JPS63274806A - Detection of shape of slip in continuous annealing furnace - Google Patents

Abstract

PURPOSE:To enable detection of the position and size of a heat buckle and the inclination, etc., of a steel sheet by subjecting the image information obtd. by a TV camera to signal processing. CONSTITUTION:Image data A is converted B to one-dimensional displacement data arrangement Z(i), (i=1-N) by preprocessings including binarization, fining of lines, and supplementation of dropout points and is made into the displacement data arrangement X(i), (i=1-N) in the transverse direction of the steel sheet where high-frequency components are removed by moving average. The components except the heat buckle contained in X(i) are subjected to recurrence with the pass position of the steel sheet as a constant term (P0), the inclination of the steel sheet as a proportional term (P1.i), the C camber of the steel sheet as a sinusoidal half-wave term (P2.sin(pi.i/N)), and the general shape of the steel sheet as the 2/3 wave of the full sinusoidal wave + sinusoidal wave (P3.sin(2pi.i/N)+P4.sin(3pi.i/N)). The heat buckle arrangement h(i), (i=1-N) is formed D by subtracting the total sum X0(i), (i=1-N) from (i), (i=1-N) of the one- dimensional displacement data arrangement. The heat buckle is detected by adding the decision of a crest value to h(i).

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method for detecting the shape of slip in a continuous annealing furnace, and more specifically, detecting the shape of a slip in a continuous annealing furnace from a widthwise displacement distribution signal of a strip, for example, a steel plate, running in a continuous annealing furnace. The present invention relates to a method for detecting inclination in the width direction, C warpage, and general shape.

[Conventional technology]

A flat steel plate running in a continuous annealing furnace is affected by uneven tension distribution and temperature distribution in the width direction of the plate, shear force due to meandering of the steel plate, hearth roll crown to prevent meandering, etc. Compressive stress occurs in the direction. It is known that when this compressive stress exceeds the buckling strength of the steel plate, heat buckling occurs in the steel plate.

This heat buckle appears to be located in the center of the board, or
This appears as vertical wrinkles in a portion of the roll taper, and if the degree of buckling progresses further, it will induce plate breakage, causing long-term line stoppages and flaws in the roll. Such phenomena often occur particularly during high-temperature threading of extremely soft materials with low buckling strength, during tension fluctuations such as during acceleration and deceleration, and in zones where the temperature rapidly changes during heating and cooling.

In addition, the steel plate running in the continuous annealing furnace may tilt in the width direction,
It may exhibit a bend in the width direction called C-curvature, or a shape such as an ear wave or a middle elongation.

Detection of the shape of steel plates has been proposed in the past. For example, in Japanese Patent Application Laid-Open No. 59-176609, a hot rolled steel material is scanned with multiple photoelectric elements, the detection output corresponding to the temperature is quantized, and a multivalued signal is generated. It is disclosed that a binarization threshold is determined based on a density histogram by integrating the density histogram, and a shape is detected. Furthermore, JP-A No. 59-178127 discloses that the central part of the strip in the width direction is pressed with a roll to deform it, and that the warpage of the strip is detected by shining light on the deformed shape and detecting the reflected light. .

[Problem that the invention seeks to solve]

Although such shape detection methods exist, it is difficult to apply them to detecting the shape of a steel plate running in a continuous annealing furnace. The reason for this is that in a continuous annealing furnace, it is a major requirement to maintain a uniform temperature as much as possible across the width of the sheet during the heating, soaking, and cooling processes in order to obtain the desired material properties. It is difficult to detect the shape of

In addition, since the continuously annealed steel plate is then turned into a product,
It is not preferable to press and deform the central part of the board width with a roll.

Conventionally, in continuous annealing furnaces, operators visually check the television screens installed in each furnace to determine the occurrence of heat buckling, and then reduce the tension on the steel plate.
This reduced the compressive force caused by the hearth rolls and prevented the steel plate from breaking. Also.

The degree of inclination of the steel plate in the width direction, C-curve shape, etc. was detected by visually observing the television screen.

However, this makes it difficult to accurately or continuously detect the shape of the strip.

