JP2001150113A - Method for controlling molten metal surface level in continuous caster and controlling unit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a control method, in which the variation of molten metal surface level due to the non-stationary bulginess or the periodic disturbance caused by the eccentricity of pinch rolls is suppressed in a continuous caster. SOLUTION: A notch filter 21 is interposed in a molten metal surface level control system control loop and also, the control system for adding the molten metal surface level deviation into a computing output in a molten metal surface level control law 16 is constituted through a band pass-filter 22, a phase compensator 23 and a phase compensating gain 24. Frequency of the notch filter and the band pass-filter is adjusted to the pre-obtained periodic disturbance frequency. The periodical disturbance frequency is obtained from a peak component obtained by executing the frequency analysis to the variation of the molten metal surface level and may be automatically set to the notch filter and the band pass-filter. Control parameters of the notch filter, the molten metal surface level controller and the phase compensating gain may be better to be automatically set. When the frequency analisis is executed, the periodical disturbance frequency can be obtained at high speed by using a turning type frequency analyzing method having a variable frequency generator.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

TECHNICAL FIELD The present invention relates to a molten metal (hereinafter, referred to as molten metal).
A level control method and level for preventing a level change in a mold caused by unsteady bulging of a slab generated in a secondary cooling zone or eccentricity of a pinch roll in a continuous casting machine for molten metal (also referred to as a molten metal). It relates to a control device.

[0002]

2. Description of the Related Art FIG. 1 is a schematic diagram showing a molten metal level control system of a continuous casting machine. In the figure, reference numeral 1 denotes molten metal, 2
Is a tundish, 3 is an immersion nozzle, 4 is a mold, 5 is a slab, 6 is a solidified shell, 7 is an unsolidified portion, 8 is a secondary cooling zone roll, 9 is a pinch roll, and 10 is a drive for driving a pinch roll. A motor, 11 is a level gauge, 12 is a level controller, 13 is a stopper driving device for controlling the opening of the stopper according to the output of the level controller 12, and 14 is a stopper.

In FIG. 1, a molten metal 1 poured into a mold is cooled by a mold 4 to form a solidified shell 6, and a slab 5 is formed inside the unsolidified portion 7 while the solidified portion 7 is gradually solidified. While being supported by the next cooling zone roll 8, it is sequentially pulled downward.

[0004] The level control is performed as follows.
A control law (control function by proportional / integral operation) is provided so that the level of the molten metal 1 is detected by the level gauge 11 and the deviation from the set level (level deviation) becomes zero. The molten metal level controller 12 drives the stopper 14 via the stopper driving device 13 to control the inflow amount of the molten metal 1. As a result, even if disturbances such as a change in casting conditions and clogging of the immersion nozzle 3 occur, the molten metal level is maintained at the set value.

[0005] Fig. 2 is a schematic view showing the state of occurrence of unsteady bulging. Fig. 2 (a) shows the state when the slab is expanded.
(b) shows the case where the slab shrinks. In this figure, the same elements as those in FIG. 1 are represented by the same reference numerals.

As shown in FIG. 2A, the solidified shell 6 of the slab 5
When the thickness is not sufficient, the slab 5 is easily deformed, and the slab 5 expands between the secondary cooling zone rolls 8 due to the static pressure of the molten metal. If the amount of molten metal supplied to the mold is constant, the level of the molten metal drops as shown by arrow A. If the solidified shell 6 is easily deformed, the portion once expanded is pressed again by the secondary cooling zone roll 8, and the portion no longer in contact with the secondary cooling zone roll 8 expands. The expansion and contraction are repeated in a specific portion of the slab 5 as described above, but the expansion shape of the slab is constant from the specific position of the secondary cooling zone roll 8. At this time, the slab 5
Does not fluctuate, so that the level of the molten metal in the mold 4 does not fluctuate. Once dropped, the level can be restored by increasing the molten metal supply.

However, when the solidification shell 6 becomes hard while maintaining the wavy shape for some reason and the casting proceeds, as shown in FIG. Therefore, the volume of the cast slab changes, and the level of the molten metal rises as indicated by arrow B.

[0008] This is repeated, and a change in the level of the molten metal occurs with the roll interval as a cycle. This state is called unsteady bulging. When the fluctuation of the level of the unsteady bulging property becomes large, the quality of the slab is deteriorated and there is a risk of breakout. Such unsteady bulging tends to occur in steel types having a high carbon content or steel types having a high alloying component.

Separately from this, even when the pinch roll has eccentricity (bending of the roll), the unsolidified portion of the cast slab is periodically lowered and released, so that a periodic fluctuation of the molten metal level occurs. I do.

Generally, the roll intervals of the secondary cooling zone are not all the same, and the roll interval is small in the roll segment near the mold and large in the segment far from the mold. Two or more types of roll spacing segments are used. Therefore, the above-mentioned unsteady bulging level level fluctuation may include not only one type but also two or more types of frequency components.

A plurality of pinch rolls having different diameters may be used in one continuous casting machine, and not only one type but also two or more types of metal surface level fluctuation caused by the eccentricity of the pinch rolls may be used. Types of frequency components may be included.

The combination of the level change caused by the unsteady bulging and the level change caused by the eccentricity of the pinch roll is called a periodic disturbance, and its frequency is called a periodic disturbance frequency.

[0013]

In order to suppress the unsteady bulging described above, for example, Japanese Unexamined Patent Publication No. HEI 4-65742 discloses a method of making the intervals of the rolls of the secondary cooling zone non-uniform, thereby making it possible to reduce the regular periodic bulging. A technique for preventing temperature fluctuation (fluctuation in solidified shell thickness) is disclosed. However, the technique disclosed in this publication requires a large number of types of spare segments, which causes an increase in equipment costs.

Japanese Patent Application Laid-Open No. 10-314911 discloses a phase compensator having a characteristic frequency adjusted to a non-stationary bulging frequency in order to advance the phase of the level difference. Input, add the output of the phase compensator to the operation output of the fluid level controller (control command to the controller of the sliding nozzle or stopper), and compensate for the phase lag due to the integral characteristics of the mold, to provide unsteady bulging Techniques for preventing a change in the level of the molten metal are disclosed.

However, in the technique described in Japanese Patent Application Laid-Open No. 10-314911, since the fluctuation component of the periodic disturbance frequency is input to the level controller, the output of the level controller and the operation of the phase compensator are calculated. There are problems such as interference with the result and inability to cope with a case where a plurality of periodic disturbance frequencies exist.

An object of the present invention is to deal with an unsteady bulging level level fluctuation which fluctuates every moment or a periodic disturbance frequency / amplitude of a level level fluctuation caused by pinch roll eccentricity,
It is another object of the present invention to provide a method and apparatus for controlling a level of a continuous casting machine, which can effectively prevent a level change of the level even when a plurality of types of periodic disturbance frequencies are present.

[0017]

FIG. 3 is a block diagram showing the control system of FIG. In FIG.
The same elements as those shown in FIG. FIG. 1 shows a prior art level control system. In the figure, reference numeral 15 denotes a deviation calculator for calculating the difference between the set value of the molten metal level and the deviation, 16 denotes a control law unit for executing a proportional / integral operation, 17 denotes a transfer function of a stopper driving device, and 18 denotes a stopper. , 19 is the transfer function of the mold, and 20 is the transfer function of the level gauge. In addition, SP in the figure is a set level (mm) of the bath level, PV is a bath level value (mm) measured by the bath level meter, and MV is an output value (mm) of the control law unit 16.

In FIG. 3, a control law section 16 is expressed as a transfer function for performing a normal proportional-integral control (PI control).
(T) is the following equation (1). In the Laplace-transformed frequency band, the output Y (s) is
Each is expressed by the equation (2).

[0019]

(Equation 1)

[0020]

(Equation 2) Therefore, the transfer function of the level controller 12 is expressed by the following equation (3) in the frequency range.

[0021]

(Equation 3) In the following, each element of the block diagram is represented in the frequency domain.

The transfer function 17 of the stopper driving device is expressed by the following equation (4) as a first-order lag system including a dead time.

[0023]

(Equation 4) The transfer function 18 of the stopper is a system including dead time (5)
It is expressed by an equation.

F (s) = K _q e ^-sLw (5) Since the mold level of the transfer function 19 of the mold rises as a result of the integration of the opening degree of the stopper, the equation (6) is obtained. Can be represented by

Y (s) = 1 / As (6) Since the transfer function 20 of the level gauge is a second-order lag system,
This is regarded as a first-order lag + dead time system and is expressed by equation (7).

[0026]

(Equation 5) Here, K _p and T _p are proportional gains (−) of the control law unit 16.
And integration time (s), T _c and L _c are the time constant (s) and the dead time (s) of the transfer function 17 of the stopper driving device,
K _q is the gain (−) of the flow characteristic of the transfer function 18 of the stopper, L _w is the dead time (s) until the molten metal falls from the stopper to the mold, A is the cross-sectional area of the mold (mm ² ), T
_s and L _s represent the time constant (s) and the dead time (s) of the transfer function 20 of the level gauge.

In the feedback control system, since a signal related to control makes one round of the control block, a route of one round is also called a control loop. The product of the transfer functions of each control block of the control loop is called loop gain.

When the control system of the prior art as shown in FIG. 3 is used, periodic fluctuations in the molten metal level may be further increased. The reason for this is that the loop gain at a specific frequency of the feedback control system shown in the figure is larger than 1 (a signal that has made one round of the control loop becomes larger than the original signal), and the control system becomes unstable. .

FIG. 4 is an example of a graph showing the change in the level of the molten metal when the unsteady bulging occurs. As shown in the figure,
Molten metal surface level variations due to an increase in the casting speed V _c unsteady bulging is increased and smaller when lowering the casting speed V _c. In high-speed casting, local temperature non-uniformity tends to occur, and a discontinuous temperature step occurs when the speed is changed, so that unsteady bulging tends to occur.

FIG. 5 is a graph showing the control system gain of the control system shown in FIG. 3, that is, the magnitude of the change of the bath level with respect to the disturbance input as a frequency comparison. The vertical axis in the figure is the control system gain r of the control system (= amplitude of level change / amplitude of disturbance input). In the figure, the control system gain r is 1.
In the portion exceeding 0, the level change is amplified with respect to the disturbance of the frequency and superimposed on the disturbance input, so that the level change range further increases.

FIG. 6 is a graph showing a frequency spectrum of the unsteady bulging level level fluctuation shown in FIG.
In the figure, the frequency f ₂ (Hz) at which the amplitude reaches a peak (hereinafter also referred to as a peak component) is determined by the casting speed V _c (m / m
in) and the secondary cooling zone roll spacing d (mm), Required by This frequency is a periodic disturbance frequency caused by unsteady bulging.

[0032] casting speed V _c is in the continuous casting of midweight slab thickness of about 90~120mm of the slab is 3~8m / min
Reach Secondary cooling zone roll spacing is 160-250m
m and the diameter of the pinch roll is 160 to 190 mm, so that the periodic disturbance frequency is 0.1 to 0.5 H
It occurs in the band of z. Depending on the cooling conditions, a level change of a harmonic that is twice or three times the fundamental frequency rarely occurs, but only the number of corresponding frequencies increases.

As shown in FIG. 5, the control system gain r is 1.0.
If there is a portion exceeding the frequency range and the periodic disturbance frequency is included in the frequency domain, the control tends to be unstable. FIG. 7 is a graph showing an example of a change in the level of the molten metal in the control system including a plurality of periodic disturbance frequencies.

FIG. 8 is a graph showing a frequency spectrum of the fluctuation of the molten metal level shown in FIG. In the example shown in FIG.
There are three peaks of the periodic disturbance frequency, f ₁ ,
Assuming that f ₂ and f ₃ (Hz), the relations of equations (9) to (11) hold.

[0035] Here, V _c is the casting speed (m / s), R _SC is the radius of the pinch roll (mm), d ₁ is the interval (mm) between the secondary cooling zone rolls immediately below the mold, and d ₂ is the further lower part. This is the roll interval (mm) (see FIG. 1).

The fluctuation of the low frequency corresponding to the above-mentioned f ₁ is not a fluctuation of the unsteady bulging level caused by the interval of the secondary cooling zone roll, but a pinch roll (roll radius;
R _SC ) itself is caused by the eccentricity (bending of the roll) of the molten metal level.

