Patent Application
20200338609
This application claims priority under 35 U.S.C. § 119 to Chinese Patent Application No. 201910347862.4 filed Apr. 28, 2019, the entirety of which is incorporated by reference herein.
The present invention belongs to the field of strip rolling, and particularly relates to a method for channel decoupling of a whole-roller flatness meter for a cold-rolled strip.
With the advantages of high performance and high precision, cold-rolled strips are widely used in industrial manufacturing sectors such as automobiles, home appliances, construction and electronics. Cold-rolled strips are high-value-added (HVA) products. The technical level of strip production demonstrates a country's ability to produce iron and steel and has already become a characteristic of a strong country of iron and steel industry and an important symbol of a country's industrialization level. Flatness is an important quality indicator of cold-rolled strip. Poor flatness will cause difficulties to the subsequent processes and lead to the occurrence of accidents like strangled rolls or broken strips and may even damage the rolling mill in serious cases.
Flatness detection is the key to control flatness and improve the flatness quality. Flatness meters are necessary high-end instruments for producing advanced cold-rolled strip and realizing intelligent production process. There are various types of cold-rolled strip flatness meters, such as multi-piece type (segmented type), probe type and whole-roller type. Flatness meters technology has been monopolized by a few large international companies for a long time. In recent years, China has made major progress in the research of cold-rolled strip flatness meters, and independently developed the whole-roller flatness meter.
The whole-roller flatness meter is provided therein with 2 to 4 elongated holes in the axial direction near the surface inside the detection roll. A series of sensors are installed in the holes. The sensors each have an axial width of 26 mm and the detection channels are distinguished by the corresponding axial position of each sensor in the holes. The whole-roll type flatness meter is different from the probe type flatness meter in which sensors are spirally arranged on the detection roll. There are fewer mounting holes on the detection roll of the whole-roll flatness meter and the sensors are arranged next to each other along a straight line, which simultaneously detect the flatness of the strip on the same cross section to ensure synchronous flatness detection. However, because the sensors are arranged next to each other in the holes, the adjacent channels overlap and coupled obviously, causing errors in flatness detection. Therefore, it is necessary to decouple the channels for accuracy. At present, there are no accurate decoupling methods reported in the world.
An objective of the present invention is to provide a method for eliminating interference between channels of a whole-roller flatness meter, so as to improve the flatness detection accuracy.
The present invention provides a method for channel decoupling of a whole-roller flatness meter for a cold-rolled strip, including the following steps:
a, setting a channel number n and a channel breadth b of the flatness meter;
b, obtaining an influence matrix under the condition of signal interference between the channels, which includes the following steps:
b1, making a temporary variable i=1;
b2, making a temporary variable j=1;
b3, using a calibration device to apply a calibration force to an i channel of the flatness meter;
b4, recording an analog/digital (AD) influence value of the i channel on a j channel;
b5, determining whether j=n is true; if yes, going to b6; if not, making j=j+1 and returning to b4;
b6, determining whether i=n is true; if yes, going to b7; if not, making i=i+1 and returning to b3;
b7, making a temporary variable i=1;
b8, making a temporary variable j=1;
b9, calculating an influence coefficient βji=αji/αjj of the i channel on the j channel;
b10, determining whether j=n is true; if yes, going to b11; if not, making j=j+1 and returning to b9;
b11, determining whether i=n is true; if yes, going to b12; if not, making i=i+1 and returning to b9; and
b12, forming an influence matrix
c, calculating an inverse matrix (
d, using the inverse matrix of the influence matrix to decouple the channels according to a measured signal of the flatness meter; and
e, obtaining flatness distribution after channel decoupling.
Preferably, step d includes the following steps:
d1, setting a detection force signal Hi of the flatness meter, i ranging from 1 to n, and forming a column vector
d2, multiplying the inverse matrix (
Preferably, step e includes the following steps:
e1, setting a total strip tension T, a strip breadth B and a mean strip thickness h, and calculating a mean strip tensile stress σmean=T/(Bh);
e2, dividing the strip breadth B by the channel breadth b and rounding to obtain a temporary integer mi;
e3, determining whether m1 is an odd number; if yes, making a strip-covered channel number of the flatness meter m=m1 and going to e4; if not, making the strip-covered channel number of the flatness meter m=m1+1, and going to e4;
e4, making a left boundary number of the strip-covered channel number of the flatness meter mz =(n−m)/2+1 and a right boundary number of the strip-covered channel number of the flatness meter my=n−(n−m)/2;
e5, calculating a mean force
and
e6, setting an elastic modulus E and a Poisson's ratio ν of a strip, and calculating true flatness distribution
where i ranges from mz to my.