In addition, there is a problem that the operator cannot constantly monitor the television camera, and there has been an urgent need to automate the shape detection 2 judgment.

The present invention irradiates a strip running in a continuous annealing furnace with laser light, photographs the laser irradiated part with a television camera, and performs signal processing on the image information obtained.
The purpose is to detect a heat buckle, detect its size and two positions, and also automatically identify and detect the inclination of a steel plate, C warp, and film shape.

[Means for solving problems]

In order to achieve the above object, the present invention pre-processes heat buckle image data captured from a television camera by binarizing it, thinning it, and filling in missing points to create a one-dimensional displacement data array z(i), i=1 Displacement data array in the width direction X(i), i
= 1 to N, and the components other than the heat buckle included in the displacement data array
t・i).

The C warp of the steel plate is expressed as a half-sine term (P2・sin (π・i/
N)), the general shape of the steel plate is a full sine wave and a three-half sine wave (
P3-sin(2π・i/N) + P4・5in(
3c・i/N)), and the sum of these X0(i
), i = 1 to N as X(i
), i=1~N to generate a heat buckle data array h(i), i=l~N, and the h (
Heat buckles are detected by adding peak value determination to i). Furthermore, calculate the regression equation and calculate the slope of the steel plate: P1
°C warpage: P2. Shape: (Pa, Pa) components can also be detected.

The present invention will be described in detail below based on one embodiment with reference to the drawings.

FIG. 1 schematically shows an overview of signal processing in an embodiment of the present invention. FIG. 2 is an explanatory diagram showing a heat buckle detection device implementing the present invention in one embodiment.

A linear laser beam is irradiated from a projection head 2 in the width direction of the steel plate 1, and the portion irradiated with the laser beam is photographed by a camera head 3. The camera head 3 extends 3 in the width direction of the steel plate 1.
An example is shown in which images are taken by dividing, and a movable mirror 4 is attached for this purpose, but instead of this, a plurality of cameras may be installed in parallel in the width direction.

The image 1 taken by the camera head 3, for example, the image A in FIG. The signal is captured into the internal image memory of 6, undergoes one-image binarization and thinning processing in the signal processing unit 8, and is converted into a -dimensional data array. The -dimensional data is
Since missing points may occur depending on the angular relationship between the laser beam projection/reception and the steel plate, the signal processing unit 8 further performs primary complementation processing for the missing points, for example, in the steel plate width direction shown by B in FIG. A displacement signal array Z(i), i=1 to N, is obtained.

By the way, the displacement signal array Z(i) includes information on the inclination in the width direction of the steel plate pass line position 9, C warp, and general shape other than the heat buckle, in addition to information on the heat buckle of the steel plate. There is. Therefore, the signal processing unit 8 separates each of these information components and first detects the heat buckle component. Next, the method will be described.

The steel plate width direction displacement signal array Z (,i), i = 1
~N is.

First, in order to remove high frequency noise, a plurality of points are subjected to moving average processing and converted into, for example, a displacement data array X(i) in the board width direction shown by C in FIG. 1, where i=1 to N. Next, in order to separate and remove components other than the heat buckle described above, the displacement data array X(i) in the width direction of the steel plate, i=1
~N is expressed by the function formula Xo (iL
Regression is performed using i=1 to N, and heat buckle component h(i), i=1 to N, is further detected using equation (2).

X0(i)'Pa +P1 ・f+P2 ・sin(g
f)+P3 ・sin(2sf)+P4 ・5in(3
πf)...(1) (f insect i/N, i=1~N) h (i) A X (i) X o (i)
...(2) (i = 1 to N) In the formula (1), the first term PG represents the pass line position of the steel plate running in the continuous annealing furnace, and the second term Pl・f
represents the inclination of the pass line in the board width direction. Third term Pl.
sin (πf) is a term representing C warpage of the steel plate.

In addition, the fourth term and the fifth term P3 ・5in (2tc f
)+Pa·5in (3tcf) represents the general shape component of the steel plate other than the heat buckle.

h(i) is a heat buckle data array, for example shown as D in FIG. Here, P OtP 1
s Pl e pa e Pa is a coefficient representing the intensity (magnitude) of each term, as shown in equation (3) below.