In addition, a smaller frequency peak may appear, but what matters is the peak height, that is, the periodic disturbance frequency of the change in the level of the molten metal such that the amplitude exceeds 10 mm. In this case, the periodic disturbance frequency caused by the roll interval and the periodic disturbance frequency caused by the pinch roll eccentricity often become problems as shown in FIG. In the following description, a case where the frequency f _2, f ₃ of f ₁ and two unsteady bulging of melt surface level fluctuation of the one roll eccentricity due shown in FIG. 8 has occurred.

When the feedback control is performed by the control system as shown in FIG. 3, the fluctuation of the molten metal level is rather increased because of the periodic disturbance frequency, especially the periodic disturbance frequency f ₂ , f _This is because the control system causes resonance in step ₃ . The inventor first reduces the loop gain in the vicinity of the frequencies f ₂ and f ₃ of the unsteady bulging property level fluctuation, without impairing other control characteristics.
The method of suppressing the level change was investigated.

FIG. 9 is a block diagram showing an example of a control method for suppressing unsteady bulging. The figure shows
The control loop shown in FIG. 3 is provided with a notch filter 21 for suppressing the fluctuation of the periodic disturbance frequency, and the other elements are denoted by the same reference numerals as those in FIG.

In the figure, a control loop is constituted by a control law section 16, a transfer function 17 of a stopper driving device, a transfer function 18 of a stopper, a transfer function 19 of a mold, a transfer function 20 of a level gauge and a notch filter 21. ing. Although the notch filter 21 is inserted at any position in the control loop, the loop gain remains unchanged, but the example of FIG. 1 shows a configuration in which the notch filter 21 is inserted into the system of the bath surface level value PV.

FIG. 10 shows the notch filter 21 shown in FIG.
5 is a graph showing the filter gain of FIG. The transfer function of the notch filter 21 is represented by equation (12). Here, ω in the equation (12) represents an angular frequency, and ω = 2πf.

[0042]

(Equation 6) In FIG. 10, at the notch frequency f _n , the filter gain (also referred to as the (output / input) ratio or attenuation rate) is the lowest,
The attenuation rate at that time is g (hereinafter, referred to as a notch ratio). The band constant Q is a numerical value representing the sharpness of the valley shape in the figure, and is defined by the ratio of f _n to the frequency width Δf when the attenuation rate becomes (0.5) ^1/2 = 0.707 times. ,
The valley shape becomes narrower and sharper as the band constant Q increases.
If the notch frequency f _n of the notch filter 21 is adjusted to the periodic disturbance frequency f, even if unsteady bulging occurs, its amplitude is suppressed, so that it is possible to prevent the control system from increasing the amplitude. Even if the notch frequency f _n does not completely coincide with the periodic disturbance frequency f, the same effect can be obtained as long as the periodic disturbance frequency f is included in the range of the frequency width Δf of the notch filter.

(A) The inventor performed a simulation with a control system configuration as shown in FIG. The notch frequency of the notch filter 21 was adjusted to the periodic disturbance frequency due to the unsteady bulging, and the gain of the level controller 12 was adjusted.

FIG. 11 is a graph showing a change in the level of the molten metal when a simulation is performed by the control system shown in FIG. FIG.
4 and FIG. 4, the casting speed is 6 m / mi in FIG.
When the casting speed was increased to 6 m / m in FIG.
Although the amplitude at the time of “in” is larger than that at the casting speed of 3 m / min, there is no tendency to continuously increase and the improvement is seen in that the casting speed does not need to be reduced. However, there are still the following problems.

That is, since the periodic disturbance frequency f is cut off from the control system by the notch filter 21, the level fluctuation of the frequency component is promoted by the control and does not increase or diverge. Since the function of canceling the volume fluctuation by the hot water supply amount control is also reduced, the level fluctuation caused by the original disturbance itself remains as it is.

The present inventors have arranged this problem as follows. The control system essentially has an integral element of stopper opening → injection of molten metal into the mold → rise in the level of the molten metal.
Therefore, it is not sufficient to simply cut off the level change near the periodic disturbance frequency f from the control system by the notch filter 21.

As a countermeasure, the present inventor compensates for the phase lag of the opening control signal of the stopper for adjusting the amount of hot water in the vicinity of the frequency of the hot water level fluctuation, and cancels the volume fluctuation caused by the periodic disturbance. By doing so, it has been conceived that the fluctuation of the molten metal level can be suppressed.

To compensate for the phase delay, the component of the periodic disturbance frequency f of the level change is distinguished by a band-pass filter.
Then, it can be realized by performing a phase advance calculation process.
Note that the phase compensation element has a differential characteristic (an output is obtained in proportion to the change rate of the input variable), and the control output becomes excessive with respect to a rapid change of the cycle. It is necessary to reduce the gain outside the frequency band of the level of the molten metal. On the other hand, if the phase compensation calculation is performed after discriminating the level difference of the molten metal with the band pass filter, it is possible to prevent the control output from becoming excessive.

The phase compensation is preferably inserted in the control loop in parallel. The reason is that the loop gain of the control system for the frequency is reduced by the notch filter 21 to increase the stability, and only the frequency component is discriminated by the band-pass filter, and then phase compensation is performed. Is added to the output. Further, since a phase compensation element can be inserted in a control loop in parallel, a plurality of phase compensation elements can be inserted.

Based on the above idea, the present inventor conducted a test using a test continuous casting machine and performed various simulations, and as a result, obtained the following knowledge. (a) Unsteady bulging property Fluctuation in the level of the metal surface or fluctuation of the metal surface level caused by the eccentricity of the driving roll includes one or more frequency components.
For these frequencies, the roll spacing of the secondary cooling zone is two or three,
Also, the pinch roll diameter corresponds to one to three kinds, and the frequency is a value obtained by dividing the casting speed by the roll interval or a value obtained by dividing the casting speed by the pinch roll circumference. Usually, the frequency at which the unsteady bulging property level fluctuation occurs is 0.2 Hz or more. On the other hand, the fluctuation of the molten metal level due to the eccentricity of the pinch roll has a lower frequency of about 0.1 Hz.

(B) When there are a plurality of frequencies of the level change, the amplitude for each frequency is not always constant. A specific periodic disturbance frequency may not appear at all. This is because there is a roll position (a roll position in the secondary cooling zone) where the unsteady bulging property level change easily occurs due to a change in casting conditions.

(C) When the level change caused by the periodic disturbance does not occur, the level change of the non-periodic or long-period (frequency band lower than 0.1 Hz) is reduced by the ordinary PI controller. Can be sufficiently followed by the molten metal level control.

Unsteady bulging property In the case where a level change caused by a level change of the molten metal surface or an eccentricity of the driving roll occurs, a frequency component corresponding to the unsteady bulging is attenuated from a signal of a control system in order to prevent the fluctuation. It is sufficient to prevent the level fluctuation from increasing. To attenuate a specific frequency component from the control signal, the notch filter 21 may be interposed in the control loop.

(D) When the notch filter 21 is used, a phase delay of the control system occurs. This is because the phase is delayed in a frequency range lower than the notch frequency. However, the periodic disturbance frequency has a level fluctuation that constantly exists (nozzle clogging or fluctuation of the molten steel head of the tundish).
The frequency band is higher than the above frequency band, and there is generally no problem. If this band is close, a phase compensation operation for compensating for this may be used. That is, the level difference is input to the phase compensation calculator, the output is added to the calculated output of the level controller 16, and the result is given to the stopper driving device. In this way, the level difference is transmitted to the stopper driving device in a feedforward manner, so that the phase delay can be compensated. This phase compensation calculation unit can be realized by a combination of a band-pass filter having a response characteristic of a specific frequency, a phase compensator, and an appropriate gain constant.

Since the phase compensation calculation itself does not include a delay element, a plurality of phase compensation calculation outputs can be added. (e) Unsteady bulging level level fluctuation or level fluctuation due to eccentricity of pinch roll does not always depend on the specific roll interval and roll diameter (or roll circumference), and the magnitude of the fluctuation is not constant. The characteristic 21 (notch frequency or notch ratio), the band-pass filter characteristic (band-pass frequency), or the gain of the bath level controller 12 cannot be fixed to a constant value. As a countermeasure, if the frequency analysis is performed by constantly observing the change in the level of the molten metal, the frequency and amplitude of the peak component can be constantly grasped, and the change state of the unsteady bulging can be captured. If the characteristic parameters of the bandpass filter are automatically set, it is possible to cope with a process change that changes every moment.

(F) To perform the above-mentioned frequency analysis at a high speed, a fast Fourier transform (FFT) is used.
By using the method of n), online calculation is possible. However, the FFT method must sample the fluctuation of the surface level over a period of 50 seconds or more. Even if the FFT operation itself is fast, the frequency analysis result is found at least 50 seconds after the start of the measurement. Therefore, when the casting speed changes and the periodic disturbance frequency changes at the timing until the end of the sampling section, or when a new periodic disturbance frequency occurs, the frequency before the casting speed or the like changes for automatic setting is used for automatic setting. The frequency after the change becomes available only after the end of the next sampling interval, so that the responsiveness is not always sufficient.

As a countermeasure against this, if a variable frequency oscillator which is always tuned to the periodic disturbance frequency is provided and the periodic disturbance frequency is obtained from the oscillation frequency, the periodic disturbance frequency can be obtained at a higher speed. It is possible to quickly respond to a change in casting speed.

The present invention has been completed based on the above findings, and the gist thereof is as follows. (1) In the level control method of a continuous casting machine using a level controller, the periodic disturbance frequency of the level change caused by unsteady bulging or eccentricity of the pinch roll is determined. During the control loop of the system,
A variable component of a specific notch frequency is selectively attenuated, and a notch filter adjusted such that the notch frequency includes the periodic disturbance frequency is interposed. A band-pass filter that transmits and adjusts the band-pass frequency to include the periodic disturbance frequency; a phase compensator whose phase compensation frequency is adjusted to include the periodic disturbance frequency; A phase compensation gain unit that is multiplied and output to form a phase compensation calculation unit,
And a method for inputting a level difference to the phase compensation calculation section, and adding an output of the phase compensation calculation section to a calculation output of the level control section, the level control method for the continuous casting machine.

(2) The frequency of the peak component is determined by performing a frequency analysis of the change in the level of the molten metal, and the frequency of the peak component is calculated from among these frequencies.
The level control method for a continuous casting machine according to the above mode (1), wherein the periodic disturbance frequency is obtained by selecting a frequency of 1 Hz or more.

(3) The amplitude of the peak component obtained by the frequency analysis is obtained, the notch ratio of the notch filter is set according to the amplitude, and the frequency of the peak component obtained by the frequency analysis is set to 0. The method for controlling a level of a continuous casting machine according to the above mode (2), wherein the control gain of the level controller is set according to the lowest frequency of 1 Hz or more.

(4) The phase compensation gain of the phase compensation calculation section is increased or decreased by a small amount, a predetermined time control is performed, and then the variation of the output value of the band-pass filter is measured. The continuous casting machine according to the above (2) or (3), wherein a new value of the phase compensation gain is set based on the relationship of the direction of increase and decrease of the fluctuation amount so that the fluctuation amount decreases. Level control method.

(5) The notch filter and the phase compensation operation unit are replaced with a composite notch filter and a composite phase compensation operation unit, respectively, and the composite notch filter is provided for each of a plurality of periodic disturbance frequencies. A plurality of notch filters which are adjusted so that the notch frequencies include the respective periodic disturbance frequencies and which are connected in series, and wherein the composite phase compensation operation unit comprises a plurality of phase compensation operation units and at least one composite phase compensation operation unit A plurality of phase adders, and the composite phase compensation operation unit receives the level difference from each of the plurality of phase compensation operation units, and outputs the output of each of the plurality of phase compensation operation units to the composite phase compensation operation unit. The result added by the arithmetic unit adder is configured to be added to the arithmetic output of the level controller, and each of the plurality of phase compensation arithmetic units is The bandpass filter, the phase compensator, and the phase compensator are configured such that each bandpass frequency and phase compensation frequency correspond to each periodic disturbance frequency. The method according to any one of the above (1) to (4), wherein the molten metal level is adjusted so as to include the molten metal level.

