JP2003245761A - Control method of molten steel surface level inside mold in continuous casting machine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a control method at a low cost by which a molten steel surface level inside a mold can be stably maintained even if an operation condition is changed in continuous casting. <P>SOLUTION: The method controls the molten steel surface level in the mold by increasing/decreasing a cast slab drawing-out speed to an amount equivalent to an increase/decrease amount calculated from the deviation and the increase/ decrease logic of the cast slab drawing-out speed set in advance so as to make the deviation of a temperature measured value by a temperature sensor embedded in the mold and a control target temperature set in advance to be smaller. Sample data are extracted successively from a population comprised of data in time series of the temperature measured value by the temperature sensor and a histogram regarding the sample data is formed and displayed. The control of the molten steel surface level with high accuracy is achieved by correcting the control target temperature so as to bring the histogram close to a normal distribution curve. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for controlling a molten metal level in a mold in continuous casting.

[0002]

2. Description of the Related Art In a continuous casting machine in which a molten metal is continuously supplied to a mold and a slab is continuously drawn out of the mold, the level of the molten metal in the mold varies during casting. It also reduces the quality (especially the quality of the surface) and, if significant, also induces breakouts. Therefore, in order to stably produce a high quality slab, the molten metal level in the mold during casting is always maintained at a predetermined level,
In particular, it is important to prevent sudden changes in the molten metal level.

Therefore, conventionally, in order to prevent the fluctuation of the molten metal level in the mold, the fluctuation of the molten metal level is detected by some means, and the molten metal inflow amount from the tundish to the mold is increased or decreased to keep the molten metal level constant. The automatic control method is adopted so that On the other hand, when casting is performed with a constant amount of injection into the mold without a flow rate adjusting mechanism at the tundish discharge port like many billet continuous casting machines, fluctuations in the molten metal level are detected by some means A method of automatically controlling the molten metal surface level to be constant by increasing or decreasing the speed, that is, the withdrawal speed of the slab is adopted.

As means for detecting the level of molten metal in the mold, γ
Wire system, image system, eddy current sensor system, copper plate temperature detection system, etc. are adopted.

In the γ-ray method, a γ-ray source and a radiation sensor are provided in pairs on both sides near the level of the molten metal of the mold, and the level of the molten metal is obtained by measuring the transmission intensity of radiation, which is extremely good. Although it has a responsiveness, it is becoming obsolete in recent years because it uses a radioactive substance harmful to the human body as a radiation source.

In the image system, a camera installed diagonally above the surface of the molten metal in the mold is used to monitor the brightness of the molten metal from a bird's-eye view, and the image is image-processed to obtain the molten metal level. Then, there is a drawback that it becomes undetectable.

In the eddy current sensor system, a pair of transmitter and sensor are installed on the upper part of the mold, and the primary magnetic flux from the transmitter generates an eddy current in the copper plate of the mold. This is a method in which the magnetic flux is detected by a sensor and the molten metal level is measured. It has good responsiveness and is safe for the human body, but since the sensor's external dimensions are relatively large and it is necessary to install the sensor immediately above the molten metal surface, install the sensor in the mold with a small cross section of the billet continuous casting machine. Not applicable due to lack of space.

In the copper plate temperature detection method, a temperature sensor (for example, a thermocouple) is embedded in a mold copper plate and the change in the copper plate temperature is detected to indirectly grasp the molten metal level. This method has the drawback of being less responsive than other methods, but it is safe for the human body, does not impair the workability of the operator, can be applied to small cross-section molds, and the sensor itself is inexpensive. In recent years, it has become the mainstream of the method for detecting the molten metal level in the billet continuous casting machine.

FIG. 14 schematically shows the temperature distribution of the copper plate in the flow direction of the cast product at the position of the depth D (the position of the broken line) from the inner surface of the mold copper plate. The copper plate temperature at this position rises from above the molten metal level toward the molten metal level, reaches a maximum value slightly below the molten metal level, and then gradually decreases as it further decreases.

As shown in FIG. 14, when the temperature sensor is embedded in a position where the temperature of the copper plate tends to rise with respect to the flow direction of the slab, that is, in the copper plate directly below the molten metal level, the temperature of the copper plate detected by the temperature sensor is detected. The (sensor temperature) rises because the entire temperature distribution curve moves upward as the temperature rises, and decreases as the temperature rises because the entire temperature distribution curve moves downward as the temperature rises.