Compared with the prior art, the present invention has the follow advantages:
The present invention decouples the channel of the whole-roller flatness meter by inverting the influence matrix and multiplying with the detection force vector, thereby reproducing the true force vector and flatness distribution and improving the flatness detection accuracy.
1. motor, 2. calibration bracket, 3. calibration beam, 4. calibration weight, 5. detection roll, 6. bearing seat, 7. pressure roller, 8. calibration rod, 9. sensor, 10. elongated hole, and 11. Detection channel.
The technical solutions in the examples of the present invention are clearly and completely described with reference to the accompanying drawings in the examples of the present invention. As will be apparent, the described examples are merely a part rather than all of the examples of the present invention. All other examples obtained by a person of ordinary skill in the art based on the examples of the present invention without creative efforts should fall within the protection scope of the present invention.
In the present invention, in order to reduce a flatness detection error caused by the coupling between adjacent channels 11, the channels need to be accurately decoupled. As shown in
a, set a channel number of the flatness meter n=57, a channel breadth b=26 mm, where the channel number is equal to the number of sensors in one elongated hole, that is, a plurality of sensors 9 are provided one after another along the elongated hole 10.
b, obtain an influence matrix under the condition of signal interference between the channels. A calibration force is applied to an i channel by a calibration device, and an analog/digital (AD) influence value of the i channel on a j channel is recorded. In an actual calibration process, the AD influence values of the i channel on ≤i −2 and >i+2 channels are approximately zero. Therefore, it is only necessary to record the AD influence value of the i channel on an i−1 channel, the i channel and an i+1 channel. The recorded results are shown in Table 1 below. In the table, a first column indicates a code of a channel applying a calibration force, a second column indicates an AD influence value of a channel applying a calibration force on a previous channel, a third column indicates an AD influence value of a channel applying a calibration force on itself, and a fourth column indicates an AD influence value of a channel applying a calibration force on a subsequent channel.
The main structure of the calibration device is shown in
Influence coefficients βji in the influence matrix are respectively calculated with the data of Table 1, as shown in Table 2 below. j is a row number of the matrix, also ranging from 1 to n, and i is a column number of the matrix, ranging from 1 to n. The influence coefficients in the influence matrix
c, calculate an inverse matrix (
Elements in columns 9 to 16 of (
Elements in columns 17 to 24 of (
Elements in columns 25 to 32 of (
Elements in columns 33 to 40 of (
Elements in columns 41 to 48 of (
Elements in columns 49 to 56 of (
Elements in column 57 of (
d, use the inverse matrix of the influence matrix to decouple the channels according to a measured signal of the flatness meter, specifically as follows:
d1, set a detection force signal Hi of the flatness meter, i ranging from 1 to n, and form a column vector
d2, multiply the inverse matrix (
e, calculate flatness distribution after channel decoupling, specifically as follows:
e1, set a total strip tension T=64 kN, a strip breadth B=1150 mm and a mean strip thickness h=1.0 mm, and calculate a mean strip tensile stress σmean=T/(Bh)=55.65 MPa.
e2, divide the strip breadth B by the channel breadth b to get 44.23, and round to obtain a temporary integer m1=45.
e3, determine that m1 is an odd integer, and make a strip-covered channel number of the flatness meter m=m1=45.
e4, make a left boundary number of the strip-covered channel number of the flatness meter mz=(n−m) /2+1=(57−45) /2+1=7, and a right boundary number of the strip-covered channel number of the flatness meter my=n−(n−m)/2=57−(57−45)/2=51.
e5, calculate a mean force
e6, set a strip's elastic modulus=210000 MPa E and Poisson's ratio ν=0.3, and calculate
true flatness distribution
where i ranges from mz to my. The calculation results are shown in the third column in Table 12 below. If the detection force vector
According to the second and third columns of Table 12 and
The present invention decouples the channel of the whole-roller flatness meter by inverting the influence matrix and multiplying with the detection force vector, thereby reproducing the true force vector and flatness distribution and improving the flatness detection accuracy.
Finally, it should be noted that the above examples are merely intended to illustrate the present invention, rather than to limit the technical solutions described in the present invention. Therefore, those of ordinary skill in the art should understand that although this specification describes the present invention in detail with reference to the above-mentioned examples, the present invention can still be modified or equivalently replaced. All technical solutions and improvements made without deviating from the spirit and scope of the present invention should be covered by the scope of the claims of the present invention.
| Number | Date | Country | Kind |
|---|---|---|---|
| 201910347862.4 | Apr 2019 | CN | national |