(6) When the frequency of the peak component is determined by performing the frequency analysis of the level change, the signal of the level change and the signal of the variable frequency oscillator are multiplied in real time,
The oscillation frequency of the variable frequency oscillator is changed based on a difference signal between the level difference obtained from the multiplication result and the respective frequencies of the variable frequency oscillator, and the oscillation frequency is changed to a periodic disturbance frequency of the level change. The method for controlling a molten metal level of a continuous casting machine according to the above mode (2), wherein tuning is performed and a periodic disturbance frequency is obtained from the oscillation frequency at that time.

(7) A phase plane trajectory shape judging device is interposed in the control loop, and a signal of a periodic level change of the molten metal level and a hot water supply mechanism mounted on a tundish for supplying molten steel into the mold are provided. An ellipse obtained by inputting the opening degree command to the phase plane trajectory shape determiner and plotting the fluctuation amount of the signal of the water level fluctuation and the fluctuation amount of the opening degree command of the hot water supply mechanism on the same plane. The phase advance amount by the phase compensator and / or the phase compensation gain by the phase compensation gain unit are corrected and set so that the roundness of the shape of the phase plane trajectory is improved and the area is reduced. The method for controlling a molten metal level of a continuous casting machine according to the above mode (1) or (6), wherein

(8) The control system selectively attenuates the fluctuation component of the specific notch frequency with the level sensor, the level controller, and the notch frequency is the periodic disturbance of the level. A notch filter adjusted to include a frequency, a band-pass filter selectively transmitting a variable component of a specific band-pass frequency, and adjusting the band-pass frequency to include the periodic disturbance frequency; A phase compensator whose compensation frequency is adjusted to include the periodic disturbance frequency, and a phase compensation gain unit that multiplies an input signal by a phase compensation gain by the phase compensator and outputs the result are connected in series. A notch filter that receives a level signal from the level sensor and sends a level change signal to the level controller. The band-pass filter has a function of receiving a level signal from the level sensor and transmitting a band-pass frequency signal adjusted to include a periodic disturbance frequency to the phase compensator. The phase compensator has a function of sending a phase compensation frequency signal adjusted to include a periodic disturbance frequency to a phase compensation gain unit, and further, the phase compensation calculation unit A molten metal level control device for a continuous casting machine having a function of adding to a calculation output of a surface level controller.

(9) The control system has a frequency detection calculation unit for detecting the periodic disturbance frequency.
(8) The molten metal level control device of the continuous casting machine according to the item (8). (10) The frequency detection calculation unit includes a variable frequency oscillator, and multiplies a signal of the periodic level change output from the level sensor and a signal oscillated from the variable frequency oscillator in real time. Then, based on the level difference obtained from the multiplication result and the difference signal between the frequencies of the variable frequency oscillators, the oscillation frequency of the variable frequency oscillator is changed, and the oscillation frequency is changed to the period of the periodic level change. The level control device for a continuous casting machine according to the above mode (9), further comprising a function of tuning to a periodic disturbance frequency and obtaining a periodic disturbance frequency of a periodic level change from the tuned oscillation frequency.

(11) The control system has a phase plane trajectory shape determiner for adjusting the output of the phase compensation calculation unit to be added to the calculation output of the level controller, An elliptical phase surface trajectory obtained by plotting on the same plane the amount of fluctuation of the signal of the periodic level level fluctuation input to the determiner and the amount of fluctuation of the opening command of the hot water supply mechanism of the continuous casting machine. A function to correct and set the amount of phase advance by the phase compensator and / or the phase compensation gain by the phase compensation gain unit so that the roundness of the shape is improved and the area is reduced. The level control device for a continuous casting machine according to any one of items (8) to (10), characterized in that:

[0068]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 12 is a block diagram showing a control method according to the present invention. 9, the same elements as those in FIG. 9 are denoted by the same reference numerals. Reference numeral 22 is a band pass filter of the figure, 23 is a phase compensator, 24 is a phase compensation gain unit having a phase compensation gain K _g. In the figure, the band-pass filter 22, the phase compensator 23 and the phase compensation gain unit 24
Are collectively enclosed by a dotted line, and are collectively referred to as a phase compensation calculation unit 25.

The level compensation deviation is input to the phase compensation calculation unit 25, and the output of the calculation result is added to the output of the control law unit 16 by the output addition unit 26.
7 is given a command value.

In the figure, a notch filter 21 includes a deviation calculating section 15 for calculating a level difference of the control system and a control law section 16.
As described above, the effect of the notch filter 21 is the same regardless of where it is inserted in the control loop, and as shown in FIG. This is equivalent to the case where it is inserted between the calculation units 15.

FIG. 13 is a graph showing the relationship between the frequency of the bandpass filter 22 and its gain (transmittance). At the band pass frequency f _b , the transmittance becomes maximum (the transmittance value at this time is referred to as a band pass ratio: h).

In the present invention, the band pass frequency f _b
Is adjusted to the unsteady disturbance frequency f. The transfer function of the bandpass filter 22 is given by the following equation (13). However, (13)
In the equation, ω = 2πf _b , and when the bandpass frequency is adjusted to the periodic disturbance frequency f, ω = 2πf.

[0074]

(Equation 7) FIG. 14 is a graph showing the relationship between the input and output of the phase compensator 23. The phase of the output of the phase compensator 23 advances by 90 ° with respect to the input signal as shown. That is, the phase compensation corresponds to performing a differential operation. The transfer function of the phase compensator 23 is represented by the following equation (14). Here, ω (phase compensator frequency) in equation (14) is the frequency ω of the bandpass filter 22.
_b or the same value as the unsteady disturbance frequency ω.

[0075]

(Equation 8) The phase compensation gain section 24 is a section for adjusting the amplitude of the signal passing through the band pass filter 22 and the phase compensator 23. That is, the phase compensation gain of the input signal: K _g multiplies.

As shown in FIG. 12, the bandpass filter 22,
The phase compensator 23 and the phase compensation gain unit 24 are connected in series.
Subsequently, when the phase compensation calculation unit 25 is configured, a specific frequency
Number f _bThe phase can be advanced only for. Soshi
The phase compensation calculation unit 25 is controlled by the level control system.
Along with the control law part 16 and the notch filter 21 in the loop,
The interposition of the notch in the control loop
The insertion of the filter 21 causes a disturbance that is constantly present.
Non-periodic or long-periodic (lower than 0.1 Hz
Band) The control sensitivity around the frequency of the level change decreases.
Problems can be solved. Also, a phase compensation calculation unit
25 is near the periodic disturbance frequency because the phase is advanced by 90 °.
If the control becomes stable without promoting the disturbance amplitude nearby,
This has the effect.

FIG. 15 is a graph showing a disturbance suppression simulation by the control system shown in FIG. The figure shows a case where a volume fluctuation of the molten metal corresponding to a frequency of 0.25 Hz and an amplitude of ± 10 mm is applied as disturbance due to unsteady bulging. As shown in the figure, although the volume fluctuation disturbance is not directly changed to the molten metal level level by the control but is suppressed, the amplitude of the fluctuation is ± 5 mm.

FIG. 16 is a graph showing a simulation result of disturbance suppression by the control system of the present invention shown in FIG. The conditions of the disturbance frequency and the disturbance amplitude are the same as those in FIG. As shown in FIG.
mm, and the amplitude is halved compared to FIG.

The periodic disturbance frequency changes depending on the casting speed. Further, as described above, there are several types of roll intervals in the secondary cooling zone, and it is not possible to specify a roll interval type that induces unsteady bulging depending on casting conditions. Therefore, if the notch frequency and bandpass frequency of the control system are kept constant, there is a problem that it is impossible to follow a change in casting conditions. As a countermeasure, the control method of the present invention provides a method of detecting a periodic disturbance frequency and automatically adjusting a notch frequency and a bandpass frequency.

FIG. 17 is a block diagram showing a control method for automatically adjusting the notch frequency and the band pass frequency according to the present invention. 12, the same elements as those in FIG. 12 are denoted by the same reference numerals.

FIG. 14 shows a block corresponding to the portion (notch filter 21, control law section 16 and phase compensation calculation section 25) surrounded by the two-dot chain line in FIG. 12, and shows the transfer function 17 of the stopper driving device and the transfer of the stopper. The function 18, the transfer function 19 of the mold, and the transfer function 20 of the level gauge are omitted. The frequency analysis unit 27 shown in the figure is a device that performs a frequency analysis of a change in the level of the molten metal and detects an amplitude for each frequency, and a suitable example is a fast Fourier transform device or FFT calculation software.

In addition, the present invention provides a method of frequency analysis using a variable frequency oscillator, which will be described in detail later. Hereinafter, an example in which the FFT is used as the frequency analysis unit 27 will be described.

The frequency analysis unit 27 detects the peak frequency of the change in the level of the molten metal level, regards the peak frequency as a periodic disturbance frequency, and automatically sets the notch frequency of the notch filter 21 and the bandpass frequency of the bandpass filter 22. In FIG. 17, a dotted arrow from the frequency analysis unit to the notch filter 21 and the band-pass filter 22 indicates automatic setting of the frequency.

The periodic disturbance frequency is 0.1 to 0.5 Hz
In general, the above-described automatic setting of the frequencies of the notch filter 21 and the band-pass filter 22 targets only the peak frequency of 0.1 Hz or more. Even if there is a component of the frequency 0 Hz corresponding to the average value of the level of the molten metal,
The frequency analysis can be calculated with double precision, or the deviation of the level change can be ignored by performing the frequency analysis calculation. FIG. 17 shows a case in which the level difference is used as an input for frequency analysis.

By automatically setting the frequencies of the notch filter 21 and the band-pass filter 22 as described above, the characteristics of the control system can be automatically adjusted to follow changes in casting conditions. Level fluctuation can be suppressed stably.

By the control method described above, it is possible to suppress the unsteady bulging level change or the level change caused by the pinch roll eccentricity. As in the prior art, the suppression of the constantly existing non-periodic or long-period (lower than 0.1 Hz) level control of the liquid level is accurately performed by the control by the proportional integral (PI) operation of the level controller. Done.

[0087] To improve the control accuracy and response speed, it is desirable to increase the control gain K _p of the molten metal surface level control 12. However, if the control gain _Kp is too large, there is a problem that the fluctuation with respect to disturbance increases. The value of an appropriate control gain K _p is varied by casting conditions. Thus, the control gain K _p may be desirable to automatically set. The control when the gain K _p is varied, the phase compensation gain K _g notches ratio g and the phase compensation calculation unit 25 of the notch filter 21 so that the entire control system is harmony to adjust is desirable.

The present invention also provides a method for automatically setting the gain of these control elements, which will be described below. The targets for automatically setting various gains according to the present invention are the control gain K _p , the notch ratio g, and the phase compensation gain K _g of the fluid level controller 12.

FIG. 18 is a block diagram showing a method for automatically setting various gains in the control method of the present invention. In the figure, the same elements as those in FIG. 17 are denoted by the same reference numerals. Also, as in FIG. 17, only the portion surrounded by the two-dot chain line frame in FIG. 12 is shown.

Reference numeral 28 in FIG.
Is a control gain setting unit, and 30 is a phase compensation gain setting unit. In the figure, a notch ratio setting unit 28 controls the notch filter 21 in accordance with the amplitude (peak height) of the fluctuation of the level of the periodic disturbance frequency obtained by the frequency analysis unit 27.
Is set. Control gain setting unit 28 sets the control gain K _p in accordance with the periodic disturbance frequency obtained by the frequency analyzing unit 27. The phase compensation gain setting section 30 sets the phase compensation gain K _g while observing the output of the band-pass filter 22. In FIG. 18, g and K _p
Frequency setting unit 27 → notch ratio setting unit 2
8 → dashed line toward notch filter 21, frequency analysis unit 27
→ A dashed line toward the control gain setting unit 29 → the control law unit 16 and a dashed line from the output of the bandpass filter 22 → the phase compensation gain setting unit 30 → the phase compensation gain unit 24. The details of the automatic setting method will be described below.