It is common practice to utilize this characteristic to automatically control the molten metal level. For example, in a billet continuous casting machine with a constant amount of molten metal injected from the tundish into the mold, when the molten metal surface level rises (that is, the sensor temperature rises), the casting speed (that is, slab drawing speed) increases, while When the level decreases (that is, the sensor temperature decreases), the casting speed (that is, the slab drawing speed) is decreased, and automatic control is performed to maintain the molten metal level at a constant level.

Such automatic control of the molten metal level is specifically performed by the following method. First, a set temperature Tc as a control target temperature with respect to the sensor temperature Ts, and a slab drawing speed increase / decrease logic that is a control factor for obtaining an increase / decrease amount of the slab drawing speed (hereinafter simply referred to as “increase / decrease logic”).
Say. ) The coefficient of Lv and the like are predetermined. Then, when the sensor temperature Ts measured online is higher than the set temperature Tc, the slab drawing speed v is increased, and conversely, when it is low, the slab drawing speed v is decreased. The increase / decrease amount Δv of the slab drawing speed is calculated from the deviation (ΔT = Ts−Tc) between the sensor temperature Ts and the set temperature Tc and the increase / decrease logic Lv.

[0013]

However, the proper set temperature Tc and the increase / decrease logic Lv are different for each continuous casting machine and for each type of steel to be cast, and it is not easy to determine these proper values, and conventionally, it is difficult for the operator to set them. I depended on my experience and intuition.

That is, the temperature distribution of the copper plate is the structure and operating conditions of the mold (material of the copper plate of the mold, taper amount of the mold, temperature and flow velocity of the mold cooling water, burying depth of the temperature sensor, surface condition of the copper plate, mold lubrication, molten steel). Composition and temperature, casting speed, etc.)
It changes in various ways. Therefore, as schematically shown in FIG. 14, when the steel plate A and the steel plate B have greatly different copper plate temperature distributions, the set temperature Tc and the increase / decrease logic Lv determined to be appropriate for the steel type A are not appropriate for the steel type B as they are. As a result, the automatic level control will not be performed normally.

Further, even when casting the same steel type, the surface condition of the copper plate, the condition of mold lubrication, the temperature and composition of the molten steel, etc. change from moment to moment during operation and the temperature distribution inside the copper plate changes. Therefore, if the set temperature Tc and the increase / decrease logic Lv that are initially determined to be appropriate are used as they are, the control of the molten metal level may not be normally performed.

Therefore, originally, the set temperature Tc and the increase / decrease logic Lv should be changed each time in response to such a change of steel type and a change of operating conditions, but the changing procedure is complicated and the work is not performed. Not realistic. Therefore, it is the actual situation that the value is fixed empirically or set to the greatest common denominator by trial and error regardless of the operating conditions.

Therefore, the temperature distribution of the copper plate changes momentarily due to changes in operating conditions, the initially set temperature Tc and the increase / decrease logic Lv deviate from the proper range, and the molten metal level becomes unstable, resulting in poor quality of the slab. In severe cases, a breakout is a problem. In particular, in recent years, casting has been speeded up to improve productivity, but the higher the speed, the more difficult automatic control becomes and the serious problem is that the molten metal level is not stable. Moreover, in order to make the solidified state of the slab in the mold sound, there are increasing cases in which a multi-stage taper mold or a parabolic taper mold is adopted instead of the conventional single taper mold. In the case of a single taper mold, the mold taper amount relative to the solidification shrinkage amount of the slab is maintained constant even if the molten metal level moves up or down, but in the case of a multistage taper mold or a parabolic taper mold The amount of mold taper with respect to the amount of solidification shrinkage changes as the molten metal level rises and falls. Therefore, in the multi-stage taper mold and the parabolic taper mold that have been frequently used in recent years, stable maintenance of the molten metal surface level becomes more important.

Therefore, an object of the present invention is to provide a low-cost control method capable of stably maintaining the molten metal level in the mold even if the operating conditions change.

