JP2001337703A - Method of detection for fluctuating frequency of signal and control for molten metal level in continuous casting, and detecting device for fluctuating frequency of signal and controller for molten metal level in continuous casting - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a detecting method for fluctuating frequency of signals which permits frequency component of signal fluctuation to be detected for a short detecting time with high response, while keeping necessary resolution for frequency. SOLUTION: The data is sampled with a prescribed period and converted digitally (10), and stored successively by a number of power of two to generate a data group. Whenever a data group is generated, the data group is put in time order by a number of power of two to generate the first data line (11) and it is analyzed with the fast Fourier transform to obtain the spectrum of the first fluctuating frequency (12). In the first data line, all the data in data groups other than the latest data group are permuted to zero to generate the second data line (15), and it is analyzed with the fast Fourier transform to obtain the spectrum of the second fluctuating frequency (16). Then the third fluctuating frequency spectrum is obtained through multiplication or geometrical mean of each oscillation of the first and the second fluctuating spectra at the same frequency (14, 14a), to gain the peak frequency of the third fluctuating frequency spectrum.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal for detecting a fluctuating frequency of a signal having a periodic fluctuation by performing fast Fourier transform analysis on data sampled at a predetermined period and digitally converted from the signal having a periodic fluctuation. The method of detecting the variation frequency of the signal, the method of detecting the variation level of the signal, the method of controlling the level of the molten metal in continuous casting, the method of sampling and digitally converting the data having a periodic variation from a signal having a periodic variation. The present invention relates to a signal fluctuation frequency detecting device that detects a fluctuation frequency of a signal having a periodic fluctuation by performing a fast Fourier transform analysis, and a continuous casting level surface control device including the signal fluctuation frequency detecting device. is there.

[0002]

2. Description of the Related Art As shown in FIG. 4, a continuous casting machine is provided with a tundish 21 for temporarily storing molten metal (molten metal) poured from a ladle (not shown). An immersion nozzle 25 for conducting molten metal (molten steel) 29b is connected to the bottom. The lower end of the immersion nozzle 25 is inserted into a cylindrical mold 27. The immersion nozzle 25 is provided with a slide type molten steel injection amount adjusting nozzle 24, and the gate plate of the molten steel injection amount adjusting nozzle 24 is advanced and retracted by the cylinder type nozzle opening adjusting device 22 to thereby adjust the molten steel injection amount adjusting nozzle. 24 is adjusted.

Then, a molten metal 29b is supplied from the immersion nozzle 25 into the mold 27, and the cast piece 29 cooled by contact with the wall of the mold 27 and having a solidified shell 29a formed around is drawn out of the mold, and cooled to the center by water cooling. The cooling is advanced to cast the slab 29 continuously. In such a continuous casting machine, powder is supplied between the inner wall of the mold 27 and the molten metal 29b in order to prevent oxidation of the molten steel surface in the mold 27 and lubricate the mold 27 with the slab 29. The molten metal in the mold 2
If the level of the molten metal surface 9b fluctuates, the powder gets caught in the molten metal 29b, causing subcutaneous defects of the slab 29, so it is important to maintain the molten metal surface level at the target level.

In general, fluctuations in the level of the molten metal in the mold 27 are caused by unevenness in the flow rate of the molten steel in the immersion nozzle 25 with time and fluctuations in the drawing speed of the slab 29 such as the inflow or outflow of the molten steel into the mold 27. The amount of disturbance is due. Therefore, a level detector 30 for measuring the level of the molten metal facing the molten metal 29b in the mold 27 is installed, and the level controller 23 determines the deviation between the target level and the level of the molten metal measured by the level detector 30. In order to make the obtained deviation zero, the amount of change in the opening of the molten steel injection amount adjusting nozzle 24 is calculated by PID calculation, and given to the nozzle opening adjusting device 22 as an opening instruction to adjust the molten steel injection amount. The gate plate of the nozzle 24 is moved forward and backward, and the opening degree of the molten steel injection amount adjusting nozzle 24 is adjusted.

[0005] In the continuous casting machine as described above, the unsolidified slab bulging generated between the rolls in the continuous casting machine, the roll bending of the guide roll or the pinch roll, and the unbalanced eccentricity caused by the roll eccentricity occur. Periodic expansion and contraction of the solidified slab causes fluctuations in the volume of the molten steel, causing periodic fluctuations in the molten metal level.