Setting method of notch ratio g: Notch ratio is set to 1 in order to cut off disturbance of the notch frequency (adjusted to the periodic disturbance frequency in the present invention) mixed in the control loop.
There is no effect unless it is made smaller. However, if the notch ratio is too small, a phase delay occurs at a frequency lower than the notch frequency, and the level control of the molten metal becomes unstable. Therefore, when the level change caused by the periodic disturbance is large, the attenuation by the notch filter 21 is increased (the notch ratio g is reduced), and when the level change is small, the attenuation is reduced (g is increased). Alternatively, it is better not to attenuate at all (g = 1).

FIG. 19 is a graph showing an example of a method for obtaining the notch ratio g of the notch filter 21. In the figure, when the level fluctuation of the periodic disturbance frequency is large (in the example of the figure, the amplitude exceeds 2 mm), the notch ratio g is small (0.2), and the level fluctuation is small (see the figure). When the amplitude is less than 1 mm in the example), the notch ratio is increased (1.0, that is, not attenuated). The reason why the notch ratio is changed in the form of a slope in a section where the molten metal level fluctuation amplitude is 1 to 2 mm is to avoid a sudden change.

[0093] melt-surface level controller 12 controls the gain K _p of the setting method: the adjustment K _p in the peak component obtained from the frequency analysis result of the molten metal surface level fluctuation, attention is paid to the variation of relatively low frequency .

The normal unsteady bulging level level fluctuation has a frequency of 0.2 Hz or more, and that caused by pinch roll eccentricity has a frequency of about 0.1 Hz. Frequency components below that are non-periodic or long-period (frequency lower than 0.1 Hz) such as clogged nozzles that are constantly present or fluctuations in the height of the molten steel head in the tundish.
Level fluctuation. As described above, when there is a peak component at 0.1 Hz or less, if the fluctuation caused by the eccentricity of the pinch roll is to be cut off from the control system by the notch filter 21, the phase delay caused by the notch filter 21 greatly affects the control loop. It works to increase the level fluctuation of the molten metal and make it unstable. Therefore, in this case, the control gain _Kp is reduced (at the expense of some response speed) to prevent the control from becoming unstable.

In the case where there is no level fluctuation around 0.1 Hz due to the eccentricity of the pinch roll and only the suppression of the fluctuation of the non-stationary bulging level level is set, the notch frequency of the notch filter 21 is 0.2 Hz. This problem does not occur in the above high band since the influence of the steady disturbance at a lower frequency on the band is reduced.

[0096] Figure 20 is a graph showing a correction coefficient setting example of the control gain K _p of the molten metal surface level control 12 with respect to the frequency of the molten metal surface level fluctuation of the present invention. In the example of the figure, in case the lowest frequency of the molten metal surface level fluctuation is less than 0.1 Hz, and the presence of the level fluctuation in the vicinity of 0.1 Hz, reduced K _p When inserting the notch filter 21 to the control system and shows an example that remain control gain based on the K _p when 0.2Hz larger. Between 0.1 and 0.2 Hz, the correction coefficient is changed in a slope shape to avoid a sudden change. Here, the reference control gain is a control gain of the metal level controller adjusted with a steel type such as low carbon steel in which unsteady bulging hardly occurs.

Method for setting K _{g of} phase compensation gain: As described above, the phase compensation calculation unit 25 performs a differential calculation. The differential operation is effective in compensating for a phase lag because the control in the suppression direction is performed “in advance” so as not to increase the fluctuation by performing “look-ahead” of the fluctuation of the bath surface level. However, when the disturbance signal has a small variation in high frequency, the action of suppression is increased by the differential operation, and the variation is rather increased. Such high-frequency fluctuations vary depending on the characteristics and configuration of each device in an actual process, process parameters unique to a continuous casting machine, and the like, and thus it is difficult to automatically set them by using certain conditional expressions. In the present invention, the control is executed by increasing or decreasing the phase compensation gain of the phase compensation calculation unit by a very small amount, and as a result, it is observed whether or not the fluctuation of the surface level of the frequency increases or decreases. A method of searching for an optimum value by resetting the phase compensation gain is used. As an example, a method of obtaining the phase compensation gain of the phase compensation calculation unit 25 by trial and error as described below will be described.

The initial value of the phase compensation gain K _g is set in advance in the phase compensation calculating section 25, and the phase compensation gain K _g is set.
The level of _g is increased / decreased by a small amount to execute the level control, and it is evaluated whether the level change (amplitude of level level deviation e) has decreased or increased during the level control. Phase compensation gain K _g
Is increased or decreased as a result, if the molten metal level fluctuation increases, it means that K _g has been increased or decreased in the wrong direction. Therefore, K _g is increased or decreased in the opposite direction to the previous increase or decrease operation. As a result of increasing or decreasing the phase compensation gain K _g , if the fluctuation of the molten metal level decreases, it means that the adjustment direction of K _g has increased or decreased in the correct direction, and further increased or decreased in the same direction to obtain a more optimal value of K _g . Increase further to look for.

Such an operation is repeated a finite number of times, and the optimum K _g , that is, the K _g that minimizes the change in the level of the molten metal, is selected from among them and set as a new K _g . Or,
This operation may be always performed every time, and the optimum _Kg may be always maintained.

It is preferable that the magnitude comparison of the fluctuations in the level of the molten metal is made by comparing and evaluating by means of the root mean square, which has a simple arithmetic processing. In order to apply to an actual process, it is desirable to take an integral multiple of the fundamental period T ((sec); reciprocal of the band-pass frequency = 1 / f _b ) as one trial time. In addition, in order to eliminate the influence of the transient state, it is desirable to start the root mean square calculation at least one cycle T (second) after changing K _g for each trial. Since the optimization of the phase compensation gain is related to the phase compensation calculation unit 25, it is preferable that the level of the bath surface level be determined by calculating a root mean square only for a specific frequency, that is, a component of a bandpass frequency (also a phase compensator frequency). . Therefore, in the present invention, the observation of the root mean square of the fluctuation of the level of the molten metal is performed by the bandpass filter 2.
Seeking a second output value e _b.

As a method of searching for an optimum value while changing K _g by a small amount, for example, the following adaptive learning method is suitable. FIG. 21 is a flowchart showing an example of a method for setting the phase compensation gain K _{g according to} the present invention. In FIG.
In step S1, initialization is performed. In step S2, the root mean square of the change in the bath surface level obtained during the past cycle is obtained. In step S3, the determination is performed. In step S4 or step S6, the root mean W if _n is greater than the previous value W _n-1 of the molten metal surface level variation (greater than the error range epsilon) continues again controlled to increase the minute volume values of K _n. Conversely, W
If _n is the previous value W _n-1 is smaller than the value of K _n reducing small amounts. Step S5 is the case of unwanted change have been set proper K _g is. By repeatedly performing this, the optimum K _g can be always maintained.

As described above, the unsteady bulging level change or the level change caused by pinch roll eccentricity may include not only a single frequency component but also a plurality of frequency components. To address this,
It is desirable that the notch filter 21 interposed in the control loop has a characteristic of attenuating a plurality of types of frequencies.
This means that a plurality of notch filters 21 having one notch frequency are connected in series, and the whole
It can be realized by assuming that it is 1, that is, by constructing a composite notch filter, replacing it with a conventional notch filter, and inserting it into a control loop.

Similarly, for the phase compensation calculation unit 25,
In order to correspond to a plurality of frequencies, a plurality of phase compensation calculation units 25 having one bandpass frequency are connected in parallel, and the whole is regarded as one phase compensation calculation unit, that is, a composite phase compensation calculation unit is configured. This can be realized by replacing the conventional phase compensation calculation unit.

FIG. 22 is a block diagram showing a control system corresponding to a plurality of periodic disturbance frequencies according to the present invention. 12, the same elements as those in FIGS. 12, 17, and 18 are denoted by the same reference numerals, and when there are a plurality of the same elements, the branch numbers are assigned. In addition, as in FIGS. 17 and 18, only a portion surrounded by a two-dot chain line frame in FIG. 12 is shown. In the figure, composite notch filters are notch filters 21-1, 21-2 and 21-.
3 is formed by connecting three notch filters in series. The composite phase compensation operation unit 32 includes three phase compensation operation units 25-1, 25-2, and 25-3 and a composite phase compensation operation unit adder 33. The level difference is input to the three phase compensation calculation units 25-1, 25-2, and 25-3, and their outputs are added together by the composite phase compensation calculation unit adder 33. The compensation operation units 25-1, 25-2, and 25-3 are connected in parallel. The phase compensation calculation unit 25-1 includes a band-pass filter 22-1, a phase compensator 23-1, and a phase compensation gain unit 24-1. Phase compensation calculation unit 25-
The same applies to 2, 25-3. Further, the result added by the composite phase compensation calculation unit adder 33 is added to the output of the control law unit 16 by the output adder 26 to become a control signal to the stopper driving device.

The notch frequency of the notch filter 21-1 is set to one periodic disturbance frequency f ₁ by the automatic frequency setting function of the frequency analysis unit 27 described above, and the band pass frequency of the band pass filter 22-1 is also set. It is set to the same repetitive disturbance frequency f _1. Similarly, notch filters 21-2 and 21-3 and bandpass filter 22
-2, 22-3 are other periodic disturbance frequencies f ₂ ,
each of which is set to f _3. In FIG. 22, these automatic setting paths are indicated by dotted lines.

[0106] In FIG. 22 further previously mentioned notches ratio g and the phase compensation gain K _g automatic setting function of each of the notch filters 21-1,21-2,22-3 and each of the phase compensation gain section 24 -1, 24-2, 24-
3 is performed. These automatically set paths are indicated by broken lines. However, the blocks corresponding to the notch ratio setting unit 28 and the phase compensation gain setting unit 30 shown in FIG. 18 are omitted, and the notch filter and the phase compensation gain unit are set directly from the frequency analysis unit 27. Is shown.

Next, another mode of the frequency analysis unit 27 in the method of automatically setting the periodic disturbance frequency to the notch filter 21 and the band-pass filter 22 in FIG. 17 will be described.

As described above, when the frequency analysis in the level control of the continuous casting machine is performed by FFT, the periodic disturbance frequency to be detected is 0.1 to 0.5 Hz. Among these, the periodic disturbance frequency due to the roll interval in the secondary cooling zone is 0.2 to 0.5 Hz. The difference in the roll interval (distance) of the secondary cooling zone is 10 to 15%. Therefore, the resolution of the frequency analysis needs to secure about 0.02 Hz, and the number of samples for the FFT analysis is 2 ⁹ (= 51
2) It is above. Generally, the control sampling period is set to 0.
This is on the order of one second, so that the minimum time required for sampling is 51.2 seconds.

On the other hand, in the continuous casting machine, the casting speed is increased or decreased after starting or ending the casting, or the quality is checked.
Increasing / decreasing casting speed for timing adjustment.
As the speed increases and decreases, there is a change in the position of the final solidification point (crater end), which changes the state of unsteady bulging, that is, changes in the type of roll interval that causes unsteady bulging. Disturbance frequencies may appear suddenly. In response to the change in the type of the periodic disturbance frequency and the change in the frequency, the FFT method starts sampling from the point in time when the change is completed, the frequency analysis result is obtained, and the setting of the control system parameter of the present invention is not performed. It is performed after 50 seconds or more have elapsed after the end of the change of the frequency or the like. It is desirable to shorten this time from the viewpoint of improving quality / yield and reducing breakout.

Another aspect of the frequency analysis unit 27 provided by the present invention is a method of obtaining the frequency while tuning to the periodic disturbance frequency of the level change by the internal variable frequency oscillator. In the following description, this method is referred to as tuning frequency analysis (PLL; also abbreviated as Phase Lock Loop).

FIG. 23 is a block diagram showing a frequency analysis method according to the tuning type frequency analysis method of the present invention.
In the figure, reference numeral 34 denotes a variable frequency oscillator, 35 denotes a multiplier, 36 denotes a low-pass filter, and 37 denotes a frequency measuring device. The combination of these elements is the frequency analysis unit 27
Is composed.