[0019]

The gist of the present invention which can solve the above problems is as follows. A first aspect of the invention is a deviation between a temperature measurement value of a mold and a preset control target temperature during casting in a continuous casting machine that continuously draws a slab from the mold while continuously supplying molten metal to the mold. A method for controlling the molten metal level in the mold by increasing or decreasing the slab drawing speed or the molten metal supply amount by the amount of increase or decrease calculated from the deviation and a preset control factor. Sample data is sequentially extracted from a population consisting of time-series data of temperature measurement values, basic statistics are calculated for this sample data, and / or a histogram is created, and selected from these basic statistics. Hot water in a mold in a continuous casting machine, characterized in that the control target temperature and / or the control factor is modified so that at least one index and / or histogram approaches a proper pattern. A level control method.

A second aspect of the present invention is to perform casting in a continuous casting machine in which molten metal is continuously supplied to a mold while continuously extracting a slab from the mold. In the method for controlling, the sample data is sequentially extracted from a population consisting of time series data of temperature measurement values of the template, basic statistics are calculated for this sample data, and / or a histogram is created, and the basic statistics are calculated. A method for controlling a level of molten metal in a mold in a continuous casting machine, which comprises detecting an abnormality in a mold based on at least one index and / or a histogram respectively selected from the amounts.

[0021]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in more detail below.

A continuous casting machine for carrying out the molten metal level control method of the present invention is, for example, a billet continuous casting machine in which the amount of molten metal injected from the tundish into the mold is constant, and it is in a copper plate of the mold. A thermocouple is embedded as a temperature sensor at a predetermined depth from the inner surface of the copper plate. In order to carry out the control method of the present invention, as will be described later, only one temperature sensor may be used, or a plurality of temperature sensors provided at different levels may be used.

First, the case where the control method of the present invention is implemented by using one temperature sensor will be described.

A thermocouple (temperature sensor) is used as an initial value in advance.
The control target temperature (set value) Tc with respect to the temperature measurement value according to, the coefficient of the steel strip withdrawing speed increasing / decreasing logic Lc, which is a control factor, and the like are determined and input to the control computer. Temperature value Ts measured by thermocouple (temperature sensor) after starting casting
The deviation ΔT (= Ts−Tc) between the control target temperature Tc and the control target temperature Tc is sequentially calculated. Next, the slab drawing speed increase / decrease amount Δv is calculated from the deviation ΔT and the increase / decrease logic Lc. Then, the slab drawing speed (that is, the casting speed) v is increased or decreased by this Δv (v → v + Δv). As a result, the deviation ΔT approaches zero (that is, the temperature measurement value Ts approaches the control target temperature Tc), so that the molten metal level in the mold is controlled to be constant.

Further, the temperature measurement value Ts is continuously accumulated in a control computer or the like, time series data of this temperature measurement value Ts is statistically processed, and the control target temperature (setting) is set based on the statistical data. By correcting the value Tc, it is possible to further improve the accuracy of controlling the molten metal level in the mold. Hereinafter, a method of statistical processing and a method of correcting the control target temperature (set value) Tc based on the statistical data will be described in detail.

FIG. 1 is a diagram showing a relationship between a copper plate temperature distribution at a thermocouple (temperature sensor) embedded depth position and a set temperature (control target temperature) Tc when casting a steel type A. In addition, FIG. 2 shows 0.5 from a population consisting of time series data of temperature measurement values continuously measured by a thermocouple (temperature sensor) during casting of steel type A and accumulated in a control computer.
Extract sample data for 3 minutes every second (sampling)
However, this sample data is a histogram.
In this case, since the set temperature Tc was proper, the position (temperature) of the apex of the obtained histogram has the set temperature Tc.
It is almost the same as c and approximates a normal distribution curve.

Next, FIG. 3 shows a case where the steel type A is cast under the same operating conditions as in FIG. 1, but when only the set temperature Tc is set higher than the appropriate value, the copper plate temperature distribution and the set temperature ( It is a figure which shows the relationship with control target temperature) Tc. By setting the set temperature Tc higher than that in the case of FIG. 1, the entire temperature distribution of the copper plate moves upward, and accordingly, the molten metal level becomes higher than that in the case of FIG. Then, FIG. 4 shows a histogram of the sample data of the temperature measurement values in the same manner as described above. In this case, the set temperature Tc
Is set to be higher than an appropriate value, the position (temperature) of the apex of the histogram does not match the set temperature Tc, is largely distorted toward the high temperature side, and the height (frequency) of the apex is also reduced.