However, as described above, the level detector 3
In the mold level control method in which the feedback control is performed based on the detection result of 0, it is difficult to prevent the above-described periodic mold level fluctuation. When the fluctuation of the periodic molten metal level increases, the quality of the cast slab deteriorates and breakout easily occurs. JP-A-10-314
No. 911 proposes a mold level control device in a continuous casting machine using an FFT (Fast Fourier Transform) analyzer for mold level control in a mold.

[0007]

In principle, FFT (Fast Fourier Transform) has a problem that a large number of data points is required to secure frequency resolution. For example, when the sampling period is 0.1 second, the relationship between the frequency resolution Δf and the number of sampling points N (the FFT needs to be a power of 2) is expressed by equation (1). Δf = ((1 / 0.1) / 2) / N (1) Therefore, in order to satisfy Δf <0.01 Hz, it is necessary that
Two points of data are required, and the calculation time = detection time is 5
1.2 seconds is required. If the variable frequency component changes in this section, it is detected as if both the frequency components before and after the change exist simultaneously.

On the other hand, if the number of sampling points is reduced to avoid such a problem, for example, if N = 128, then Δf ≒ 0.04 Hz from the equation (1), and if there are a plurality of fluctuating frequency components in this range. If you can't tell. The present invention has been made in view of the circumstances described above, and in the first and second inventions, while ensuring necessary frequency resolution, a short detection time, good responsiveness, and fluctuation of a signal having a periodic fluctuation. It is an object of the present invention to provide a method for detecting a fluctuating frequency of a signal which can detect the frequency component of the signal. Third
It is an object of the present invention to provide a continuous casting level control method for controlling a continuous casting level using the signal fluctuation frequency detecting method according to the first and second aspects of the present invention.

According to the fourth and fifth aspects of the present invention, there is provided a signal fluctuation frequency detecting apparatus capable of detecting a frequency component of a signal fluctuation having a periodic fluctuation with a short detection time and good response while securing necessary frequency resolution. The purpose is to: In a sixth aspect of the present invention, it is an object of the present invention to provide a continuous casting level control apparatus for controlling a continuous casting level using the signal fluctuation frequency detecting apparatus according to the fourth and fifth aspects.

[0010]

According to a first aspect of the present invention, there is provided a method for detecting a fluctuating frequency of a signal, comprising the steps of:
Each time a data group is created, a first data string is created by arranging the data groups in a time series of a power of 2 each time one data group is created. Fourier transform analysis is performed to obtain a first variable frequency spectrum, and a second data string is created by replacing all data of the data group except the latest data group in the first data string with 0. Fast Fourier transform analysis of the data sequence to obtain a second variable frequency spectrum, and multiplication or geometric averaging of the respective amplitudes of the obtained first and second variable frequency spectra at the same frequency to obtain a third variable frequency spectrum, The fluctuation frequency of the signal having the periodic fluctuation is obtained by obtaining the peak frequency of the obtained third fluctuation frequency spectrum.

In the signal fluctuation frequency detecting method according to the second invention, a band including the fluctuation frequency is estimated in advance, and each amplitude of the third fluctuation frequency spectrum included in the estimated band is multiplied by a predetermined weighting factor. It is characterized by matching.

A third aspect of the present invention is a continuous casting machine for continuously drawing a slab from a mold while supplying molten metal into the mold from a nozzle whose opening is controlled. In the method for controlling the level of the molten metal in the mold in a time series, and controlling the opening degree of the nozzle in order to suppress the fluctuation of the level of the molten metal, the level of the molten metal is controlled in time series. Frequency of the signal obtained by the measurement
A signal fluctuation frequency detection method according to claim 1 or 2, and the opening degree of the nozzle is controlled based on the detected fluctuation frequency.