As in FIG. 17, a level signal or a level deviation signal (fluctuation level fluctuation) including the periodic disturbance frequency is input to the frequency analysis unit 27, but inside the frequency analysis unit 27. It is input to the multiplier 35. On the other hand, a sine wave is input from the variable frequency oscillator 34 to the multiplier 35, and the output of the multiplication result is once passed through the low-pass filter 36, so that a “beat” corresponding to the difference in the bath surface level and the frequency difference of the variable frequency oscillator is obtained. Is extracted.
The frequency of the variable frequency oscillator 34 is changed according to the value of the “beat (frequency difference signal)”.

The frequency measuring device 37 is a variable frequency oscillator 34
And the time from when the output value becomes zero (zero cross) to the next time when the output value becomes zero is T / 2 (T
Is a cycle (second)), and f = 1 / T is a frequency. This frequency is the oscillation frequency of the variable frequency oscillator 34.

As shown in FIG. 23, by configuring the level control system using the frequency analysis unit 27 based on the PLL method, the periodic disturbance frequency and the frequency of the variable frequency oscillator 34 always coincide (tune).

The above tuning principle can be mathematically explained as follows. Assuming that a component of a specific periodic disturbance frequency ω _i (expressed as an angular frequency) included in the signal e (t) of the fluid level fluctuation is v _i , and a signal of the variable frequency oscillator 34 having the frequency ω _p is v _p , Here, φ _p and φ _i are the phases of the respective signals at time 0.

Normally, ω _p ≠ ω _i , but when the frequency of the variable frequency oscillator 34 is continuously changed, the frequency difference can be replaced with a phase difference. That is, if the angular frequency ω _i in Expression (16) is replaced with ω _p and the phase itself is considered to change with time, Expressions (15) and (16) are Here, θ _p (t) = φ _p and ω _i t + φ _i = ω _p t
+ Θ _i (t). That is, θ _p (t) and θ _i (t)
Can be regarded as a periodic disturbance frequency and a phase change applied to the variable frequency oscillator 34.

The output of the multiplier 35 is the product of the equations (17) and (18). It is expressed as

In FIG. 23, when the result of the multiplication is passed through a low-pass filter 36, the second term on the right side of the equation (19) is deleted, and only the first term remains. If this is v _d (t), Becomes Equation (20) corresponds to the “beat” of the output of the multiplier 35 or a frequency difference signal.

In the present invention, the frequency of the variable frequency oscillator 34 is changed in accordance with v _d (t) expressed by the equation (20). As described above, a change in the angular frequency is regarded as a phase change (phase shift). Is defined by changing the speed of the phase change θ _p (t) so as to be proportional to v _d (t),

[0120]

(Equation 9) Can be expressed as

Or

[0122]

(Equation 10) Can also be expressed as Here, K ₀ is a gain for changing the frequency of the variable frequency oscillator 34.

From the expression (22), the expression (17) is

[0124]

[Equation 11] It is expressed as The integral term in {} of the equation (22) represents the cumulative total of the phase shift of the variable frequency oscillator 34 performed from time 0 to the current time t by feeding back the “beat” to the variable frequency oscillator 34. and that although, as described above, is equivalent to varying the angular frequency omega _p.

Here, {θ _p (t) −θ in equation (20)
Confirm the final value of _i (t) + π / 2｝ or ω _p . Then, from equations (21) and (20),

[0126]

(Equation 12) However, it is _{K L = 1/2 · K} 0 A p A i.

Here, the periodic disturbance frequency at t = 0
Δω = ω_i−ω _pThen, equation (15)
~ From equation (18),It can be expressed as.

Alternatively,

[0129]

(Equation 13) It can be expressed as.

From equations (25) and (27),

[0131]

[Equation 14] Is derived.

FIG. 24 is a phase diagram related to Θ (t) for examining the solution of the differential equation of equation (28) according to the PLL method of the present invention. In the figure, if | Δω | <K _L , the graph of the phase state intersects the horizontal axis, so that −π / 2 <
DΘ (t) / dt = 0 in the range of Θ (t) <π / 2, that is, Θ (t)
= There is a solution that satisfies a certain condition.

In the figure, Θ (t) can approximate a sine curve by a straight line near Θ (t) = 0, so that equation (28) is

[0134]

(Equation 15) It can be expressed as.

Therefore, as an approximate solution of the differential equation with respect to Θ (t) in equation (29),

[0136]

(Equation 16) Is obtained.

Equation (30) shows that when t = 、,

[0138]

[Equation 17] Becomes

The expression (30) or (31) indicates that the phase difference Θ (t) is
= ∞ indicates that it converges to Δω / K _L , where v _d (t) in equation (20) is = は · A _p A _i sin ｛Δω /
It stabilizes at a constant value of K _L ｝.

Further, v _p (t) in the equation (17) is obtained at t = ∞. Assuming that the oscillation frequency ω _p of the variable frequency oscillator 34 is ω _i
It converges to. This state is a tuning state.

FIG. 25 is a graph showing a simulation result of a state where the oscillation frequency of the variable frequency oscillator 34 is tuned to a signal having a periodic disturbance frequency. The figure in block diagram in Figure 23, the input corresponds to the periodic disturbance frequency _{v i (t) = sin (} ω i t), the output v _p (t) = sin a variable frequency oscillator 34 (ω _p t), Multiplier 3
5 shows the output of the low-pass filter 36 and the output of the low-pass filter 36 with time. In the figure, at time 0, v phase difference between the _i and v _p rather shows a state in which there is a difference in frequency (f _p = ω _p /2π=0.3Hz: variable, f _i = ω _i
/2π=0.33 Hz: constant).

[0142] In this figure, v _i is the period of time 0-5 seconds, the output v _d of the low-pass filter 36 after passing gradually increases, v phase difference of _p and v _i is increasing (omega _p of phase gradually Is late).

Next, from time 5 to 15 seconds, the output v _d after passing through the low-pass filter 36 further increases, ω _p gradually increases, and the expansion of the phase difference decreases. Finally, the time 15 seconds after the output v _d of the low-pass filter 36 after passing through almost constant value, omega _p is equal to approximately omega _i, the phase difference is held constant. This state is a synchronized state.

As described above, the tuning type frequency analysis (P
Since the tuning can be detected in about 15 to 20 seconds by using the LL) method, the periodic disturbance frequency can be obtained in a shorter time as compared with the frequency analysis of the FFT method.

As shown in the graph of FIG.
L method is characterized in that the time required for tuning as the difference between the oscillation frequency f _p and periodic disturbance frequency f _i of the variable frequency oscillator 34 at time 0 time is small is shortened. That is, when the periodic disturbance frequency f _i occurs suddenly (step change), if given the oscillation frequency close to the periodic disturbance frequency f _i that is expected in advance to the variable frequency oscillator 34 requires to tune Time can be reduced.

[0146] Further, PLL method of the present invention for seeking always continuously tuning frequency has Seiritsu' relationship always f _p ≒ f _i. Therefore, even if the repetitive disturbance frequency f _i gradually increases or gradually decreases (ramp-like change), the oscillation frequency f
_{Since p} successively follows f _i , the periodic disturbance frequency can be obtained with almost no time delay, and the responsiveness is high.

However, in the frequency analysis by the PLL method of the present invention, amplitude information of the periodic disturbance cannot be obtained. Therefore, a method for setting the control parameters based on the amplitude information, for example, as described in the summary (3) of the present invention. The notch ratio of the notch filter 21 cannot be automatically set. Therefore, immediately after the start of casting when the amplitude of the level change greatly changes, or when the casting speed is rapidly increased or decreased, the use of the PLL method is avoided, and the automatic setting of the periodic disturbance frequency and control gain by the FFT method is performed. Corresponding, or only the automatic setting of the periodic disturbance frequency is immediately supported by the PLL method,
It is preferable to use a control method that takes advantage of each feature, such as performing amplitude detection by FFT in parallel and automatically setting the notch ratio and the like.

(Second Embodiment) Next, a second embodiment of the present invention will be described.
An embodiment will be described. In the following description, only portions different from the first embodiment will be described, and common portions will be denoted by the same reference numerals in the drawings, and redundant description will be omitted as appropriate.

FIG. 26 is a view showing the configuration of the above-mentioned Japanese Patent Application Laid-Open No. 10-3149.
It is explanatory drawing which shows typically the technique disclosed by Japanese Patent Publication No. 11-No. As shown in the figure, in this technique, a phase compensator 23 is controlled by a control law unit in order to advance a phase of a level fluctuation (a level signal) having a periodicity in the level controller 12 to compensate for the delay. 16 and the control gain output from the phase compensator 23 is multiplied by a control gain to multiply the output of the advance element by a control gain to adjust the opening degree of the hot water supply mechanism.

FIG. 27 is an explanatory diagram schematically showing the first embodiment described above. As shown in the figure, in the first embodiment, a level notch filter 21 for cutting off a signal having a frequency including a periodic disturbance frequency is added to the control loop of the level controller 12 so as to increase the level. The fluctuation is prevented from increasing.

By the way, in the first embodiment, the control signal for determining the opening degree of the hot water supply mechanism (adder 2 in FIG. 26)
Though the determination and adjustment of the gain of the output signal from the control signal 6 are considered, the phase of the control signal is fixed.

However, if the gain of the phase compensator 23 is changed to increase or the phase is changed, the control characteristics become unstable in the high frequency range. The control may become unstable due to the ripples.

The term “undulation” as used herein refers to a change in the level of the molten metal in the mold, but not a change in the level of the molten metal due to the volume change of the molten steel in the unsolidified portion inside the slab due to the bulging of the slab. This is simply the fluctuation of the molten metal level occurring near the surface of the molten steel in the mold.

Therefore, in the present embodiment, the gain and the phase of the phase compensator 23 are controlled in order to suppress the fluctuation of the periodic molten metal level and to stabilize the control of the disturbance in the high frequency range due to the undulation of the molten metal level in the mold. The amount of phase advance of the compensator 23 is optimized together.

FIG. 28 is an explanatory diagram showing control blocks of the level controller 12 of this embodiment. As shown in the figure, the level controller 12 of the present embodiment is different from the level controller 12 of the first embodiment shown in FIG. It has been added.

The phase plane trajectory shape determiner 40 receives a control command for opening and closing a hot water supply mechanism such as a sliding nozzle to adjust the amount of hot water supplied, and a signal of the level change of the level (level level signal). This is an arithmetic unit that determines the appropriateness of the phase compensation amount and the gain from the obtained phase plane trajectory shape.

The phase plane trajectory shape judging unit 40 plots the fluctuation amount of the input level change signal and the fluctuation amount of the opening degree command of the hot water supply mechanism on the same plane to thereby determine the phase plane trajectory shape. Draw.

FIGS. 29 to 31 show the phase advance amount and the phase plane trajectory shape judging unit 4 by the level controller 12.
FIGS. 29 (a) to 31 (a) are explanatory diagrams showing the relationship between the phase surface trajectory shape drawn by 0 and FIG. 29 (a) to FIG. 31 (a); 29 (b) to FIG. 31.
(B) shows the phase plane trajectory shape, the horizontal axis represents the level change of the molten metal level, and the vertical axis represents the opening degree command of the hot water supply mechanism. FIG. 29 shows a case where the amount of phase advance by the phase compensator 23 is 0 °, and FIG.
3 shows a case where the phase advance amount is 90 °, and FIG.
The case where the phase advance amount by the phase compensator 23 is 45 °.

As shown in FIGS. 29A and 29B, when the phase advance by the phase compensator 23 is 0 °, that is, when the phase advance by the phase compensator 23 is not proper, The phase plane trajectory drawn by the trajectory shape determiner 40 is linear.

As shown in FIGS. 30A and 30B, the amount of phase advance by the phase compensator 23 is 90 °.
In other words, when the phase advance amount by the phase compensator 23 is appropriate, the phase plane trajectory shape drawn by the phase plane trajectory shape determination unit 40 has a substantially perfect circular shape.

Further, FIGS. 31 (a) and 31 (b)
As shown in the figure, the amount of phase advance by the phase compensator 23 is 45
°, that is, when the amount of phase advance by the phase compensator 23 is an intermediate value between the cases shown in FIGS. 29 and 30 and is not appropriate, the phase plane trajectory shape drawn by the phase plane trajectory shape determination unit 40 is An elliptical shape changes according to the amount of phase advance.