Next, FIG. 5 shows a case where the steel type A is cast under the same operating conditions as those shown in FIGS. 1 and 3, but at the set temperature Tc.
3 shows the copper plate temperature distribution and the set temperature (control target temperature) Tc when it is set lower than the above-mentioned appropriate value, which is the opposite of FIG.
It is a figure which shows the relationship with. By setting the set temperature Tc lower than that in the case of FIG. 1, the entire temperature distribution of the copper plate moves downward, and accordingly, the molten metal level becomes lower than that in the case of FIG. Then, FIG. 6 shows a histogram of sample data of temperature measurement values in the same manner as in FIGS. In this case, since the set temperature Tc was set too low below the appropriate value, the histogram shows that the position (temperature) of the apex is the set temperature Tc.
It does not correspond to, and is greatly distorted to the low temperature side, and the height (frequency) of the apex is also reduced.

FIG. 7 is a diagram showing the relationship between the copper plate temperature distribution and the set temperature (control target temperature) Tc when casting different steel grades, steel grade A and steel grade B, respectively.
The set temperature Tc was the same for both steel types. Further, FIG. 8 shows a histogram of sample data of temperature measurement values of each of the steel types A and B in the same method as in FIGS. In this case, since the set temperature Tc is set to an appropriate value for the steel type A, the histogram of the steel type A is such that the position (temperature) of its apex is almost the same as the set temperature Tc and is approximate to the normal distribution curve. There is. On the other hand, steel type B
However, since the set temperature Tc was out of the proper value,
In the histogram of steel type B, the position (temperature) of its apex does not match the set temperature Tc, it is distorted to the high temperature side, and the height (frequency) of the apex is also reduced.

As is clear from the above, the proper set temperature Tc differs depending on the operating conditions, and the set temperature Tc
When is appropriate, the histogram is close to a normal distribution curve, but the histogram is distorted as the set temperature Tc deviates from an appropriate value.

[First Embodiment] Therefore, as one embodiment of a method for correcting the control target temperature (set value) Tc by using one temperature sensor, for example, in a cab of a continuous casting machine, a control computer screen is displayed. The above histogram is displayed above so that it can be visually monitored. The histogram display should be automatically updated at regular intervals. When the operator looks at the histogram and determines that the deviation from the normal distribution curve is unacceptable, the set temperature Tc is adjusted so as to approach the normal distribution curve that is the proper pattern. After the adjustment, the change in the shape of the histogram is observed, and Tc is readjusted if necessary. In this way, since the fine adjustment can be performed while observing the change in the shape of the histogram, the optimum set temperature Tc can be easily obtained.
It is possible to maintain a stable surface level.

On the other hand, conventionally, since the operator's experience and intuition of judging whether the set temperature Tc is appropriate or not is judged from the fluctuation state of the molten metal level, the experience and skill may differ between operators. There was variation in the adjustment of the set temperature Tc, and the level of the molten metal could not be stabilized.

Further, even when the proper set temperature Tc is set at the initial stage, it may become improper due to variations in casting conditions (for example, variations in molten steel temperature, variations in casting speed). In this case, since the shape of the above-mentioned histogram is deformed and it can be easily determined that the initially set temperature Tc is no longer appropriate, adjustment work of the set temperature Tc is performed by the same method as described above. be able to.

In place of the histogram, basic statistics (sample number, total, average, median,
Mode, maximum, minimum, range, variance, standard deviation, standard error, skewness, kurtosis, etc.), and at least one index of this basic statistic, for example, the maximum representing the position of the vertex of the histogram. The same control can be performed by selecting numerical values such as the frequency and the skewness indicating the degree of distortion of the histogram, displaying these numerical values in a graph over time, and monitoring. It is also preferable to use these time-lapse graphs and the histogram together, because the operator's judgment can be more objective and accurate.

Further, in the above description, the set temperature Ts
However, even if the set temperature Ts is appropriate, if the increase / decrease logic Lc is not appropriate, the level of the molten metal may hunt and the molten metal surface may not be stably maintained. In this case, a change appears, for example, that the histogram becomes wider than in the normal case. Therefore, in such a case, the increase / decrease logic Lc can be modified.