X(i)! Ninety people・f+i・5in(πf)+P
3 ・5ln (2sf) + P4 ・5xn (3πf) (
f=i/N, i=1~N) The heat buckle data array h(
i), i=1~N, and information other than the heat buckle pass line rank 1 of the steel plate! P o is the slope Pt in the board width direction
, C warp Pl, and the wave shape P3rP4 in the width direction of the steel plate are:
It is sent to the microcomputer 7. The microcomputer 7 has a heat buckle data array h in the board width direction.
(i) The wave height value h*ax of i=1 to N and its width direction position ii t wax are determined.

Furthermore, among M sets of heat buckle data arrays h(i), i=1 to N, which are continuous in one time series, those whose h wax values exceed a predetermined reference value are predetermined. If the number of pieces exceeds the permissible upper limit set for m, then h
Depending on the size of max, an alarm signal indicating a large heat buckle or a small heat buckle is sent to the alarm device 12. Then, there are successive L sets of heat buckle data arrays h(i).

When the number of h max values exceeding the reference value hO among i=1 to N is less than a predetermined Ω, a command to turn off the alarm is sent to the alarm device 12 . In addition, information on the pass line position PO + slope Pt, C warp -2, and -film shape (F3 + P4) of the steel plate is obtained by adding appropriate dimensions to each time information, heat buckle position, size. The information is output to the printer 10 as numerical data. Furthermore, if it is determined that heat buckling has occurred, the image data is recorded as a file in the external storage device 11. In addition, adjustment ITV13. and IT for the driver's cab
V14 can display any image, either raw image data or data subjected to image signal processing, via the image input/output device 6.

The microcomputer 7 has a function of controlling the entire detection system, that is, the camera head 3, the image input/output device 6. It also has a function of controlling and supervising the image signal processing section 8 and the like.

〔Effect of the invention〕

As explained above, in the present invention, the position and size of the heat buckle are determined by applying signal processing using mathematical regression to the width direction displacement signal of the steel plate obtained from a system using a laser and a television camera. The tilt of the steel plate in the continuous annealing furnace.

C. Warp and shape can be detected.

Currently, a detection method employing the present invention is in operation on an actual line and is achieving a high detection rate. In the future, it is expected that it will not only have the function of monitoring the behavior of steel sheets in a continuous annealing furnace as described above, but also function as a sensor for heat buckle suppression control and a sensor for controlling the shape of steel sheets.

[Brief explanation of drawings]

FIG. 1 is a block diagram schematically showing the image data processing method or process of the present invention. FIG. 2 is a block diagram showing the configuration of an apparatus that implements one embodiment of the present invention.

Claims (3)

[Claims] (1) Laser light is irradiated onto the strip running through a continuous annealing furnace, the irradiated area is photographed with a television camera, and the resulting image is subjected to signal processing such as binarization, thinning, and missing point complementation in the width direction. A displacement signal is used as a displacement signal, and this displacement signal is moved averaged at multiple points to remove high-frequency noise, and a displacement data array in the width direction of the strip is used. 1. A method for detecting the shape of a strip in a continuous annealing furnace, characterized by performing regression using a functional equation expressed by the regression method, and detecting the pie line position, inclination in the width direction, C warpage, and wave shape of the strip from the regression. (2) Laser light is irradiated onto the strip running through a continuous annealing furnace, the irradiated area is photographed with a television camera, and the resulting image is subjected to signal processing such as binarization, thinning, and missing point complementation in the width direction. A displacement data array X (
i), a regression equation X_0(i) is obtained by regressing the displacement data array using a functional equation representing the pass line position of the strip, the inclination in the width direction, the C warp, and the wave shape.
) to obtain the heat buckle data array h(i) by subtracting the regression formula
x and repeat the above process to obtain the wave height value h_m
A method for detecting the shape of a strip in a continuous annealing furnace in which heat buckling is determined to have occurred when _a_x exceeds a predetermined reference value h_0 for a predetermined allowable number of times or more. (3) A strip in a continuous annealing furnace according to claim (2), characterized in that when heat buckling occurs, the size of the heat buckle is determined based on the peak value within a predetermined allowable number of times. shape detection method.