Assuming that the state of the molten metal surface in the mold is an ideal state in which no undulation of the molten metal surface occurs, the phase advance amount by the phase compensator 23 is set to 90 ° and the phase advance amount by the phase compensator 23 is set. By setting the gain of the compensation command large, the phase plane orbital shape converges to a small perfect circle shape while exhibiting various transient response characteristics showing elliptical shapes.

FIG. 32 is an explanatory diagram showing the relationship between the phase advance amount by the level controller 12 and the phase surface trajectory shape drawn by the phase surface trajectory shape determiner 40 when exhibiting the transient response characteristic. FIG. 32 (a) shows the relationship between the level change (solid line) and the opening degree command (dashed line) of the hot water supply mechanism, while FIG. 32 (b) shows the change in the phase plane trajectory shape. The horizontal axis represents the level change of the hot water level, and the vertical axis represents the opening degree command of the hot water supply mechanism.

However, actually, by setting the gain of the phase compensation command by the phase compensator 23 to be large, or by making the phase advance amount by the phase compensator 23 close to 90 °, the undulation of the molten metal surface in the mold is achieved. May occur, and the instability in the high frequency range may increase, causing divergence.

FIG. 33 shows the amount of phase advance by the level controller 12 when the instability in the high-frequency range is increasing due to the rise of the level of the level in the mold, and the phase plane orbit shape determiner 4.
FIG. 33 (a) is an explanatory view showing the relationship between the phase surface trajectory shape drawn by 0, and FIG. 33 (a) shows the relationship between the level change (solid line) and the opening degree command (dashed line) of the hot water supply mechanism; FIG.
3 (b) shows a change in the shape of the phase surface trajectory. The horizontal axis represents the level change of the molten metal level, and the vertical axis represents the opening degree command of the hot water supply mechanism.

The reason why such a divergence phenomenon occurs is that the phase oscillation of the phase compensator 23 is 0 degree, that is, the level fluctuation outside the band of the band-pass filter 22 provided at the preceding stage of the phase compensator 23 is caused. Because it occurs. This level vibration cannot be suppressed even if the amount of hot water supplied by the hot water supply mechanism is adjusted since the level fluctuation is not accompanied.

Therefore, in the present embodiment, the gain K of the phase compensation command and the phase of the phase compensation command set by the phase compensator 23 based on the phase plane orbit shape obtained by the phase plane orbit shape determiner 40. The advance amount φ is applied and corrected to an optimum value in the following procedure.

FIG. 34 is a diagram showing a phase plane orbit. That is, the shape of the phase plane trajectory between the S / N opening command and the level variation signal is determined by the phase difference between the S / N command using the S / N opening command on the y-axis and the level variation signal on the x-axis and the level variation signal. As shown in FIG.

Here, the phase difference: ψ = arcsin (b / a). 35, FIG. 36, FIG. 37, FIG. 38 and FIG.
It is a figure which shows the shape determination parameter of a phase plane trajectory.

As shown in FIG. 35, the angle formed by the radial axis of the ellipse is denoted by 、, and the rectangular Y circumscribing the elliptical phase plane trajectory
The angle formed by a straight line connecting the origin and the point of contact with the side parallel to the axis is θ, and the lengths of the sides are A and B.

When the phase compensation is small, as shown in FIG. 36, θ is almost the same angle as Θ. However, when the phase compensation advances and the phase difference: 徐 々 に gradually approaches 90 °, FIG. As shown, θ becomes smaller (0 degrees) than 度.

When the control gain is small, as shown in FIG. 38, the y-axis is smaller than the x-axis, so that B> A results in a horizontally long ellipse and rectangle. Conversely, when the control gain is large, as shown in FIG. 39, the y-axis is larger than the x-axis, so that B <A and a vertically long ellipse and rectangle.

Based on these, the shape of the phase plane trajectory, the amount of phase compensation, and the situation of the gain are patterned. 40 to 44 show the phase plane orbit shape obtained by the phase plane orbit shape determiner 40 and the phase surface orbit shape after the control by the phase compensator 23 of the present embodiment. FIGS. 40 (a) to 44 (a) show the phase plane trajectory shapes before control, and FIGS. 40 (b) to 44 (b) show 3 shows the shape of the phase plane orbit.

The phase compensator 23 is set to an initial value K ₀ (in this embodiment, 90 °) of the gain K of the phase compensation command. This initial value K ₀ is set in advance by an appropriate means such as a confirmation experiment or a simulation.

The phase compensator 23 sets the gain K of the phase compensation command to the initial value K ₀ unless the instability of the high frequency band appears.
Start changing the settings. The gain K of the phase compensation command
If the instability of the high frequency range already exists at the time when is the initial value K ₀ , the amount of phase advance by the phase compensator 23 is temporarily subtracted from 90 ° by an appropriate increment to reduce it. When the setting of the gain K of the phase compensation command is changed, the phase compensator 23 may temporarily add an appropriate increment to the setting change amount.

If the instability in the high frequency range occurs during the change of the gain K of the phase compensation command, the gain K at the time of occurrence is temporarily subtracted from the gain K to change the setting. That is, when the phase plane trajectory shape obtained by the phase plane trajectory shape determiner 40 is as shown in FIG. 40A, the angle θ is smaller than Θ and close to 0 °, so that the phase advance amount φ is appropriate. Although there is, the value A / B of the side of the rectangle circumscribing the phase plane orbit
Is a small horizontally long flat ellipse, it can be understood that the gain K of the phase compensation command is smaller than an appropriate value. Therefore, in this case, the phase compensator 23, the gain K of the phase compensation command, as long as the instability of the high-frequency region does not appear, it should set significantly changed from the initial value K _0.

When the phase plane trajectory shape obtained by the phase plane trajectory shape determiner 40 is as shown in FIG. 41 (a), the angle θ is smaller than に and close to 0 °. Is appropriate, but since the value A / B of the side of the rectangle circumscribing the phase plane trajectory is a vertically long flat ellipse, it can be seen that the gain K of the phase compensation command is larger than an appropriate value. Therefore, in this case, the phase compensator 23, the gain K of the phase compensation command, as long as the instability of the high-frequency region does not appear, continue to configuration changes decreases from the initial value K _0.

When the phase plane trajectory shape obtained by the phase plane trajectory shape determiner 40 is shown in FIG. 42 (a), the angle θ is large and close to か ら, so that the phase advance amount φ
Is smaller than the proper value, and the value A / B of the side of the rectangle circumscribing the phase plane trajectory is a long, flat oblong ellipse, indicating that the gain K of the phase compensation command is larger than the proper value. Therefore, in this case, the phase compensator 23
Together with is increased set change phase advance φ of the phase compensation command, the gain K of the phase compensation command, as long as the instability of the high-frequency region does not appear, continue to configuration changes decreases from the initial value K _0.

When the phase plane trajectory shape obtained by the phase plane trajectory shape determiner 40 is as shown in FIG. 43 (a), the angle θ is large and close to Θ, so that the phase advance amount φ
Is smaller than the appropriate value, and the value A / B of the side of the rectangle circumscribing the phase plane trajectory is a vertically long flat ellipse, which indicates that the gain K of the phase compensation command is smaller than the appropriate value. Therefore, in this case, the phase compensator 23
Sets the phase advance amount φ of the phase compensation command to a large value and changes the gain K of the phase compensation command to a large value from the initial value K ₀ unless instability in a high frequency range appears.

Further, when the phase plane trajectory shape obtained by the phase plane trajectory shape determiner 40 is shown in FIG. 44 (a), the angle θ is large and close to Θ, so that the phase advance φ is Since the value A / B of the side of the rectangle circumscribing the phase plane trajectory is smaller than the appropriate value and the value A / B is close to 1, it can be seen that the gain K of the phase compensation command is close to the appropriate value. Therefore, in this case, the phase compensator 23 changes the phase advance amount φ of the phase compensation command to a large value.

As a result, FIGS. 40 (a) to 44 (a)
In any of the cases shown in FIGS.
As shown in (b), both the phase advance amount φ of the phase compensation command and the gain K of the phase compensation command approach the optimum value, and accordingly, the phase plane orbital shape approaches a small perfect circular shape. To go.

FIG. 45 is a flow chart showing a decision logic when the phase compensator 23 of this embodiment sets and changes the phase advancing amount φ of the phase compensation command and the gain K of the phase compensation command to optimum values. It is.

In FIG. 45, steps (hereinafter, “S”)
1), the initial value K ₀ of the gain K of the phase compensation command, and the initial value φ ₀ of the phase advance amount φ of the phase compensation command =
The control is started at 90 °, a change amount ΔK of the gain K of the phase compensation command, and a change amount Δφ of the phase advance amount φ of the phase compensation command. Then, the process proceeds to S2.

At S2, it is determined whether or not the phase is optimal based on the obtained phase plane trajectory shape. If it is determined that it is optimal, the process proceeds to S3, and if it is determined that it is not optimal, the process proceeds to S4.

At S3, a new set value K ₀ of the gain K of the phase compensation command is obtained as K ₀ = K ₀ + ΔK. Then, the process proceeds to S5. On the other hand, in S4, the new set value φ ₀ of the phase advance amount φ of the phase compensation command is φ ₀ =
It is calculated as φ ₀ −Δφ. Then, control goes to a step S12.

In S12, it is determined whether or not the phase is optimal based on the obtained phase plane trajectory shape. If it is determined that it is optimal, the process proceeds to S13, and if it is determined that it is not optimal, the process proceeds to S4.

In step S13, the phase of the phase compensation command is advanced.
New set value of quantity φ₀Is φ₀= Φ ₀-Δφ / 2ⁿage
Is calculated. Then, control goes to a step S14. In S14
And, as in S13, φ₁~ Φ_{n + 1}Is calculated. So
Then, the process proceeds to S15.

In S15, it is determined whether or not the obtained phase plane trajectory shape is optimal. If it is determined that it is optimal, the process proceeds to S13, and if it is determined that it is not optimal, the process proceeds to S16.

In S16, the change amount Δφ of the phase advance amount φ of the phase compensation command is set as Δφ = −Δφ.
Then, control goes to a step S17. In S17, n ≦ n
It is determined whether it is _max . If n ≦ n _max , the process proceeds to S13. If n ≦ n _max , the process proceeds to S15.
Move to

At S5, it is determined whether or not the obtained phase plane orbital shape is optimal. If it is determined that it is optimal, the process proceeds to S3, and if it is determined that it is not optimal, the process proceeds to S6.

At S6, a new set value K ₀ of the gain K of the phase compensation command is calculated as K ₀ = φ ₀ −ΔK / 2 ⁿ . Then, control goes to a step S7. In S7, S6
Similarly, φ ₁ to φ _{n + 1} are calculated. Then, control goes to a step S8.

At S8, it is determined whether or not the obtained phase plane orbital shape is optimal. If it is determined that it is optimal, the process proceeds to S6, and if it is determined that it is not optimal, the process proceeds to S9.

In S9, the change amount ΔK of the gain K of the phase compensation command is set as ΔK = −ΔK. Then, the process proceeds to S10. In S10, it is determined whether or not n ≦ n _max . S if n ≦ n _max
The process proceeds to S11, and if n ≦ _nmax , the process proceeds to S11.

In S11, the control ends. FIG.
Is determined by the decision logic shown in FIG.
When both the phase advance amount φ of the phase compensation command and the gain K of the phase compensation command are set and changed to the optimum values,
It is a figure showing the situation where each of gain K and phase advance φ converges.

As shown in FIG. 46, the gain K is changed by S3 in FIG. 45, and the phase advance amount φ is determined by S4 in FIG.
Is further changed, and the gain K is further changed in S3 of FIG. 45, whereby both the phase advance amount φ of the phase compensation command and the gain K of the phase compensation command are set and changed to the optimum values.

As described above, according to the present embodiment, the elliptical roundness of the phase plane trajectory obtained as a phase plane diagram of the fluctuation signal of the molten metal level and the opening control command of the hot water supply mechanism is improved. , The amount of phase advance φ and the gain K by the phase compensator 23 can both be corrected and set to optimal values so that the area decreases.

