KR20010008193A - Eccentricity diagnosis method and apparatus in rolling mill - Google Patents

Abstract

PURPOSE: A method and an apparatus for diagnosing eccentricity of a rolling mill is provided to prevent producing inferiority and the shutdown of operations due to eccentricity generated in the rolling equipment, and improve the performance of controlling the eccentricity by extracting eccentricity components with time area data instead of spectrum of a frequency area, thereby improving the lifespan of the rolling equipment and the productivity. CONSTITUTION: A method for diagnosing eccentricity of a rolling mill includes the steps of adding outputs obtained by passing sampled rolling mill load data through a plurality of digital band pass filters(31-36) with an adder(37), wherein the digital band pass filters are formed in the structure to be changed in a same direction according to a rotation velocity of a roll, reducing influence of noise by a noise control filter(38), extracting a data relating to eccentricity components generated in the process of rolling from a time area, displaying the extracted eccentricity information in forms of time functions, making index relating to existence and reason of the eccentricity by the sampled rolling mill load data and the extracted eccentricity related data, and comparing a current value of the index with a reference value to determining the existence and the reason of the eccentricity.

Description

Eccentricity diagnosis method and apparatus in rolling mill

The present invention relates to a method and apparatus for diagnosing eccentricity occurring in a rolling mill during rolling. The rolling mill suffers from eccentricity due to the continuous running of the rolls and the center of the rolls that do not coincide with the axis of rotation. Therefore, there is a need for a method and apparatus for automatically diagnosing the occurrence of eccentricity.

In the conventional technique, the eccentricity is diagnosed by obtaining a frequency spectrum of data relating to the rolling load and analyzing the magnitude of the component related to the rotational frequency of the backup roll. However, the frequency spectrum method has a problem that it is very difficult for an operator to determine the cause of the eccentricity caused by looking at the spectrum, and it is difficult to cope with the variation of the rotation frequency because the spectrum continuously changes according to the change of the rotation frequency.

The present invention is to solve the above problems, by showing the waveform form in the time domain rather than the frequency spectrum difficult to decode the diagnosis results of the eccentric, the operator can immediately determine the cause of the eccentric problem caused by looking at the diagnosed waveform An object of the present invention is to provide an eccentric diagnostic method and apparatus.

1 is a block diagram of an eccentric diagnostic apparatus of the present invention

2 is a block diagram of an eccentric component data detection algorithm of FIG.

3A is an explanatory view of an embodiment according to the present invention (when the rotating shaft is inconsistent with the roll center)

Figure 3b is an illustration of an embodiment (when a part of the roll is worn) according to the present invention

3C is an explanatory diagram of an embodiment (complex case) according to the present invention;

* Explanation of symbols for the main parts of the drawings

10: Load cell

11: tachometer

12: Back up roll

13: Work roll

14 material

15: input unit

20: Eccentricity diagnosis device

21: data detection unit related to the eccentric component

22: eccentricity determination unit

23: screen display

31: Digital band pass filter with center frequency equal to rotation frequency

32: Digital band pass filter with center frequency twice the rotation frequency

33: Digital band pass filter with center frequency three times the rotation frequency

34: Digital band pass filter with center frequency four times the rotation frequency

35: Digital band pass filter with center frequency 5 times rotational frequency

36: Digital band pass filter with center frequency 6 times rotation frequency

37: adder

38: noise suppression filter

39: center frequency calculation unit of the digital band pass filter

Preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. 1 shows the rolling load from the load cell 10 through the input unit 15, the rotational speed of the backup roll from the tachometer 11, and the rolling load from the eccentric component-related data detection unit 21. From the data, frequency components of 1, 2, 3, 4, 5, and 6 times the rotation frequency corresponding to the rotational speed are extracted, and the eccentric determination unit 22 extracts certain components from the extracted eccentric component related data. Diagnosing whether an eccentricity problem is caused by a cause, and the screen display unit 23 can show the waveform of the detected eccentric component and the cause of the diagnosed eccentricity.

The operating principle of the present invention having the configuration as described above will be described in detail. Sampling time for each rolling load and rotation speed from load cell 10 and tachometer 11 Every time you get Rolling load measured at the first time point And roll speed When you receive The operation of the eccentric component related data detector 21 is started to extract the eccentric related data contained in the. The principle of this operation is as follows.

