimal feedforward control operating point.

However, in the automatic sheet thickness control means as described above, the control gain value A is adjusted to increase or decrease depending on the sign of the entrance side sheet thickness deviation ΔH and the exit side sheet thickness deviation Δh. If it is not possible to compare the signs of the inlet thickness deviation ΔH and the outlet thickness deviation Δh at the same point, the direction of increase or decrease of the control gain value A may be reversed, which may increase the outlet thickness deviation instead. . Since it is difficult in actual operation to always make the tracking error zero at various rolling speeds, it may not be possible to control the exit plate thickness with high accuracy.

Conventionally, when the average amplitude value of the outlet plate thickness deviation Δh of a rolling mill is not within a predetermined dead band value, the ratio of the average amplitude value of the outlet plate thickness deviation Δh to the average amplitude value of the inlet plate thickness deviation ΔH has been conventionally used. A method has also been proposed in which the control gain value A is corrected depending on the direction of change before and after the passage of time to prevent poor control results even if there is some tracking error (see Japanese Patent Application No. 62-316129). ).

[Problems to be Solved by the Invention] In any case, in the above-mentioned conventional automatic plate thickness control means, when adjusting the control gain value A, the increase/decrease (-amount per control cycle) is always fixed. There is. For this reason, as shown in the topsoil, if the increase or decrease of the control gain value A is fixed, the speed at which the control gain value A converges to the optimal value, the degree of convergence (how close it can get to the optimal value) ) cannot be improved together. Therefore, conventionally, the increase/decrease is set to the 'middle' value in Table 1 below.

Table Also, O: Good, Δ: Normal, ×: Bad The present invention attempts to solve such problems, and by making the increase/decrease of the control gain value variable, the optimum value of the control gain value can be determined. An object of the present invention is to provide an automatic plate thickness control method for a rolling mill that can further improve plate thickness accuracy by improving the convergence speed and degree of convergence.

[Means for Solving the Problems] In order to achieve the above object, the automatic plate thickness control method for a rolling mill of the present invention changes the increase/decrease per control cycle when increasing/decreasing a predetermined control gain value. The present invention is characterized in that the increase or decrease is set smaller as the predetermined control gain value approaches the optimum value.

[Function] In the above-described automatic plate thickness control method for a rolling mill of the present invention, as the predetermined control gain value approaches its optimum value, the increase/decrease per control cycle when adjusting the predetermined control gain value becomes smaller. Since the settings are changed, the speed of convergence toward the optimum value is fast at the beginning of control, and as the value approaches the optimum value, the degree of convergence to the optimum value improves.

[Embodiments of the Invention] Hereinafter, an automatic plate thickness control method for a rolling mill as an embodiment of the present invention will be explained with reference to the drawings.
(e) is a flowchart for explaining the procedure, FIG. 2 is a block diagram showing an apparatus to which the method of this embodiment is applied, and FIG. 3 is a graph showing the effect in comparison with the conventional method.

First, the configuration of the apparatus to which the method of this embodiment is applied will be explained with reference to FIG. 2. In this FIG. 4 is a rolling roll of a rolling mill that rolls the rolled material 1; 5 is an inlet side plate thickness gauge that measures the thickness of the rolled material 1 before rolling and outputs it as an inlet thickness deviation ΔH. , 6 is an outlet thickness gauge that measures the thickness of the rolled material 1 after rolling and outputs it as an outlet thickness deviation Δh, 7 is a pulse generator that detects the input speed of the rolling mill, and 8 is an input plate thickness meter 5. The control calculator 1 takes into consideration the traveling time of the rolled material 1 to the rolling roll 4.
2 (described later) is an output timing circuit that delays the signal (calculated value of the roll gap ΔS) by a considerable time and outputs it; 9 is a rolling machine hydraulic pressure reduction control device that controls the thickness of the rolled material 1 by changing the roll gap; , 10 is a feedback control device that feeds back the integral amount of the exit side plate thickness deviation Δh to the rolling mill hydraulic pressure reduction control device 9.