As described above, the present invention contributes to the improvement of the accuracy of the molten metal level control when the normal casting state is maintained. However, when an abnormality occurs in the mold as described below. Even in the case of the above, abnormalities can be detected early and accurately, and more effective level control of the molten metal surface becomes possible.

That is, if an abnormal phenomenon such as seizure between the slab and the mold, a crisping phenomenon due to an increase in frictional resistance between the slab and the mold, or an abnormality in molten metal inflow occurs in the mold, FIGS. 12 and 13 are shown. As described above, the normal shape of the histogram is distorted. FIG. 12 is an example of a histogram when crimping occurs in a cast piece during casting, and FIG. 13 is an example of a histogram when a molten metal surface is disturbed due to an abnormality in molten metal inflow. Therefore, by monitoring the change of at least one index selected by the operator from the histogram or the basic statistic, it becomes possible to detect the abnormality early and accurately, and there is also an effect that the breakout can be predicted easily and surely.

[Second Embodiment] Control target temperature (set value)
As another embodiment of the method for correcting Tc, a plurality of temperature sensors may be used. 9 to 11 are views showing a case where two temperature sensors A and B are used together. As shown in FIG. 9, while controlling the temperature sensor A as a main sensor to a set temperature (control target temperature) Tc, while referring to the temperature sensor B as a sub sensor, the molten metal level control is performed. 10 and 11
In addition, the temperature measurement values Ts by the temperature sensors A and B,
The histogram of the sample data of Ts' is shown. FIG. 10 shows the case where the set temperature Tc is appropriate, and the histograms of both the temperature sensors A and B are close to the normal distribution curve.
FIG. 11 shows the case where the set temperature is higher than the proper value, and the histogram of the temperature sensor A is distorted to the high temperature side and the height of the apex is lowered. Further, the histogram of the temperature sensor B moves to the high temperature side, and the height of the apex rises to form a sharp curve.

Therefore, at least one index (for example, mode value, skewness, kurtosis, etc.) selected from a histogram and basic statistics obtained by statistically processing sampled data of temperature measurement values by a plurality of thermocouples is used. By displaying it, the set value Tc can be corrected more accurately.

(Third Embodiment) Instead of (or in addition to) displaying at least one index selected from a histogram and basic statistics, the at least one index (for example, mode value, skewness, kurtosis, etc.) ) Is selected, and a system for automatically correcting the set value Tc and the increase / decrease logic Lc for making this index close to an appropriate value is constructed, and by using this system, more accurate level control can be performed. it can.

Furthermore, an automatic alarm system is constructed using the above-mentioned index, and by using this automatic alarm system, the above-mentioned various abnormalities in the mold can be automatically detected, and the alarm device and the automatic operation stop It is possible to make the molten metal level control more effective by linking with the device.

On the other hand, the conventional breakout predicting device requires a large number of temperature sensors and complicated logic and is restraining (due to seizure between the slab and the mold).
Only the breakout could be predicted, and the disturbance of the molten metal surface due to the abnormal molten metal inflow could not be detected.

Since many billet continuous casting machines have already adopted a copper plate temperature detecting type molten metal level control system, these continuous casting machines have
Since the existing temperature sensor can be used as it is, the control method of the present invention can be introduced at low cost, and highly accurate level control of the molten metal can be achieved easily and inexpensively.

That is, in the conventional copper plate temperature detection type molten metal level control, usually 3 to 4 temperature sensors are embedded at different levels in the mold copper plate, and the upper temperature sensor causes an abnormal rise in the molten metal level. The detection and middle temperature sensors are used for controlling the molten metal level, and the lower temperature sensors are used for detecting the abnormal lowering of the molten metal level. Since it is applied to the molten metal level control method of the present invention, one or a plurality of temperature sensors can be arbitrarily selected and used from these temperature sensors. For example, in order to carry out the first embodiment, only one middle temperature sensor may be diverted, or in order to carry out the second embodiment, the middle temperature sensor is used as the main sensor and the lower temperature is used. The sensor may be used as a sub sensor. Of course, apart from these conventional temperature sensors,
A dedicated temperature sensor for implementing the control method of the present invention may be provided.