First, we introduce the concept of analysis window. In order to determine whether the eccentricity is in real time, it is appropriate to diagnose with the data acquired during approximately one revolution of the roll. Therefore, the number of data required for one diagnosis is promised as an analysis window. It is written as,. here Indicates when the analysis window is newly modified. Considering the meaning of It can be seen that is expressed as Equation (1).

-------------------------(One)

Is a constant value, but the rotational speed of the backup roll may vary during rolling, so it is appropriate for diagnosis. Should be changed according to the rotation speed.

The output of the tachometer Is usually given in units of revolutions per minute (rev / min), Rotation frequency of the backup roll at the first time point (cycle / sec) Is the same as Equation (2).

-------------------------------------------------- -----(2)

Also the rotation frequency Is the number of data acquired during one rotation of the backup roll. Is as shown in equation (3).

-------------------------------------------------- -(3)

Then the appropriate By introducing the concept of mean can be expressed as in Equation (4).

----------------------------------(4)

This operation end Is done once every new save, and will be used next time. this doggy The sum of Calculated by averaging Also here Depends on the rotational speed of the roll Function to compensate for errors, for example, when the rotational speed decreases, doggy When calculating the size of the analysis window, the analysis window becomes smaller than the number of data obtained during one revolution of the roll when the speed decreases, and thus cannot be an appropriate analysis window. Therefore, as a way of correcting this problem, Multiplied by May be selected from a number greater than 1 and less than 2 with reference to the operating conditions of the rolling mill.

In the present invention, the information about the rolling load Since multiple digital band pass filters 31, 32, 33, 34, 35 and 36 are used to extract only the components related to the eccentricity, the center frequency of each digital band pass filter must be determined. But if the rotation speed of the roll changes The eccentric components included in the fluctuate at the same speed, so the center frequency must be changed according to the rotational speed of the roll to accurately filter the eccentric information.

Center frequency of the digital bandpass filter that filters the signal equal to one times the eccentric frequency at the second point in time Is calculated by the center frequency calculating section 39 of the digital band pass filter and should be changed as in Equation (5) using the result of Equation (2).

-------------------------------------------------- ----- (5)

But Is selected by Equation (5), when the rotational speed of the roll fluctuates rapidly, the digital band pass filter having the center frequency determined by Equation (5) does not accurately extract the current eccentricity information. By observing the trend You need to change. Usually this calibration process is simple and straightforward to use because it is straightforward. Is calculated by equation (6) rather than equation (5).

---------------------------------- (6)

Equation (6) is Is calculated whenever is changed, where Is the weight for the correction term, if the rotational speed fluctuates Is a number between 0.5 and 1 Is chosen from 0.1 to 0.5.

Digital Bandpass Filters It has a function to filter the eccentric information contained in the Where angular velocity and B are bandwidths, the transfer function The second-order analog Butterworth type is designed by discretizing the Tustine approximation method and the approximate discrete transfer function Suppose that Equation (8).

------------------------------------------------ (7 )

---------------------------------------------(8)

here

to be. In addition, as a digital band pass filter, a Chebyshev type or Bessel type filter may be used, and the order of the filter may be a second order or more.

However, since the effects of the eccentricity vary depending on the cause of the eccentricity, a plurality of digital band pass filters are required to detect all types. In the present invention, a method of arranging a plurality of digital band pass filters in parallel form and summing the outputs in the adder 37 was selected. As an example, the number of digital band pass filters was selected as six. That is, the output of each of the digital band pass filters 31, 32, 33, 34, 35, 36. , Are computed with the difference equation as in equation (9).

------------------------- (9)

Where the center frequency of each filter Is obtained from equation (6). On the basis of And when Is calculated by equation (10).

--------------------------------------------- (10)

The output of each digital bandpass filter Are added together, Eccentricity Information Included in Is given by Eq. (11).

-------------------------------------------------- -(11)

Equations (4) and (6) doggy Is performed once, but Eq. (11) is performed every sampling period.

Or the output of a digital bandpass filter , Noise may be included in the noise suppression filter 38 Is averaged as in Eq. (12) Final value of eccentricity components extracted from Get

--------------------------------------- (12)

this Is calculated, the calculation of the eccentric component related data detector 21 is completed. Since the noise suppression filter 38 is a function for reducing the influence of various noises included in the rolling load data, the noise suppression filter 38 may be installed in front of the digital band pass filters 31, 32, 33, 34, 35, and 36. .