According to a fourth aspect of the present invention, there is provided an apparatus for detecting a fluctuating frequency of a signal, comprising: means for sampling a signal having a periodic fluctuation at a predetermined period; means for digitally converting the signal sampled by the means into data; Means for sequentially storing power-of-two numbers of digitally converted data to create a data group; and each time the means creates one data group, the data groups are arranged in time-series power-of-two numbers to form first data. Means for creating a sequence, first analysis means for performing fast Fourier transform analysis on the first data sequence created by the means to obtain a first variable frequency spectrum, and excluding the latest data group from the first data sequence Means for creating a second data string in which all data of the data group is replaced with 0, and fast Fourier transform analysis of the second data string created by the means to obtain a second variable frequency spectrum And Mel second calculating means, by the first analyzing means first variation frequency spectrum and a second analyzing means for multiplying or geometric mean respective amplitude at the same frequency of the second variation frequency spectrum obtained was determined,
It is characterized by comprising means for obtaining a third fluctuation frequency spectrum, and means for obtaining a fluctuation frequency of the signal having the periodic fluctuation by obtaining a peak frequency of the third fluctuation frequency spectrum obtained by the means.

In the signal fluctuation frequency detecting method according to the first invention and the signal fluctuation frequency detecting apparatus according to the fourth invention, the sampling means samples the signal having the periodic fluctuation at a predetermined cycle to obtain data. However, the sampled signal is converted into data by digital conversion. The means for creating a data group creates the data group by sequentially storing the digitally converted data in a power-of-two number. The means for creating the first data string includes: for each time the means for creating the data group creates one data group,
The first data string is created by chronologically arranging the powers of. The first analyzing means obtains a first fluctuating frequency spectrum by performing a fast Fourier transform analysis on the first data sequence.

The means for creating the second data string creates a second data string in which all the data of the data group excluding the latest data group in the first data string are replaced with 0, and the second analyzing means comprises: The second data sequence is analyzed by fast Fourier transform to obtain a second
Find the fluctuating frequency spectrum. The means for obtaining the third variable frequency spectrum is obtained by multiplying or geometrically averaging respective amplitudes of the first variable frequency spectrum obtained by the first analyzing means and the second variable frequency spectrum obtained by the second analyzing means at the same frequency. A third variable frequency spectrum is obtained.

The means for obtaining the fluctuation frequency of the signal obtains the peak frequency of the third fluctuation frequency spectrum obtained by the means for obtaining the third fluctuation frequency spectrum, thereby obtaining the fluctuation frequency of the signal having the periodic fluctuation. As a result, a signal fluctuation frequency detection method and a signal fluctuation frequency detection device capable of detecting a frequency component of a signal fluctuation having a periodic fluctuation with short response time and good response while securing necessary frequency resolution are realized. I can do it. In particular, when a plurality of fluctuating frequency components are present and a beat occurs in the waveform, it is possible to prevent erroneous detection occurring due to the disappearance of the fluctuating component in a section of the beat. .

According to a fifth aspect of the present invention, the apparatus for detecting a variable frequency of a signal further comprises means for multiplying each amplitude of the three variable frequency spectrums included in an externally designated band by a predetermined weighting factor. I do.

In the signal fluctuation frequency detecting method according to the second invention and the signal fluctuation frequency detecting apparatus according to the fifth invention, the multiplying means may be configured such that each of the amplitudes of the third fluctuation frequency spectrum included in the band designated from the outside is Is multiplied by a predetermined weighting factor, so that if a band in which a fluctuating frequency occurs can be estimated in advance, a fluctuating frequency detecting method and a fluctuating signal detecting apparatus for a signal that improve the detection sensitivity in the speculated and specified band Can be realized.

According to a sixth aspect of the present invention, there is provided a continuous casting level controller for continuously casting a cast metal piece from a mold while supplying molten metal into the mold from a nozzle whose opening is controlled. 5. A level controller for continuous casting, which measures the level of the molten metal in the mold in a time series and controls the opening of the nozzle to suppress the fluctuation of the level of the molten metal.
Or a device for detecting a fluctuating frequency of a signal described in 5,
The fluctuation frequency of the signal obtained by measuring the bath level in time series is detected by a fluctuation frequency detection device of the signal,
The opening degree of the nozzle is controlled based on the detected fluctuation frequency.

In the continuous casting level control method according to the third invention and the continuous casting level control apparatus according to the sixth aspect, the molten metal is supplied from the nozzle whose opening is controlled into the mold while the molten metal is supplied into the mold. With a continuous casting machine that continuously pulls out a slab, the level of the molten metal in the mold is measured in time series, and the opening of the nozzle is controlled to suppress the fluctuation of the molten metal level. The fluctuation frequency of the signal obtained by measuring the surface level in time series,
The fluctuation frequency detection device for a signal according to claim 4 or 5 detects the fluctuation frequency and controls the opening degree of the nozzle based on the detected fluctuation frequency.