In the above embodiment, only the example in which the level of the molten metal is applied to the level control of the molten metal by adjusting the withdrawal speed of the slab is described, but the present invention is not limited to this. The same can be applied to the case where the molten metal inflow amount is adjusted at a constant cast slab drawing speed.

In the above embodiment, only the copper plate was used as the material of the mold and the steel was used as the material to be cast. However, the present invention is not limited to these, and various metals are appropriately selected according to the purpose. It is possible.

[0047]

As described above, according to the present invention,
Even if the type of metal to be cast or the casting conditions change, it is possible to easily and easily achieve highly accurate control of the molten metal level in the mold.

[Brief description of drawings]

FIG. 1 Copper plate temperature distribution and set temperature (control target temperature) Tc
It is a figure which shows the relationship with (when Tc is appropriate).

FIG. 2 is a histogram of sample data extracted from a population consisting of time series data of temperature measurement values in the case of FIG.

FIG. 3 Copper plate temperature distribution and set temperature (control target temperature) Tc
It is a figure which shows the relationship with (when Tc is higher than an appropriate value).

FIG. 4 is a histogram of sample data extracted from a population consisting of time series data of temperature measurement values in the case of FIG.

FIG. 5: Copper plate temperature distribution and set temperature (control target temperature) Tc
It is a figure which shows the relationship with (when Tc is lower than an appropriate value).

FIG. 6 is a histogram of sample data extracted from a population consisting of time series data of temperature measurement values in the case of FIG.

FIG. 7 is a diagram showing a relationship between a copper plate temperature distribution and a set temperature (control target temperature) Tc when casting different steel types A and B, respectively.

FIG. 8 is a diagram in which histograms of sample data extracted from each population of time series data of temperature measurement values at the time of casting of each steel type are overlapped in the case of FIG. 7.

FIG. 9 is a diagram showing a relationship between a copper plate temperature distribution and a set temperature (control target temperature) Tc when two temperature sensors are used.

FIG. 10 is a diagram in which histograms of sample data extracted from each population consisting of time-series data of temperature measurement values by each temperature sensor in the case of FIG. 9 are superimposed (T).
If c is correct).

FIG. 11 is a diagram in which histograms of respective sample data extracted from respective populations composed of time-series data of temperature measurement values obtained by respective temperature sensors in the case of FIG. 9 are superimposed (T).
If c is higher than the proper value).

FIG. 12 is a diagram showing an example of a histogram when a slab being cast is crimped.

FIG. 13 is a diagram showing an example of a histogram when the molten metal surface is disturbed due to a molten metal inflow abnormality.

FIG. 14 is a diagram schematically showing a copper plate temperature distribution in the flow direction of a cast product at a position of depth D (position of broken line) from the inner surface of the mold copper plate.

   ─────────────────────────────────────────────────── ─── Continued front page    F-term (reference) 4E004 AA08 MB04 NC01 PA07                 5H309 AA01 BB01 BB08 CC09 DD02                       EE05 EE08 FF07 GG04 JJ06

Claims (2)

[Claims] 1. When casting in a continuous casting machine for continuously drawing molten slab from the mold while continuously supplying the molten metal to the mold,
In order to reduce the deviation between the measured temperature value of the mold and the preset control target temperature, the slab drawing speed or the molten metal supply amount is increased or decreased by the increment or decrement amount calculated from the deviation and the preset control factor. A method of controlling the level of molten metal in the mold by extracting the sample data sequentially from the population consisting of time-series data of temperature measurement values of the mold, calculating basic statistics for this sample data, and / or Continuous casting characterized by creating a histogram and modifying the control target temperature and / or the control factor so that at least one index and / or the histogram respectively selected from the basic statistics are approximated to a proper pattern. Level control method in the mold in the machine. 2. When casting in a continuous casting machine for continuously drawing molten slab from the mold while continuously supplying the molten metal to the mold,
In the method of controlling the level of the molten metal in the mold based on the measured temperature value of the mold, the sample data is sequentially extracted from the population consisting of the time series data of the measured temperature value of the mold, and the basic statistics are calculated for this sample data. , And / or a histogram is created, and an abnormality in the mold is detected based on at least one index and / or the histogram selected from the basic statistics, respectively, and the molten metal level in the mold in the continuous casting machine. Control method.