Next The principle and function of the eccentricity determination unit 22 for determining the presence or absence of eccentricity and the cause of the eccentricity will be described. The function of the eccentricity determination section 22 is the analysis window (from the eccentric component-related data detection section 21) ( Data for rolling loads , Is obtained, doggy It is to determine the existence of eccentricity and to diagnose the cause of the eccentricity. First, the two variables used in the determination of the eccentricity are explained in detail.

After taking the absolute value of, Extract 10 values from, in order of largest value, and extract 8 data from the maximum and minimum values. Let's promise. Next At the first diagnosis Arithmetic mean of Is as shown in equation (13).

--------------------------------------------- (13)

this Is a variable with information about the maximum magnitude of the eccentric component that occurs within an analysis window.

Also At the first diagnosis Which is the arithmetic mean of Is as shown in equation (14).

----------------------------------------------- (14)

And Load cell 10 output in one analysis window at the first diagnosis point Arithmetic mean of, Is equal to equation (15).

-------------------------------------------------- (15)

silver Unlike always Since it is a phosphorous property, it is not necessary to take an absolute value.

The method of determining the presence of eccentricity is as follows. The eccentricity is more of a problem than the size of the rolling load due to the influence of the eccentricity. In other words, it is appropriate to determine whether eccentricity is generated by the ratio of the variation of the rolling load caused by the eccentricity in the currently measured rolling load. At the first diagnosis of

-------------------------------------------------- (16)

If you promise, Is an appropriate variable for the presence or absence of eccentricity.

The reference values of are as follows. Eccentricity, the rolling load is When fluctuating by The minimum value of If you promise to set the rolling load Rain of To

---------------------------------------------- (17)

When you promise to end May be an appropriate reference value. In other words , It is determined that an eccentricity has occurred, Can be determined according to the operating standards of the installation.

Next, a method for diagnosing the cause of the eccentricity will be described. The cause of the eccentricity may be a case where the center of the rotating shaft does not coincide with the center of the roll, a roll is worn out, or a complex case. According to the present invention, since the eccentric information detected by the digital band pass filter is shown in the form of a waveform that can be easily visually recognized in the time domain, the eccentric diagnosis apparatus automatically detects the cause of the eccentricity only by the operator looking at the waveform. In order to be able to identify the cause of eccentricity, we need a criterion that can be judged by some figure. Therefore, in order to automatically diagnose the cause of the eccentricity, it is necessary to observe the form of the influence of the eccentricity caused by each cause and to create an indicator to distinguish the causes. At the first diagnosis of

------------------------------------------------- ( 18)

If you define as Due to the inherent characteristics of the eccentric waveform according to the type of eccentricity, is usually 0.6 or more when the axis of rotation is inconsistent, and 0.3 or less when the roll is worn. And If the value is between 0.3 and 0.6, it can be diagnosed that the cause of the eccentricity is complex.

3A, 3B and 3C show an embodiment. 3A, 3B and 3C show the case where the center of the roll does not coincide with the axis of rotation (FIG. 3A), when wear occurs on a part of the roll (FIG. 3B) and when the two causes are combined (FIG. 3C). The eccentric component-related data extracted from the rolling load data acquired by the load cell 10 is displayed on the screen display unit 23 in a waveform form in the time domain. In addition, since the cause of the eccentricity in each case can be automatically diagnosed from the calculation of the index by the method for diagnosing the cause of the eccentricity presented in the present invention, the operator can immediately take appropriate measures when necessary.

According to the present invention, the eccentricity occurring in the rolling mill can be checked at all times during rolling, and the diagnosis of the cause of the eccentricity is automatically diagnosed and notified to the operator. And prevents downtime, and consequently there is an important industrial effect that greatly improves productivity. In addition, in the present invention, since the eccentric component is extracted as data in the time domain rather than the spectrum of the frequency domain, the result is directly used in an eccentric control technique capable of suppressing the influence of the eccentricity, thereby improving the performance of the eccentric control. The side effect of increasing productivity by extending the lifespan can also be expected.

Claims (1)

In the eccentric diagnostic apparatus of a rolling mill, a plurality of digital band pass filters (31) (32) (33) (in parallel form having a structure in which sampled rolling load data are shifted in the same direction according to the rotational speed of the roll) 34) (35) and (36) combine the outputs obtained by the adder (37) and noise suppression filter (38) to reduce the influence of noise to extract the eccentric component data generated during rolling in the time domain. Is shown as a function of time, and the indicators related to the eccentricity and the cause of the eccentricity are created using the sampled rolling load data and the extracted eccentric component data. Eccentricity diagnostic method of the rolling mill, characterized in that for determining the cause and the cause of the eccentricity.