Therefore, according to the present embodiment, even if the gain of the phase compensator 23 is changed or the phase is changed, or the phase is changed, the control characteristics are stable even in a high frequency range, and Also in the level control, the control is stabilized even when the surface of the molten metal in the mold is wavy. For this reason, according to the present embodiment, it is possible to suppress the periodic fluctuation of the molten metal level and to stably control the disturbance in the high frequency range due to the undulation of the molten metal level in the mold.

[0198]

EXAMPLE (Example 1) As an example of the present invention, a control simulation was performed using a control system shown in FIG. The casting conditions used for the simulation are as follows.

Mold dimensions: thickness 90 mm × width 1200 mm, casting speed: 3.0 m / min, secondary cooling zone roll spacing: 200 mm.

In order to compare the control method according to the present invention with the prior art, a control simulation using a control system including only a level controller shown in FIG. 3 was also performed. The control system gain of the entire control loop of the conventional control system is maximum at 0.25 Hz as shown in FIG. The control parameters (control gain and integration time) of the level controller of the present invention were the same as those of the conventional example.

The notch frequency of the notch filter 21 and the band-pass frequency of the band-pass filter 22 in FIG. 12 are adjusted to the same values as the above-mentioned periodic disturbance frequency, the notch ratio g = 0.2, and the control of the level controller. Gain K _p =
1.0 and the phase compensation gain K _g = 0.8.

As a disturbance condition assuming a disturbance generated in the medium-thick high-speed continuous casting machine, a periodic disturbance frequency f ₂ : 0.25
Hz (= casting speed / roll interval), amplitude 1080 cm ³
/ Second volume change (corresponding to ± 10 mm / second at the molten metal level) was applied to the control system. This frequency corresponds to the resonance frequency of the control system of the prior art.

FIG. 47 is a graph showing a control result according to the prior art. FIG. 48 is a graph showing a control result according to the present invention. In the control example according to the prior art in FIG. 47, the actual level change of the molten metal level increases to ± 25 mm with respect to the applied volume fluctuation disturbance. As a result of this control, the actual fluctuation in the level of the molten metal could be suppressed to about ± 15 mm. However, this level of level variation can result in significant quality defects or breakouts in the actual continuous casting process.

As shown in FIG. 48, in the control example according to the present invention, initially, ± 10 mm close to the volume fluctuation disturbance amplitude.
Fluctuations in the level of the bath surface were observed, but ± 5 in about 10 seconds thereafter.
mm. This level variation is acceptable in an actual continuous casting process.

As described above, in the prior art, when the frequency of the disturbance (unsteady bulging) becomes the same as the resonance frequency inherent to the control system, the fluctuation of the molten metal level becomes extremely large. The frequency is suppressed in the control loop, and the volume compensation due to bulging is canceled out by adjusting the amount of hot water supplied by the phase compensation calculation unit (band-pass filter and phase compensator). It was found that it was possible to suppress the fluctuation of the fluent level.

(Embodiment 2) Using the control system of the present invention shown in FIG. 22, unsteady bulging and level change caused by pinch roll eccentricity coexist, and the frequency of the unsteady bulging type level change is A simulation was performed in the case of automatically setting the parameters of the control system in the case of change. As shown in FIG. 22, the targets of the automatic setting are the frequency of the notch filter and the band-pass filter, the control gain of the surface level controller, the notch ratio, and the phase compensation gain. The casting conditions used for the simulation were as follows.

Mold size: thickness 90 mm × width 1200 mm, casting speed V _c : 2.0 to 5.0 m / min, pinch roll diameter R _SC : 100 mm Secondary cooling zone roll interval: d ₁ = 200 mm, d ₂ = 25
0 mm (2 types).

Periodic disturbance frequency and amplitude: f ₁ = V _c / 2πR _SC = 0.05 to 0.13 Hz, ₂ mm f ₂ = V _c / d ₁ = 0.17 to 0.42 Hz, 3 mm f ₃ = V _c / d ₁ = _0.13~0.33Hz, the 3mm simulation sequentially increasing casting speed from the start of casting, unsteady bulging with a roll distance d ₁ is generated between them (time T1), then the non-stationary by d ₂ It is assumed that bulging occurs due to superposition (time T2) and that periodic disturbance due to pinch roll eccentricity occurs.

FIG. 49 is a graph showing a control result by the automatic setting function of the present invention. As shown in the figure, when the unsteady bulging does not occur at first, the level (square root of the root mean square of the level difference every 4 seconds) was stable, but at the time T1, the first unsteady bulging was Fluctuations in level increased when they occurred. After a while, the control parameters were optimized and the level change of the molten metal level was reduced. Time T2
Then, when new unsteady bulging occurred, the fluctuation of the molten metal level increased again, but it became stable for a while. Next, when a periodic disturbance due to pinch roll eccentricity occurred at time T3, the fluctuation of the molten metal level slightly increased, but soon stabilized.

From the above simulation, it was found that F of the present invention
It has been found that the automatic setting of parameters such as the notch frequency and the control gain of the level controller by the FT method can effectively cope with a change in the condition of unsteady bulging.

Example 3 A simulation was performed to compare the automatic setting of the notch frequency and the band-pass frequency by the FFT method of the present invention with that by the PLL method of the present invention.

The casting conditions for the simulation were the FFT method,
The PLL method was as follows. Casting speed: 3.0m
/ Min, the casting speed is increased to 3.6 m / min in 10 seconds, the secondary cooling zone roll interval: 180 mm, and as a result, the periodic disturbance frequency due to unsteady bulging: 0.2
It increased from 78 Hz to 0.333 Hz.

FIG. 50 is a graph showing conditions of the casting speed and the periodic disturbance frequency during the simulation of the FFT method of the present invention. In the figure, in the FFT method, sampling for frequency analysis is started from time 0.
Before collecting 12 samples, the casting speed changed and the periodic disturbance frequency changed. At the end of the sampling interval, the detected frequency is the value before the casting speed was increased, and the detection of the frequency after the change was delayed until the end of the current sampling interval.

FIG. 51 is a graph showing the change of the molten metal level by the FFT method. As shown in the figure, although the amplitude of the fluctuation of the molten metal level was about 1 mm at the beginning, it rapidly increased from the point of time t = 40 seconds when the casting speed began to increase, and t = 1.
At around 00 seconds, it increased to about 5.5 mm.

FIG. 52 is a graph showing a change in the molten metal level by the PLL method. As shown in the figure, although the amplitude of the fluctuation of the molten metal level was initially about 1 mm, it increased from the time when the casting speed began to increase. However, the maximum amplitude becomes 1.8 mm at t = about 55 seconds, and then gradually decreases to t = 100
After a second, the amplitude returned to the original value.

As can be seen from the above comparison, the PLL method makes it possible to control the molten metal level stably with respect to the fluctuation of the casting speed. (Example 4) A control simulation was performed using the molten metal level controller 12 of the second embodiment described with reference to FIGS. The casting conditions used in the simulation were the same as those in Example 1 described above.

FIG. 53 is a graph showing the results.
FIG. 53 shows the amount of phase advance by the level controller 12;
FIG. 53 (a) is a diagram illustrating the relationship between the phase surface trajectory shape drawn by the phase surface trajectory shape determiner 40, and FIG. On the other hand, FIG. 53 (b) shows the shape of the phase plane trajectory, in which the horizontal axis represents the level change of the molten metal level and the vertical axis represents the opening degree command of the hot water supply mechanism.

As shown in FIG. 53 (a), when the drawing speed Vc was increased after about 5 seconds, a periodic level fluctuation occurred. However, in this embodiment, the phase compensator 23 immediately corrected and set the gain K and the phase lead amount φ of the phase compensation command to optimum values, respectively, so that the fluctuation of the molten metal level could be suppressed after about 13 seconds.

[0219]

According to the control method and the control device of the present invention, it is possible to effectively suppress the fluctuation of the molten metal level due to the periodic disturbance caused by the unsteady bulging property or the eccentricity of the pinch roll in the continuous casting machine. Further, according to the present invention, even when the periodic disturbance frequency changes, the parameters of the control system can be optimally followed, and stable control can always be realized even when high-speed casting is performed.

Further, according to the control method and the control apparatus of the present invention, even if the gain of the phase compensator is changed or the phase is changed, or the phase is changed, the control characteristics are stabilized even in a high frequency range. In the actual control of the mold level, even if the mold level in the mold rises, the control is stable. Therefore, it is possible to suppress the periodic fluctuation of the molten metal level and to stabilize the control against the disturbance in the high frequency range due to the undulation of the molten metal level in the mold.

[Brief description of the drawings]

FIG. 1 is a schematic diagram showing a molten metal level control system of a continuous casting machine.

FIGS. 2A and 2B are schematic diagrams showing a state of occurrence of unsteady bulging. FIG. 2A shows a case where a slab expands, and FIG. 2B shows a case where the slab shrinks.

FIG. 3 is a block diagram showing a control system of FIG. 1;

FIG. 4 is an example of a graph showing a change in the level of the molten metal when unsteady bulging occurs.

FIG. 5 is a graph showing a control system gain of the control system shown in FIG. 3, that is, a magnitude of a change in the molten metal level with respect to a disturbance input, as a frequency comparison.

FIG. 6 is a graph showing a frequency spectrum of the unsteady bulging level level fluctuation shown in FIG. 4;

FIG. 7 is a graph showing a change in the level of the molten metal in the control system including a plurality of periodic disturbance frequencies.

FIG. 8 is a graph showing a frequency spectrum of a change in the level of the molten metal shown in FIG. 7;

FIG. 9 is a block diagram illustrating an example of a control method for suppressing unsteady bulging.

FIG. 10 is a graph showing a filter gain of the notch filter shown in FIG. 9;

FIG. 11 is a graph showing a change in the molten metal level when a simulation is performed by the control system shown in FIG. 9;

FIG. 12 is a block diagram illustrating a control method of the present invention.

FIG. 13 is a graph showing the relationship between the frequency of a bandpass filter and its gain (transmittance).

FIG. 14 is a graph showing a relationship between an input and an output of the phase compensator.

FIG. 15 is a graph showing a disturbance suppression simulation by the control system shown in FIG. 9;

FIG. 16 is a graph showing a simulation result of disturbance suppression by the control system of the present invention shown in FIG.

FIG. 17 is a block diagram illustrating a control method for automatically adjusting a notch frequency and a bandpass frequency according to the present invention.

FIG. 18 is a block diagram showing a method for automatically setting various gains in the control method of the present invention.

FIG. 19 is a graph showing an example of a method for calculating a notch ratio g of a notch filter.

FIG. 20 is a graph showing an example of setting a correction coefficient of the control gain _Kp of the liquid level controller with respect to the frequency of the liquid level fluctuation of the present invention.

FIG. 21 is a flowchart illustrating an example of a method of setting a phase compensation gain K _{g according to} the present invention.

FIG. 22 is a block diagram showing a control system corresponding to a plurality of periodic disturbance frequencies according to the present invention.

FIG. 23 is a block diagram showing a frequency analysis method according to the tuning type frequency analysis method of the present invention.

FIG. 24 is a phase diagram related to Θ (t) for studying the solution of a differential equation according to the PLL method of the present invention.

FIG. 25 is a graph showing a simulation result of a state where the oscillation frequency of the variable frequency oscillator is tuned to a signal having a periodic disturbance frequency.

FIG. 26 is an explanatory diagram schematically showing a technique disclosed in Japanese Patent Application Laid-Open No. 10-314911.

FIG. 27 is an explanatory diagram schematically showing the first embodiment.

FIG. 28 is an explanatory diagram showing control blocks of a molten metal level controller according to the second embodiment.

FIG. 29 is an explanatory diagram showing the relationship between the phase advance amount by the level controller and the phase plane trajectory shape drawn by the phase plane trajectory shape determiner when the phase advance amount by the phase compensator is 0 °; FIG. 29 (a) shows the relationship between the level change (solid line) and the opening degree command (dashed line) of the hot water supply mechanism, while FIG. 29 (b) shows the phase plane trajectory, The horizontal axis represents the level change of the hot water level, and the vertical axis represents the opening degree command of the hot water supply mechanism.