Reference numeral 12 denotes a control calculator which receives the measured value (ΔH) of the entrance plate thickness gauge 5 and calculates the rolling reduction amount and corrects and changes the control gain value A. Numerical setting device 13. First computing unit 14, second computing unit 15
．． It is composed of a memory 16 and an AGC calculation controller 18. The numerical value setter 13 is for setting the values of the mill constant M and the plastic constant m used in the above formula (2), and the first calculator 14 is for multiplying the entry side plate thickness deviation ΔH by m/M. , the second arithmetic unit 15 multiplies the arithmetic result from the first arithmetic unit 14 by a control gain value A to output a feedforward output amount (roll gap) ΔS, and the memory 16 holds the control gain value A. The AGC calculation controller 18 is
Based on the entrance side plate thickness deviation ΔH from the entrance side plate thickness gauge 5 and the exit side plate thickness deviation Δh from the exit side plate thickness gauge 6, FIGS. 1(a) to (e)
), the control gain value A is calculated according to the procedure described later and is output to the memory 16.

With the above configuration, plate thickness control (feedforward control) during rolling is performed as follows.

That is, the rolled material 1 is supplied from the input reel 2, rolled by the rolling roll 4, and wound onto the output reel 3.

At this time, the entry side plate thickness deviation ΔH of the rolled material 1 is measured by the entry side plate thickness gauge 5 and outputted to the control calculator 12. In this control calculator 12, based on the constants m and M input from the numerical value setter 13 and the control gain value A from the memory 16, the calculator 14.15 performs calculation according to equation (2). The roll gap ΔS is calculated and output to the output timing circuit 8. The roll gap ΔS is then output by the output timing circuit 8 to the rolling mill hydraulic pressure reduction control device 9 after a delay time corresponding to the running time of the rolled material 1, and feedforward control of the roll reduction amount is performed. Note that since the feedforward control is an open loop and cannot correct the zero point of the exit side plate thickness deviation Δh, feedback control by the feedback control device 1210 is also used.

Next, the control operation of the AGC calculation controller 18, which is a characteristic part of the method of this embodiment, will be explained with reference to FIGS. 1(a) to 1(e).

FIG. 1(a) is a flowchart of the main program. First, step Al)
, when the rolling state is reached, the entrance side plate thickness deviation average value ΔH, which is the average amplitude value of the entrance side plate thickness deviation, is calculated according to the flow shown in FIG. 1(b) (described later) (step A2),
Output side plate thickness deviation average value Δ which is the average amplitude value of outlet side plate thickness deviation
h is calculated according to the flow shown in FIG. 1(C) (described later) (step A3).

Then, the plate thickness deviation attenuation rate Xj is calculated according to the flow shown in FIG. 1(d) (described later) (step A4).

This plate thickness deviation attenuation rate Xj is a value for determining whether to increase or decrease the control gain value A.

Next, it is determined whether the average exit plate thickness deviation value Δh is within the dead zone E (step A5).

As a result of the judgment, when Δh<H, that is, when it is within the dead zone E, sufficient plate thickness accuracy has been obtained, so
The process immediately proceeds to step A7 without changing the control gain value A. On the other hand, when Δh≧E, that is, when it is not within the dead zone E, sufficient plate thickness accuracy has not been obtained, so the process proceeds to step A6. As described later, according to the flow shown in FIG. 1(e), the control gain value A is determined by the plate thickness deviation attenuation rate Xj.
to improve plate thickness accuracy.

In step A7, it is determined whether to continue the calculation or not by referring to the switches on the operation panel. To continue, return to step A2.

Although steps A2 and A3 are shown to be executed sequentially in FIG. 1(a), it goes without saying that parallel processing is more desirable. Furthermore, the order of execution of steps A4 and A5 may be reversed.

The engineering drawing (b) is a flowchart of a subprogram for calculating the entrance side plate thickness deviation average value ΔH in step A2.
The measurement signal from the inlet side plate thickness deviation ΔH is measured at a predetermined sampling pitch (control cycle).

(i=1”n) and read sequentially (step A8),
The absolute value of these entry side plate thickness deviations ΔH1 (ΔH1 → ΔH
il) (step A9), the entrance side plate thickness deviation Δ
The average value ΔH of H is determined by the following equation (3) (step A10). Here, n is the number of samples.

ΔH-” (1/n) (IΔH, l+lΔHZ l0...
... +1ΔHn l ) ... (3) Figure 1 (C)
is a flowchart of a subprogram that calculates the average outlet thickness deviation Δh in step A3, and as shown in FIG. , H to entry side plate thickness deviation.

Find the point at which the rolled material l reaches the exit plate thickness meter 6 at the same point as the measurement position (step A11), and read the exit plate thickness deviation Δhi at the same point as the measurement position of the input plate thickness deviation ΔH (step A11). A12). And, as before,
Absolute value of exit side plate thickness deviation Δhi (Δh, → 1Δhil)
After performing (step A13), the exit side plate thickness deviation Δh,
Find the average value Δh using the following formula (4) (step A1
4).

Δh=(1/n)(lΔh 11 + lΔh2+...
... +1Δh nl) ... (4) Fig. 1(d) is a flowchart of a subprogram for calculating the plate thickness deviation attenuation rate Xj in step A4, and is shown in Fig. 1(d). As such, the plate thickness deviation attenuation rate Xj is determined by the following formula (5) based on the inlet side plate thickness deviation average value ΔH obtained in step AIO and the outlet side plate thickness deviation average value Δh obtained in step A14 (step A15). ).

Xj=Δh/ΔH...(5) The smaller the value of this plate thickness deviation attenuation rate Xj, the lower the control gain value A.
It can be said that the value of is appropriate and the exit side plate thickness deviation is small. Then, it is determined whether Xj is the first value X□ (
Step A16). If the judgment is Yes, the previous Xj
Since there is no (Xj→), the comparison of Xj>Xj (which will be performed in step A17, which will be described later) cannot be performed, so the process returns and proceeds to step A7, while if the determination is No, the process proceeds to step A5.

FIG. 1(e) shows that in step A6, the control gain value A
This is a flowchart of a subprogram for changing the control gain value A, as shown in FIG.
It is determined whether the increase and decrease are repeated N times (an arbitrarily set value) in succession (step A17).

In the case of No determination, it is determined that the control gain value A is still approaching its optimum value, and the process proceeds to step A19 without changing the increase/decrease α, and changes the increase/decrease so far as described later. The control gain value A is changed by the amount α.

On the other hand, in the case of Yes determination, the control gain value A is
It is determined that the change adjustment has resulted in an oscillating state centered on the optimum value (see Figure 3), and in order to bring this control gain value A closer to the optimum value by subsequent change adjustment of the control gain value A. , change the increase/decrease α to a smaller value (for example, α →
α/2; Step A18).

Then, it is determined that Xj>Xj→ (step A19)
). Xj→ is the previous thickness deviation attenuation rate that was calculated last time and temporarily stored, and it is determined whether the current thickness deviation attenuation rate Xj has become larger than the previous thickness deviation attenuation rate Xj→. By doing so, it can be determined whether the current thickness deviation on the exit side is worse than the previous one. In other words, if the determination is Yes, the exit side plate thickness deviation has worsened, and step A21
On the other hand, if the determination is No, the exit side plate thickness deviation has been improved, and the process advances to step A20.

In step A20, it is determined whether or not the previous control gain value was lowered, and if the determination is Yes, the previous control gain value has been lowered and the plate thickness deviation attenuation rate Xj has become smaller, and the plate thickness deviation attenuation rate In order to reduce the control gain value A this time as well in order to reduce Xj, the process advances to step A23. On the other hand, N
In the case of o judgment, the thickness deviation attenuation rate Xj has become smaller by increasing the control gain value A last time, and in order to further reduce the thickness deviation attenuation rate Xj, the step is taken to increase the control gain value A this time as well. Proceed to A22.

In step A21, it is similarly determined whether the previous control gain value A has been lowered. At this time, if the judgment is Yes, the previous control gain value A has been lowered and the plate thickness deviation attenuation rate Xj is large, that is, the output The side plate thickness deviation has worsened, and therefore, this time, the process proceeds to step A22 in order to increase the control gain value A, contrary to the previous time. On the other hand, in the case of a No judgment,
The previous control gain value A was increased and the plate thickness deviation attenuation rate xj was large, that is, the exit side plate thickness deviation was worsened, and therefore,
This time, in order to lower the control gain value A, contrary to the previous time, the process advances to step A23.

In step A22, the control gain value A is increased by an amount of increase/decrease α (the initial value or the value changed in step A18). That is, after adding the increase/decrease α to the previous control gain value A as the current control gain value A, the process proceeds to step A7.

In step A23, the control gain value A is decreased by the amount of increase/decrease α. In other words, after subtracting the increase/decrease α from the previous control gain value A as the current control gain value A, step A
Proceed to step 7.

Further, in this embodiment, step A17. At A18, the control gain value A increases N times (for example, twice) consecutively,
When the decrease is repeated, the increase/decrease α per control cycle for adjusting the change in the control gain value A is made smaller, so as shown in FIG. At the initial stage, the speed of convergence toward the optimal value is extremely fast, and as the value approaches the optimal value, the degree of convergence toward the optimal value becomes extremely good.

In FIG. 3, for example, the optimum value of the control gain value A is 0.07.
In the conventional method, the increase/decrease α is fixed at 0.02, whereas in the method of the present invention, the increase/decrease α is set to 0 from time ○ to T.
．． 08, 0.04 from T□ to T2, 0.04 from T2 to T2.
02, T3 and later, the case where the setting is changed to 0.01 is shown. Looking at the convergence speed to the optimum value in FIG. 3, the time for the control gain value A to converge to 0.12 is T4 seconds in the conventional method, and T0 seconds in the method of the present invention.
T<T4, and it is clear that the method of the present invention is faster. Also, looking at the degree of convergence to the optimal value,
In the conventional method, the control gain value A oscillates at 0.07±0.01, whereas in the method of the present invention, the control gain value A oscillates at an optimum value of 0.0.
It can be seen that it converges to 07.

Therefore, by making the increase/decrease α of the control gain value A variable, the speed at which the control gain value A converges to the optimum value can be increased.

The degree of convergence is greatly improved, and the plate thickness accuracy is further improved.

[Effects of the Invention] As described in detail above, according to the automatic plate thickness control method for a rolling mill of the present invention, as the predetermined control gain value approaches its optimum value, one control when adjusting the increase or decrease of the predetermined control gain value Since the configuration was configured to change the increase/decrease per cycle to a small value,
At the beginning of control, the speed of convergence toward the optimal value is fast, and as it approaches the optimal value, the degree of convergence to the optimal value improves, allowing highly accurate automatic plate thickness control to be executed, which has a significant effect on both product quality and productivity. can be obtained.

[Brief explanation of drawings]

1 to 3 show an automatic plate thickness control method for a rolling mill as an embodiment of the present invention, and FIGS. 1(a) to 3(e) are flowcharts for explaining the procedure, and FIG. 3 is a block diagram showing an apparatus to which the method of this embodiment is applied, and FIG. 3 is a graph showing the effect in comparison with the conventional method. 1-rolled material, 2-input reel, 3-output reel. 4-Rolling roll, 5-Inlet side plate thickness gauge, 6-Outlet side plate thickness gauge, 7
- pulse generator, 8- output timing circuit, 9-
Rolling mill hydraulic pressure reduction control device, 10-feedback control device, 12-control calculator, 13-numeric setting device, 14-first
Arithmetic unit, 15-second arithmetic unit, 16-memory, 18-AG
C calculation controller. Figure 1 (G) Figure 1 (b) Figure ■ (d) C]

Claims (1)

[Claims] Measure the average value of the entrance side plate thickness deviation at the entrance side of the rolling mill of the rolled material, and change the roll reduction amount at each predetermined control cycle based on the result of multiplying the average value of the input side plate thickness deviation by a predetermined control gain value. The plate thickness of the rolled material is feedforward controlled, and the average value of the exit side plate thickness deviation of the rolled material at the same position as the measurement position of the average value of the input side plate thickness deviation is measured, and the output side plate thickness is In an automatic plate thickness control method for a rolling mill equipped with an automatic gain control function that increases or decreases the predetermined control gain value based on the average value of the deviation and the average value of the entrance side plate thickness deviation, the predetermined control gain value is It is possible to change the increase/decrease per control cycle when adjusting the increase/decrease, and the increase/decrease can be changed by
An automatic plate thickness control method for a rolling mill, characterized in that the predetermined control gain value is changed to a smaller value as the predetermined control gain value approaches an optimum value.