Thus, a method for controlling the level of a continuous casting metal level and a method of controlling the level of a continuous casting capable of detecting the frequency component of the fluctuation of the level signal of the continuous casting while ensuring the necessary frequency resolution and with a short response time and with good responsiveness. A level control device can be realized. In particular, when a plurality of fluctuating frequency components are present in the level signal and a beat is generated in the waveform, an erroneous detection that occurs due to the disappearance of the fluctuating component in the section of the beat is apparently detected. Can be prevented.

[0022]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below.
This will be described with reference to the drawings showing the above. Embodiment 1 FIG. FIG. 1 is a block diagram showing a main configuration of a signal fluctuation frequency detection method and a signal fluctuation frequency detection apparatus according to an embodiment of the present invention. The signal fluctuation frequency detecting apparatus amplifies a given signal into a predetermined voltage signal, converts the signal into a predetermined voltage signal, and outputs the amplified signal.
(Fast Fourier Transform) An operation unit 2 for analysis, a display monitor 3 for displaying the result of the FFT analysis by the operation unit 2, and a result of transmitting the result of the FFT analysis by the operation unit 2 to a higher-level control system 5 used for control. And a communication control unit 4.

FIG. 2 is a block diagram showing a configuration example of a main part of the arithmetic unit 2. The arithmetic unit 2 samples a given voltage signal (analog signal) at a predetermined cycle, converts the sampled signal into a digital signal, and outputs the converted signal as data.
A moving block table 11 (means for creating a first data string) for creating and storing a data frame (first data string) based on the data output from the sampler 10; and a data frame (means for creating a first data string) based on the data output from the sampler 10. Detection block table 15 for creating the second data string)
(Means for creating a second data string).

Each time the moving block table 11 and the detection block table 15 create a data frame, the arithmetic unit 2 performs an FFT analysis of the created data frame by a first FFT unit 12 (first analysis means) and A second FFT unit 16 (second analysis means);
A first spectrum table 13 and a second spectrum table 17 for respectively storing a first variable frequency spectrum and a second variable frequency spectrum, which are the results of fast Fourier transform performed by the FT section 12 and the second FFT section 16, respectively.
And a multiplying means 14 for multiplying each amplitude (intensity) at the same frequency of the first variable frequency spectrum and the second variable frequency spectrum stored in the first spectrum table 13 and the second spectrum table 17, respectively. Geometric averaging means 14a for geometrically averaging the amplitudes and switching means 14b for switching whether or not to geometrically average the multiplied amplitudes
And

The arithmetic unit 2 also includes a band (f _{a to} f _b ) including a fluctuating frequency, which is presumed and specified externally, as shown in FIG. 3, and each amplitude of the frequency included in the band. Weighting table 19 for storing a weighting coefficient (here, 2) for multiplying (intensity);
Alternatively, the weighting coefficient stored in the table 19 is added to each amplitude of the frequency included in the band stored in the weighting table 19 of the third variable frequency spectrum created by multiplying or geometrically averaging the amplitudes at the same frequency by the geometric mean means 14a. Multiplying means 18 and multiplying means 1
8 has a third spectrum table 20 for storing a third variable frequency spectrum multiplied by the weighting coefficient.

The operation of the thus-configured signal fluctuation frequency detecting device will be described below. The signal conversion amplifier 1 amplifies and converts a given signal into a predetermined voltage signal and outputs the signal. The sampler 10 of the arithmetic unit 2 samples the voltage signal output from the signal conversion amplifier 1 at a cycle of, for example, 0.1 second, digitally converts the voltage signal, and outputs the data. The moving block table 11 stores, for example, 64 (2 ⁶ ) pieces of data output from the sampler 10 in chronological order, creates a block X ₀ (i) (one data group), and goes to one block X ₀ ( Each time i) is created, the storage position is shifted in block units, for example, 8 (2 ³ ) blocks X ₀ (i)
A data frame (first data string) in which the blocks X ₇ (i) are arranged in time series is created.

On the other hand, the detection block table 15 stores the data output by the sampler 10 in a time series of 64 (2 ⁶ ).
The block X ₀ (i) (one data group) is created and stored, and the block X ₀ (i) is updated every time one block X ₀ (i) is created. For example, eight blocks X ₀ (i) are created. ₀ (i) ~ block X ₇ out of (i), and replacing all the data included in the update block X ₀ (i) 7 blocks except X ₁ (i) ~ block X ₇ (i) 0 Then, a data frame (second data sequence) in which the blocks X ₀ (i) to X ₇ (i) are arranged in time series is created.

Each time the moving block table 11 creates a data frame, the first FFT unit 12 performs a fast Fourier transform analysis on the created data frame. As a result, the fast Fourier transform analysis is performed for each of the 64 data, so that the frequency resolution can be improved. 1st F
The first, which is the result of the fast Fourier transform analysis performed by the FT unit 12,
The variable frequency spectrum is stored in the first spectrum table 13
Is stored. Each time the detection block table 15 creates a data frame, the second FFT unit 16 performs a fast Fourier transform analysis on the created data frame. As a result, the analysis time of the fast Fourier transform analysis can be reduced while maintaining the frequency resolution at which the fast Fourier transform analysis is performed for every 64 data. The second fluctuating frequency spectrum as a result of the fast Fourier transform analysis performed by the second FFT unit 16 is stored in the second spectrum table 17.

The multiplying means 14 or the geometric averaging means 14a
The first spectrum table 13 and the second spectrum table 17 respectively multiply or geometrically average each amplitude (intensity) at the same frequency of the first variable frequency spectrum and the second variable frequency spectrum which are stored, and newly generate the third frequency spectrum. Create a fluctuating frequency spectrum. Thus, when a signal has a plurality of fluctuating frequency components and a beat is generated in the waveform, it is possible to prevent an erroneous detection occurring due to the disappearance of the fluctuating component in a section of the beat. I can do it.

The multiplying means 18 stores the table 19 for each amplitude (intensity) of the frequency included in the band stored in the weighting table 19 of the third variable frequency spectrum created by the multiplying means 14 or the geometric mean means 14a. The third variable frequency spectrum, which is multiplied by the weighting factor and multiplied by the weighting factor by the multiplying means 18, is stored in the third spectrum table 20. The third variable frequency spectrum stored in the third spectrum table 20 is displayed on the display monitor 3 as necessary, and is transmitted to the higher-level control system 5 through the communication control unit 4 to be used for controlling a process. You.

Embodiment 2 FIG. 4 is a block diagram showing a main part configuration of an embodiment of a continuous casting level control apparatus and a continuous casting level control apparatus according to the present invention. In this continuous casting level controller, a tundish 21 for temporarily storing a molten metal (molten metal) poured from a ladle (not shown) is provided. ) Immersion nozzle 25 for conducting 29b
Are connected. The lower end of the immersion nozzle 25 is a cylindrical mold 2
7 is charged. The immersion nozzle 25 is provided with a slide type molten steel injection amount adjusting nozzle 24, and the gate plate of the molten steel injection amount adjusting nozzle 24 is advanced and retracted by the cylinder type nozzle opening adjusting device 22 to thereby adjust the molten steel injection amount adjusting nozzle. 24 is adjusted.

From the immersion nozzle 25, the molten metal 29b is supplied into the mold 27, and the cast piece 29 cooled by contact with the wall of the mold 27 and having a solidified shell 29a formed around it is drawn out of the mold 27, and cooled by water cooling. The cooling is advanced to continuously cast the slab 29. The molten metal 2 in the mold 27
9b, a level detector 30 for measuring the level of the level 26 is installed. The level controller 23 calculates a deviation between the target level and the level measured by the level detector 30, and determines the deviation. Is calculated by PID calculation to give the nozzle opening adjusting device 22 as an opening command, and the gate plate of the molten steel injection amount adjusting nozzle 24 is set to zero. The opening and closing of the molten steel injection amount adjusting nozzle 24 is adjusted.

The level detector 30 detects a level change of the molten metal level 26 in the mold 27 when the periodic molten metal level level fluctuation occurs. The level signal of the mold level in the mold detected by the level detector 30 is supplied to a frequency detection arithmetic unit 31, which detects the frequency of the level change of the mold level in the mold, and Give to 23. The level controller 23 determines the frequency of the given level fluctuation,
Control is performed so as to suppress fluctuations in the level of the molten metal in the mold.

The frequency detection arithmetic unit 31 is the same as that of the first embodiment.
Built-in signal fluctuation frequency detection device described in
In the operation, the signal conversion amplifier 1 (FIG. 1) amplifies and converts the level signal of the molten metal level in the mold detected by the level detector 30 (FIG. 4) into a predetermined voltage signal and outputs the signal. The sampler 10 (FIG. 2) of the arithmetic unit 2 (FIG. 1) samples the voltage signal output from the signal conversion amplifier 1 at a cycle of, for example, 0.1 second, digitally converts the voltage signal, and outputs the data.

The moving block table 11 (FIG. 2) stores the data output from the sampler 10 in chronological order, for example, 64 times.
Each time (2 ⁶ ) blocks are stored, a block X ₀ (i) (data group) is created, and each time one block X ₀ (i) is created, the storage position is shifted in units of blocks, for example, 8
A data frame (first data string) in which (2 ³ ) blocks X ₀ (i) to X ₇ (i) are arranged in time series is created.

On the other hand, the detection block table 15 (FIG. 2)
Creates a block X ₀ (i) (data group) by chronologically storing, for example, 64 (2 ⁶ ) pieces of data output from the sampler 10, and creates one block X ₀ (i). Each time the block X ₀ (i) is updated, for example, of the 8 (2 ³ ) blocks X ₀ (i) to X ₇ (i), 7 blocks excluding the updated block X ₀ (i) A data frame (second data) in which the data included in X ₁ (i) to block X ₇ (i) are all replaced with 0, and blocks X ₀ (i) to block X ₇ (i) are arranged in time series Column).

Each time the moving block table 11 creates a data frame, the first FFT unit 12 performs a fast Fourier transform analysis on the created data frame. Thus, the fast Fourier transform analysis is performed for every 64 data, so that the resolution of the frequency of the level fluctuation of the molten metal level in the mold can be improved. The first fluctuating frequency spectrum, which is the result of the fast Fourier transform analysis performed by the first FFT unit 12, is:
It is stored in the first spectrum table. FIG. 5 is an example of the first variable frequency spectrum stored in the first spectrum table. Sampling cycle 0.1 second, number of data 6
4 points × 8 blocks = 512 points, and the resolution is sufficient, but the detection time is long at 0.1 seconds × 512 = 51.2 seconds, and the response is slow.

Each time the detection block table 15 creates a data frame, the second FFT unit 16 performs a fast Fourier transform analysis on the created data frame. As a result, it is possible to reduce the analysis time of the fast Fourier transform analysis of the level fluctuation of the molten metal level in the mold while maintaining the frequency resolution at which the fast Fourier transform analysis is performed for every 64 data. The second fluctuating frequency spectrum as a result of the fast Fourier transform analysis performed by the second FFT unit 16 is stored in the second spectrum table. FIG. 6 is an example of the second variable frequency spectrum stored in the second spectrum table.
The sampling period is 0.1 second, the number of data is 64 points × 8 blocks (the data value of 7 blocks is 0) = 512 points, but in a short section of 0.2 Hz to 0.4 Hz, 2 points
It has been detected that two fluctuation components have occurred.

The multiplying means 14 or the geometric mean means 14a
The first spectrum table 13 and the second spectrum table 17 respectively multiply or geometrically average each amplitude (intensity) at the same frequency of the first variable frequency spectrum and the second variable frequency spectrum which are stored, and newly generate the third frequency spectrum. Create a fluctuating frequency spectrum. Thereby, when a plurality of fluctuating frequency components are present in the level signal of the molten metal level in the mold, and a groan occurs in the waveform, in the section of the groan node,
Apparently, erroneous detection caused by the disappearance of the fluctuation component can be prevented.

FIG. 7 shows the multiplication means 14 or the geometric mean averaging means 1.
4a is an example of a third variable frequency spectrum created by multiplying or geometrically averaging the amplitudes (intensities) of the first variable frequency spectrum (moving block frame) and the second variable frequency spectrum (detection block frame) at the same frequency. It is. In the section of the fluctuating beat, even in a section where the characteristic frequency is not conspicuous in the second fluctuating frequency spectrum, the sensitivity is increased by multiplication or geometric averaging,
Characteristic frequencies can be emphasized.

The multiplying means 18 stores the table 19 in each amplitude (intensity) of the frequency included in the band stored in the weighting table 19 of the third variable frequency spectrum created by the multiplying means 14 or the geometric mean means 14a. The third variable frequency spectrum, which is multiplied by the weighting factor and multiplied by the weighting factor by the multiplying means 18, is stored in the third spectrum table 20.

FIG. 8 shows an example of weighting characteristics of the fluctuation frequency of the molten metal level in continuous casting. When the casting speed is 3.0 m / min, the frequency component of the periodic molten metal level fluctuation when the roll interval is two types of guide rolls of 165 mm and 200 mm, and the roll diameter of the pinch roll is 160 m
The characteristic when weighting is applied to the periodic component level fluctuation frequency component when the roll is eccentric with mφ. Predicting the fluctuation frequency of the metal surface level by the arrangement of guide rolls (not shown) that support and reduce the slab 29 (FIG. 4) of the continuous casting machine and the casting speed (the speed at which the slab 29 is pulled out from the mold 27). Therefore, the weighting coefficient in the band including the fluctuating frequency is set to be larger than 1 to increase the detection sensitivity.

The third fluctuating frequency spectrum stored in the third spectrum table 20 is displayed on the display monitor 3 as required, and is transmitted through the communication control unit 4 to the level controller 23 (the higher-level control system 5). This is transmitted to FIG. 4) and is used for control for suppressing fluctuations in the metal surface level in continuous casting. FIG. 9 is an example of a third variable frequency spectrum obtained by performing a weighting correction operation on the third variable frequency spectrum shown in FIG. A fluctuating frequency existing between 0.2 Hz and 0.4 Hz is detected with high sensitivity.

[0044]

According to the signal fluctuation frequency detecting method of the first invention and the signal fluctuation frequency detecting apparatus of the fourth invention, the required frequency resolution is ensured, the response time is short, the response is good, and the period is high. Thus, it is possible to realize a signal fluctuation frequency detection method and a signal fluctuation frequency detection device capable of detecting a frequency component of a signal fluctuation having a characteristic fluctuation. In particular, when a plurality of fluctuating frequency components are present and a beat occurs in the waveform, it is possible to prevent erroneous detection occurring due to the disappearance of the fluctuating component in a section of the beat. .

According to the signal fluctuation frequency detecting method according to the second invention and the signal fluctuation frequency detecting apparatus according to the fifth invention, if the band in which the fluctuation frequency occurs can be estimated in advance, the estimated and designated band is determined. Can be realized a signal fluctuation frequency detection method and a signal fluctuation frequency detection device that improve the detection sensitivity in the above.

According to the continuous casting level control method of the third invention and the continuous casting level control apparatus of the sixth aspect, the required frequency resolution is ensured and the response time is short and good. In addition, it is possible to realize a continuous casting level control method and a continuous casting level control apparatus capable of detecting the frequency component of the fluctuation of the continuous casting level signal. In particular, when a plurality of fluctuating frequency components are present in the level signal and a beat is generated in the waveform, an erroneous detection that occurs due to the disappearance of the fluctuating component in the section of the beat is apparently detected. Can be prevented.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a main configuration of a signal fluctuation frequency detection method and a signal fluctuation frequency detection apparatus according to an embodiment of the present invention.

FIG. 2 is a block diagram illustrating a configuration example of a main part of an arithmetic unit.

FIG. 3 is an explanatory diagram for explaining a weight coefficient.

FIG. 4 is a block diagram showing a main part configuration of an embodiment of a continuous casting level control apparatus and a continuous casting level control apparatus according to the present invention.

FIG. 5 is a characteristic diagram showing an example of a first variable frequency spectrum stored in a first spectrum table.

FIG. 6 is a characteristic diagram illustrating an example of a second variable frequency spectrum stored in a second spectrum table.

FIG. 7 is a characteristic diagram illustrating an example of a third variable frequency spectrum.

FIG. 8 is a characteristic diagram showing an example of a weighting characteristic of a fluctuation level of a metal surface level in continuous casting.

FIG. 9 is a characteristic diagram illustrating an example of a third variable frequency spectrum that has been subjected to weighting correction calculation and stored in a third spectrum table.

[Brief description of reference numerals]

REFERENCE SIGNS LIST 1 signal conversion amplifier 2 operation unit 10 sampler 11 moving block table (means for creating first data string) 12 first FFT unit (first analysis means) 14 multiplication means 14 a geometric mean means 15 detection block table (second data string 16) Second FFT unit (second analysis means) 18 Multiplication means 19 Weighting table 20 Third spectrum table 22 Nozzle opening degree adjustment device 23 Level controller 24 Molten steel injection amount adjustment nozzle 25 Immersion nozzle 26 Hot water surface 27 Mold 29a solidified shell 29b molten metal 30 level detector 31 frequency detection arithmetic unit

Claims (6)

[Claims] 1. A data group sampled from a signal having a periodic variation and digitally converted at predetermined intervals and stored in power-of-two order to create a data group. Each time one data group is created, the data group is changed to 2. The first data string is created by arranging the power series in time series, and the created first data string is analyzed by a fast Fourier transform to obtain a first fluctuating frequency spectrum. A second data sequence in which all the data of the data group except for is replaced by 0 is created, and the created second data sequence is analyzed by fast Fourier transform to obtain a second variable frequency spectrum, and the obtained first and second variable frequencies are obtained. A third variable frequency spectrum is obtained by multiplying or geometrically averaging the respective amplitudes at the same frequency of the spectrum, and the peak frequency of the obtained third variable frequency spectrum is obtained. , Thereby obtaining a fluctuating frequency of the signal having the periodic fluctuation. 2. The detection of a fluctuating frequency of a signal according to claim 1, wherein a band including the fluctuating frequency is estimated in advance, and each amplitude of the third fluctuating frequency spectrum included in the estimated band is multiplied by a predetermined weighting coefficient. Method. 3. A continuous casting machine for continuously drawing a slab from a mold while supplying molten metal into the mold from a nozzle whose degree of opening is controlled, and measures a metal surface level in the mold in a time-series manner. In the continuous casting level control method of controlling the opening of the nozzle to suppress the variation of the level, the variation frequency of the signal obtained by measuring the level in time series, Detected by the signal fluctuation frequency detection method according to claim 1 or 2, based on the detected fluctuation frequency,
A method for controlling the level of a molten metal level in continuous casting, comprising controlling the opening of the nozzle. 4. A means for sampling a signal having a periodic variation at a predetermined cycle, a means for converting a signal sampled by the means into digital data, and a means for converting the data digitally converted by the means into a power of 2 Means for sequentially storing and creating a data group, means for creating a first data string by arranging the data group in time series of a power of 2 each time the means creates one data group, First analyzing means for obtaining a first fluctuating frequency spectrum by performing a fast Fourier transform analysis on the first data string obtained, and a second analyzing means for replacing all data of the data group excluding the latest data group in the first data string with zero. Means for creating two data strings, second analyzing means for performing fast Fourier transform analysis on the second data string created by the means to obtain a second variable frequency spectrum, and means for obtaining the first variable means. The first variable frequency spectrum and the second
Means for obtaining a third variable frequency spectrum by multiplying or geometrically averaging the amplitudes of the second variable frequency spectrum at the same frequency obtained by the analyzing means, and calculating the peak frequency of the third variable frequency spectrum obtained by the means. Means for obtaining the fluctuation frequency of the signal having the periodic fluctuation by obtaining the fluctuation frequency of the signal. 5. The apparatus according to claim 4, further comprising: means for multiplying each amplitude of the third variable frequency spectrum included in a band specified from the outside by a predetermined weighting factor. 6. A continuous casting machine for continuously drawing a slab from a mold while supplying molten metal into the mold from a nozzle whose opening degree is controlled, and measures a metal surface level in the mold in a time-series manner. A level controller for continuous casting for controlling the opening degree of the nozzle to suppress the fluctuation of the level of the molten metal, comprising the apparatus for detecting a fluctuation frequency of a signal according to claim 4 or 5, The fluctuation frequency of the signal obtained by measuring the surface level in time series is detected by the fluctuation frequency detection device for the signal, and based on the detected fluctuation frequency, the opening of the nozzle is controlled. Features of continuous casting level control system.