FIG. 30 shows the amount of phase advance by the molten metal level controller when the amount of phase advance by the phase compensator is 90 °;
It is explanatory drawing which shows the relationship with the phase surface trajectory shape drawn by the phase surface trajectory shape determination device, and FIG. On the other hand, FIG. 30 (b) shows the shape of the phase surface trajectory, in which the horizontal axis represents the level change of the molten metal level and the vertical axis represents the opening degree command of the hot water supply mechanism.

FIG. 31 shows the amount of phase advance by the molten metal level controller when the amount of phase advance by the phase compensator is 45 °;
FIG. 31 (a) is a diagram illustrating a relationship between a phase surface trajectory shape drawn by a phase surface trajectory shape determiner, and FIG. On the other hand, FIG. 31 (b) shows the shape of the phase plane trajectory, in which the horizontal axis represents the level of the molten metal level and the vertical axis represents the opening degree command of the hot water supply mechanism.

FIG. 32 is a diagram showing the relationship between the amount of phase advance by the surface level controller 12 and the phase surface trajectory shape drawn by the phase surface trajectory shape determiner 40 when the transient response characteristic is exhibited. FIG. 32 (a) shows the relationship between the change of the molten metal level (solid line) and the opening degree command (dashed line) of the hot water supply mechanism, while FIG. 32 (b) shows the change of the phase plane trajectory shape. The horizontal axis represents the level change of the hot water level, and the vertical axis represents the opening degree command of the hot water supply mechanism.

FIG. 33 shows the amount of phase advance by the level controller and the phase plane trajectory drawn by the phase plane trajectory shape determiner when the instability in the high-frequency region is increasing due to the occurrence of undulation in the mold surface in the mold. FIG. 33 is an explanatory diagram showing the relationship with the shape, and FIG.
(A) shows the relationship between the level change (solid line) and the opening command (dashed line) of the hot water supply mechanism, while FIG. 33 (b) shows the change in the phase surface trajectory shape, with the horizontal axis representing the hot water. The vertical axis indicates the opening degree command of the hot water supply mechanism.

FIG. 34 is a diagram showing a phase plane trajectory.

FIG. 35 is a diagram showing shape determination parameters of a phase plane trajectory.

FIG. 36 is a diagram showing a shape determination parameter of a phase plane trajectory.

FIG. 37 is a diagram showing parameters for determining the shape of the phase plane trajectory.

FIG. 38 is a diagram showing shape determination parameters of a phase plane trajectory.

FIG. 39 is a diagram showing a shape determination parameter of a phase plane trajectory.

FIG. 40 is an explanatory diagram showing the phase plane trajectory shape obtained by the phase plane trajectory shape determiner and the phase surface trajectory shape after control by the phase compensator of the present embodiment;
0 (a) shows the phase plane trajectory shape before control, and FIG.
(B) shows the phase plane trajectory shape after control.

FIG. 41 is an explanatory diagram showing a phase plane trajectory shape obtained by the phase plane trajectory shape determination device and the phase surface trajectory shape after control by the phase compensator of the present embodiment;
1 (a) shows the phase plane trajectory shape before control, and FIG.
(B) shows the phase plane trajectory shape after control.

FIG. 42 is an explanatory diagram showing the phase plane orbit shape obtained by the phase plane orbit shape determiner and the phase surface orbit shape after control by the phase compensator according to the present embodiment;
2 (a) shows the phase plane trajectory shape before control, and FIG.
(B) shows the phase plane trajectory shape after control.

FIG. 43 is an explanatory diagram showing the phase surface trajectory shape obtained by the phase surface trajectory shape determiner and the phase surface trajectory shape after control by the phase compensator according to the present embodiment;
FIG. 3 (a) shows the shape of the phase plane trajectory before control.
(B) shows the phase plane trajectory shape after control.

FIG. 44 is an explanatory diagram showing the phase plane orbit shape obtained by the phase plane orbit shape determiner and the phase surface orbit shape after control by the phase compensator according to the present embodiment;
FIG. 4A shows the phase plane trajectory shape before control, and FIG.
(B) shows the phase plane trajectory shape after control.

FIG. 45 is a flowchart showing a determination logic when the phase compensator of the second embodiment changes the phase advance amount φ of the phase compensation command and the gain K of the phase compensation command to optimal values.

46 is a diagram illustrating the determination logic shown in FIG. 45, in which the phase compensator sets the gain K and the phase advance when both the phase advance amount φ of the phase compensation instruction and the gain K of the phase compensation instruction are changed to optimal values. It is a graph which shows the situation in which the quantity φ converges.

FIG. 47 is a graph showing a control result according to the related art.

FIG. 48 is a graph showing a control result according to the present invention.

FIG. 49 is a graph showing a control result by the automatic setting function of the present invention.

FIG. 50 is a graph showing conditions of a casting speed and a periodic disturbance frequency during simulation of the FFT method of the present invention.

FIG. 51 is a graph showing a change in the molten metal level by the FFT method.

FIG. 52 is a graph showing a change in the level of the molten metal by the PLL method.

FIG. 53 is a graph showing the results of Example 4.

[Explanation of symbols]

 1: molten metal 2: tundish 3: immersion nozzle 4: mold 5: cast piece 6: shell 7: unsolidified portion 8: secondary cooling zone roll 9: pinch roll 10: drive motor 11: level gauge 12: hot water Surface level controller 13: Stopper drive unit 14: Stopper 15: Deviation calculation unit 16: Control law unit 17: Transfer function of stopper drive unit 18: Transfer function of stopper 19: Transfer function of mold 20: Transfer of level gauge Function 21: Notch filter 22: Band pass filter 23: Phase compensator 24: Phase compensation gain unit 25: Phase compensation calculation unit 26: Output addition unit 27: Frequency analysis unit 28: Notch ratio setting unit 29: Control gain setting unit 30 : Composite compensation notch filter 32: Composite notch filter 32: Composite phase compensation calculator 33: Composite phase compensation calculator adder 34: Variable frequency oscillation 35: multiplier 36: low-pass filter 37: Frequency meter

Claims (11)

[Claims] 1. A method of controlling a level of a continuous casting machine using a level controller, the method comprising: determining a periodic disturbance frequency of a level change caused by unsteady bulging or eccentricity of a pinch roll; In a control loop of a level control system, a variable component of a specific notch frequency is selectively attenuated, and a notch filter adjusted so that the notch frequency includes the periodic disturbance frequency is interposed. A band-pass filter selectively transmitting a fluctuation component of a pass frequency and adjusting the band-pass frequency to include the periodic disturbance frequency; and a phase compensation adjusting a phase compensation frequency to include the periodic disturbance frequency. And a phase compensation gain unit that multiplies the input signal by a phase compensation gain and outputs the result, thereby forming a phase compensation calculation unit; and A level control method for a continuous casting machine, characterized in that a level difference is inputted to the phase compensating section and an output of the phase compensating section is added to a calculation output of the level controller. 2. The frequency of a peak level component is determined by performing a frequency analysis of a change in the level of the molten metal, and 0.1% of these frequencies is calculated.
2. The method according to claim 1, wherein the periodic disturbance frequency is obtained by selecting a frequency equal to or higher than Hz. 3. An amplitude of a peak component obtained by the frequency analysis is obtained, a notch ratio of the notch filter is set according to the amplitude, and a frequency of 0.1 Hz among peak components obtained by the frequency analysis is obtained. 3. The method according to claim 2, wherein the control gain of the level controller is set according to the lowest frequency. 4. The phase compensation gain of the phase compensation calculation unit is increased or decreased by a small amount, a predetermined time control is performed, and then the variation of the output value of the band-pass filter is measured. Based on the relationship of the fluctuation direction,
4. The method according to claim 2, wherein a new value of the phase compensation gain is set so as to reduce the variation. 5. A configuration in which the notch filter and the phase compensation operation unit are replaced with a composite notch filter and a composite phase compensation operation unit, respectively, wherein the composite notch filter is provided with a notch for each of a plurality of periodic disturbance frequencies. A plurality of notch filters which are adjusted in frequency to include each periodic disturbance frequency and are connected in series, and wherein the composite phase compensation operation unit comprises a plurality of phase compensation operation units and at least one composite phase compensation operation unit An adder, and the composite phase compensation operation unit receives the level difference from each of the plurality of phase compensation operation units, and outputs the output of each of the plurality of phase compensation operation units to the composite phase compensation operation unit. The result of the addition by the partial adder is configured to be added to the operation output of the level controller, and each of the plurality of phase compensation operation units includes The bandpass filter, the phase compensator, and the phase compensator are configured such that each bandpass frequency and phase compensation frequency correspond to each periodic disturbance frequency. The method of controlling a molten metal level of a continuous casting machine according to any one of claims 1 to 4, wherein the method is adjusted to include the molten metal. 6. A variable frequency oscillator is interposed in the control loop when a frequency of the peak component is determined by performing a frequency analysis of the level difference, and a signal of the level change and a signal of the variable frequency oscillator are provided. Multiply the signal in real time, based on the difference signal between the level of the molten metal level obtained from the multiplication result and the frequency of the variable frequency oscillator, change the oscillation frequency of the variable frequency oscillator,
3. The method according to claim 2, wherein the oscillation frequency is tuned to the periodic disturbance frequency of the level change, and the periodic disturbance frequency is obtained from the tuned oscillation frequency. 7. A phase plane trajectory shape judging device is interposed in the control loop, a signal of a periodic molten metal level change, and an opening of a hot water supply mechanism attached to a tundish for supplying molten steel into a mold. And the degree command is input to the phase-plane trajectory shape determiner, and the elliptical shape obtained by plotting the variation of the signal of the level difference and the variation of the opening degree command of the hot water supply mechanism on the same plane. The phase advance amount by the phase compensator and / or the phase compensation gain by the phase compensation gain unit are corrected and set so that the roundness of the phase plane orbital shape of the above is improved and the area is reduced. 7. A method for controlling a molten metal level of a continuous casting machine according to claim 1 or claim 6. And a control system for selectively attenuating a fluctuation component of a specific notch frequency, wherein the notch frequency is a periodic disturbance of the metal surface level. A notch filter adjusted to include a frequency, a band-pass filter selectively transmitting a variable component of a specific band-pass frequency, and adjusting the band-pass frequency to include the periodic disturbance frequency; A phase compensator whose compensation frequency is adjusted to include the periodic disturbance frequency, and a phase compensation gain unit that multiplies an input signal by a phase compensation gain by the phase compensator and outputs the result are connected in series. A notch filter for receiving a level signal from the level sensor and controlling a level change signal in the level control. The band-pass filter receives the level signal from the level sensor and sends a band-pass frequency signal adjusted to include a periodic disturbance frequency to the phase compensator. The phase compensator has a function of sending a phase compensation frequency signal adjusted to include a periodic disturbance frequency to a phase compensation gain unit, and further, the phase compensation calculation unit outputs the output A level controller for a continuous casting machine, having a function of adding to a calculation output of a level controller. 9. The apparatus according to claim 8, wherein the control system includes a frequency detection operation unit for detecting the periodic disturbance frequency. 10. The frequency detection calculation unit includes a variable frequency oscillator, and converts a signal of a periodic level change output from the level sensor and a signal oscillated from the variable frequency oscillator in real time. The oscillation frequency of the variable frequency oscillator is changed based on the level difference obtained from the multiplication result and the difference signal between the frequencies of the variable frequency oscillators. 10. The apparatus for controlling the level of a continuous casting machine according to claim 9, wherein the apparatus has a function of tuning to a periodic disturbance frequency and calculating a periodic disturbance frequency of the periodic level fluctuation from the tuned oscillation frequency. . 11. A phase plane trajectory shape determiner for adjusting an output of the phase compensation operation unit to be added to an operation output of the level controller, wherein the control system includes: Elliptical phase plane trajectory obtained by plotting on the same plane the amount of variation of the signal of the periodic molten metal level fluctuation and the amount of variation of the opening degree command of the hot water supply mechanism of the continuous casting machine. Has a function of correcting and setting the phase advance amount by the phase compensator and / or the phase compensation gain by the phase compensation gain unit so that the roundness of the phase is improved and the area is reduced. The molten metal level control device for a continuous casting machine according to any one of claims 8 to 10, characterized in that: