JPS64155B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a groove welding tracing control method for controlling a welding torch by accurately detecting the three-dimensional distance between an imaging device and a welding groove in a non-contact manner.

Conventionally, most of the welding in horizontal fillet welding was done manually, resulting in uneven quality and high costs due to rising labor costs. Therefore, in recent years, non-contact and highly accurate distance sensors have been developed as part of welding automation. Distance measurement using the optical cutting method has good accuracy, has been put into practical use in some cases, and is promising for the future.

FIG. 1 is a diagram showing an example of the above-mentioned practical device. That is, this device abuts the second welding base material 1b on the first welding base material 1a vertically, for example, to obtain the welding groove 2, and also provides illumination on one side of the welding groove 2. Set up device 3. This illuminator 3 irradiates a laser slit beam 4 from a direction tilted at an angle of θ from the y-axis with the z-axis as the center and hits the welding groove 2 to obtain a slit image 4' showing the cross-sectional shape of the groove. .
Reference numeral 5 designates an imaging device that captures the slit image 4' as slit image data, 6 a welding torch, and 10 a signal processing section that performs signal processing on the slit image data from the imaging device 5 to control the welding torch 6.

Specifically, the signal processing section 10 performs the following processing. That is, the slit image data captured by the imaging device 5 is transmitted to the image capture control circuit 12 in response to a capture control signal from the vertical/horizontal synchronization detection circuit 11.
takes in one horizontal line at a time, and stores this one horizontal line of slit image data in a subsequent image horizontal one line storage circuit 13 having a storage capacity of one line.
14 is the brightest point detection circuit, which is
The brightest point is found from among the slit image data of one horizontal line stored in the line storage circuit 13 and is supplied to the smoothing differentiation circuit 15. Then, the smoothing differentiation circuit 15 and the shoulder determination circuit 16 smoothed and differentiated the data of the brightest points of several lines to obtain the coordinate data of the shoulders M and N shown in FIG. 2 on the imaging surface. After that, groove center position calculation circuit 1
Since Z on the imaging plane can be known from the coordinate data of the shoulders M and N in step 7, the groove center position of the welding groove portion 2 is calculated by calculation, and based on the calculation result, the welding torch drive control calculation circuit 18 The welding torch drive device 19 controls the welding torch 6 to move to the center position of the groove.

However, in the case of the above-mentioned apparatus, the following points are considered to be problems. In other words, with this device, in the case of horizontal fillet welding as shown in FIG. 3a, although the groove center position of the welding groove 2 can be detected, the If the distance in the direction is not known, it is not possible to automate welding on the I line, which is the groove center direction. Therefore, in the case of fillet welding, the position of the welding torch 6 on the I line is fixed, but for example, the groove position may be uneven, thermal distortion may occur, or the position of the first weld base may be uneven. When fillet welding the curved second welding base material 1b to the material 1a, welding cannot be performed automatically unless the distances in the x and z directions are known.

In addition, a V-shaped welding groove 2 as shown in FIG.
In the case of welding or butt welding, the position of the welding torch 6 in the z direction is always fixed in advance, but in reality, the groove line surface often changes, and the inner surface of a box-shaped structure or steel pipe In the case of welding, etc., it is very difficult to mechanically keep the position in the z direction constant at all times. Furthermore, it is difficult to deal with misalignment during temporary attachment and multi-layer welding. In this multilayer welding, the groove center positions are points A, B, etc. on line I as shown in FIG. 3c, but in reality, the welding torch 15 and points C, D,
Unless the positional relationship with points E, F, etc. is known, the depth direction cannot be grasped, so there is a drawback that multilayer welding cannot be performed.

The present invention has been made in view of the above-mentioned circumstances, and it is possible to detect the three-dimensional distance between the imaging device and the welding groove with a simple configuration, and it is also possible to perform automatic welding with high precision for various types of welding. A method for controlling groove welding profiling is provided.

Below, in explaining one embodiment of the present invention,
First, the relationship between the principle of the optical system that is the basis for obtaining three-dimensional distances and the groove shape will be described. Generally, the imaging device 20 and the measurement target 21 have a relationship as shown in FIG. 4. In the figure, 20A is a lens, 20L is an imaging surface, and LC is an optical axis. Therefore, the imaging device 2
0 and the measurement object 21 have a relationship as shown in FIG. 4, the distance x _a between the imaging device 20 and the measurement object 21 can be expressed as x _a =·(A+B)/B (1). Here, is the focal length, A is the length of the measurement object 21, and B is the captured image 20L.
The length of the measurement target above, x _b , is the distance between the lens 20A and the imaging surface 20L. In the above formula, the focal length can be set in advance according to the measurement range, so A and B
If we know, we can find the distance x _a .

Therefore, considering the relationship between the lengths A and B and the welding groove shown in FIG. 5, the length A of the measurement object 21 corresponds to either value Y or Z in FIG. corresponds to the distance on the Y or Z imaging surface 20L. Here, since the plate thickness (or groove depth) h and the groove angle ψ of the welding groove are known as shown in FIGS. 6a and 6b, for example
If the welding base material is irradiated with laser slit light from a direction tilted at an angle of

Y=h/tanθ...(2) Z=h/tanψ...(3) Therefore, from equations (2) and (3), the length A can be determined as known. On the other hand, B is obtained from Y and Z obtained by arbitrarily determining the origins of y and z and performing image processing on the slit image data accumulated on the imaging surface 20L at that time. Therefore, it is possible to obtain the distance x _a in the Z direction based on the above principle, and furthermore, it is possible to obtain three-dimensional distance data in x, y, and z from the imaging device 20 at the M point and the N point. be able to.

However, during actual measurement, it is necessary to perform calibration once to improve accuracy, calculate the absolute value of B, and set the focal length of the lens system. Therefore, actual measurements including the calibration method will be explained with reference to FIGS. 7 to 9. First, the calibration method is performed according to the flowchart shown in FIG. First, in the block diagram, measurement conditions are set to obtain the Y and Z of a real image. Next, in the block diagram, the distance x between the measurement object 21 and the imaging device 20 is actually determined, and the focal length is determined based on this. Next, Blockha said,
After determining, by substituting Y or Z of the block diagram into equation (1) and transforming equation (1), the imaging surface 20L can be determined.
Obtain the absolute value B of the upper image. Next, when the image processing speed is fast, in the block technique, Y and Z on the imaging plane are determined multiple times and the average value is determined to improve accuracy. Then, the resolution in the y and z directions = absolute value of the image on the imaging surface/average sample value (number of dots) by image processing is determined.

Next, Figure 8 shows that under the calibration method as in Figure 7,
This is one method of performing welding tracing by obtaining three-dimensional positional information of points M and N, and this is essential information when performing multilayer welding. FIG. 9 shows a method for performing welding tracing without using three-dimensional position information.

Next, a first embodiment of the groove welding tracing control method according to the present invention realized based on the above principle will be described.
This will be explained with reference to FIG. 0 and FIG. 10th
The figure shows the overall configuration of an apparatus to which the method of the present invention is applied, in which a second welding base material 1b having a groove angle ψ as shown in FIG. 6 is placed on a first welding base material 21a, for example. A welding groove 22 is obtained by vertically contacting the welding groove 22, and an illuminator 23 is installed on one side of this welding groove 22.
Set. This illuminator 23 irradiates the welding groove 22 with a laser slit beam 24 from a direction tilted at an angle of θ from one axis with the z-axis as the center, and obtains a slit image 24' showing the cross-sectional shape of the groove. . Reference numeral 20 indicates an imaging device for capturing the slit image 24' as slit image data, 30 indicates a signal processing section, 31 indicates a welding torch driving device, and 32 indicates a welding torch.

FIG. 11 shows an example of the configuration of the signal processing section 30. This signal processing unit 30
a vertical/horizontal synchronization detection circuit 301 that detects vertical and horizontal synchronization signals from the circuit 301; an image capture control circuit 302 that captures slit image data line by line and converts it into a digital signal using the capture control signal from this circuit 301; An image line storage circuit 303 having a memory capacity of one line, a brightest point detection circuit 304 that detects the address of the maximum brightness point from the memory contents of each line, and a binary signal of the maximum brightness point. A binarization circuit or smoothing differentiation circuit (hereinafter referred to as a binarization circuit, etc.) 305 that obtains coordinate data of the shoulder N point and M point on the imaging surface 20L by performing smoothing or smoothing differentiation; Find Y and Z on the imaging surface from the data, and also calculate the preset plate thickness h,
From the groove angle ψ and the slit inclination angle θ, equation (2) and
After calculating the groove position calculation circuit 306 that calculates Y and Z of the real image based on equation (3) and the length B on the imaging surface 20L from the output signal of this circuit 306, the imaging device 2
0 and the groove portion ₂₂ , and a welding torch drive control calculation circuit 308.

Next, FIG. 12 is a diagram showing a schematic configuration of the welding torch drive device 31, in which 311 is a rail mount;
312 is an X-axis drive motor, 313 is a Y-axis drive motor, and 314 is a Z-axis drive motor, and these motors 312 to 314 drive the welding torch 32 with signals from the welding torch drive control calculation circuit 308. Control is performed in the direction of the dashed-dotted line arrow. 321 is a ψ-axis drive motor.

Next, the operation of the above device will be explained. First, conditions are set. Setting items include the plate thickness or groove depth h, the groove angle ψ, and the irradiation angle θ of the slit light by the illuminator 23, which are set in the groove position calculation circuit 306. Further, the focal length is set in the distance calculation circuit 307.

After setting the conditions as described above, the illuminator 2
3, slit light 24 of a laser, for example, is irradiated to form a slit image 24' showing the cross-sectional shape of the groove on the welding groove 22. The imaging device 20 is the welding groove portion 2
The slit images 24' formed in step 3 are sequentially captured by, for example, beam scanning and stored in an image section (imaging surface) as slit image data.
This slit image data (video signal) and vertical and horizontal synchronization signals from beam scanning are also output. Here, the vertical/horizontal synchronization detection circuit 301 sends a capture control signal to the image capture control circuit 302 every time it detects a horizontal synchronization signal, and the image capture control circuit 302 receives slit image data line by line from the imaging device 20. The image is taken in, converted into a digital signal, and stored in the subsequent image 1-line storage circuit 303. The number of samplings for AD conversion by the image capture control circuit 302 is determined in relation to the resolution, and therefore, the number of samplings simultaneously corresponds to the address of the image one-line storage circuit 303. The 1-line slit image data in the 1-line image storage circuit 303 is supplied to the brightest point detection circuit 304, which detects the address of the maximum brightness point within 1 line. Such operations are performed for each line to sequentially find the address of the maximum brightness point. Then, the address signal of the maximum brightness point determined by the brightest point detection circuit 304 is converted into a binary value by a binary conversion circuit etc. 305, and the coordinate data of the N point and M point on the imaging plane in FIG. This is then input to the groove position calculation circuit 306.
Here, binarization means to distinguish between the portion corresponding to the line segment MN and the other portions of the line formed by the maximum brightness points in FIG. 5. In other words, the part corresponding to the line segment MN has lower brightness than other parts because the light rays are projected diagonally.
Therefore, by binarization processing, it is possible to distinguish the MN part from other parts and detect the MN point as the boundary. In addition, the smoothed differential is a line formed by the maximum brightness point in FIG. 5 that is smoothed to eliminate small fluctuations (noise).
After removing , the inflection point MN is detected by differential processing. This groove position calculation circuit 306
Now, perform the subtraction operation on the coordinate data to obtain the fifth
Y and Z on the imaging plane of the figure are determined and supplied to the distance calculation circuit 307. Moreover, this circuit 307 is
Equations (2) and (3) are calculated based on the previously described condition setting values to obtain Y and Z in the real image, which are also supplied to the distance calculation circuit 307. Since the focal length is set in advance in the distance calculating circuit 307, the distance x _a is calculated using equation (1) using Y and Z. The welding torch drive control calculation circuit 308 is sent from the distance calculation circuit 307.
From x _a and Y, Z, the three-dimensional data x in Figure 10,
y and z are determined and inputted to the welding torch drive device 31 to control the welding torch 32 in a tracing manner.

Therefore, with the above configuration, even in the case of V-groove welding or butt welding as shown in FIG. Automatic welding is possible by detecting three-dimensional position information. Further, in the case of horizontal fillet welding shown in FIG. 3a, the distance in the x and z directions between the position of the welding torch 32 and the groove center position can be detected to determine the groove center line and welding can be performed.

Furthermore, multilayer welding can be performed using the above method. That is, when there are target welding positions A, B, C,...H, etc. in automatic multi-layer welding as shown in Fig. 13, since the plate thickness h and the groove angle ψ are known in advance, 3 of the N points The welding torch 32 can be controlled by determining the actual points A, B, C, .

By the way, in the case of automatic multi-layer welding shown in Fig. 13,
The welding point A of the first layer can be performed by the welding tracing method shown in FIG. 8 described above. However, if the target welding position of each layer is preset in multilayer welding, the amount of welding changes and welding defects such as undercuts and overlaps are likely to occur. Therefore, in multilayer welding as shown in Fig. 13, the three-dimensional x, y, and z position information from the imaging device 20 of points G and H is obtained from points C and D of the second layer, and based on experiments and experience. C,
Correct point D. Note that the welded shape after further welding appears on the imaging device 20 as shown in FIG. Here, since the coordinate data of the M point and the N point are separated, the equation of the straight line connecting the M point and the N point is created,
The H point is determined by a binarization circuit etc. 305,
Find the three-dimensional position of point H from MH/MN. Next, the N point is automatically driven so as to be at the center of the screen in consideration of error and accuracy, and the intersection point of the extension line of the MG and the perpendicular line from the N point is set as I. At this time, M,
The z coordinates of points G and I are the same. Also MI and MN
The y coordinates of are equal and can be expressed by equation (2). From this, X=Y・tanθ (4), and the three-dimensional position of point I can be found. Similarly, point G is determined and the three-dimensional position of point G is determined from MG/MI. Furthermore, from points G and H, welding torch 3
2. Determine the welding target positions C and D. Automatic multilayer welding can be performed by repeating this process repeatedly. In addition, the approximate welding area MGH and NHGI can be determined by this, which can be used for setting the welding conditions.

Note that the present invention is not limited to the above embodiments. For example, the traveling speed and plate thickness of the welding base material,
After the shoulder N is detected by the binarization circuit 305 in consideration of the groove angle, etc., a signal for skipping a required line may be given to the image capture control circuit 302.
In addition, the present invention can be modified in various ways without departing from its essential aspects.

As described in detail above, according to the present invention, after the slit image of the groove is captured as slit image data by an imaging device, the maximum brightness point is detected line by line and the image is captured by performing binarization conversion, etc. In addition to determining the coordinate data of the surface, the welding torch is controlled by tracing the welding torch by determining the three-dimensional distance from this data and the data of the real image determined by setting conditions in advance, so the distance between the welding torch and the groove is determined. can be detected automatically, which can dramatically improve the accuracy of automatic welding. Furthermore, it is possible to provide a groove welding tracing control method that can automatically weld multilayer welding and box-shaped structures.

[Brief explanation of drawings]

Figures 1 to 3 a to c are shown to explain the conventional method. Figure 1 is a configuration diagram of a device to which the conventional method is applied, and Figure 2 is an explanation for detecting the groove shape. Figures 3a to 3c are diagrams showing welding modes, Figures 4 to 9 are diagrams for explaining the principle for realizing the method of the present invention, and Figure 4 is the configuration of the optical system. Figure 5 is an explanatory diagram for detecting the groove shape, Figures 6a and b are side views of the groove,
Figures 7 to 9 are flowcharts showing the calibration method and welding tracing method, Figure 10 is an overall configuration diagram of a device to which the method of the present invention is applied, and Figure 11 is an example of the configuration of the signal processing section in Figure 10. Figures 12 and 12 are schematic configuration diagrams of the welding torch drive device, Figures 13 and 14.
The figure is an explanatory diagram of the multilayer welding method. 20... Imaging device, 21a, 21b... Welding base material, 22... Groove portion, 23... Illuminator, 24'...
...Slit image, 30...Signal processing unit, 31...Welding torch drive device, 32...Welding torch, 302
...Image capture control circuit, 303...Image 1 line storage circuit, 304...Brightest point detection circuit, 305
...Binarization circuit or smoothing differentiation circuit, 306...
Groove position calculation circuit, 307... Distance calculation circuit, 3
08...Welding torch drive control calculation circuit, 312...
...X-axis drive motor, 313...Y-axis drive motor, 314...Z-axis drive motor.

Claims (1)

[Scope of Claims] 1. A welding groove in which an illuminator illuminates the welding groove to form an optical image showing the groove shape, and this optical image is captured as image data by an imaging device and used for groove welding tracing. In the first welding tracing control method, after capturing the image data of the imaging device line by line and detecting the maximum brightness point from each line, two lines are obtained from the line obtained by connecting the maximum brightness points of each line. The shoulder of the welding groove on the imaging surface is detected by value conversion or smoothing differentiation, and two-dimensional data of the welding groove shape on the imaging surface is obtained from the obtained shoulder signal. The three-dimensional distance between the imaging device and the welding groove is determined using data, two-dimensional data of a real image obtained according to predetermined setting conditions, and a predetermined focal length. Groove welding tracing control method. 2 In a groove welding tracing control method in which an illuminator illuminates the welding groove to form an optical image showing the groove shape, and this optical image is captured as image data by an imaging device and used for groove welding tracing. , After capturing the image data of the imaging device line by line and detecting the maximum brightness point from each line, binarization or smoothing differentiation is performed from the line obtained by connecting the maximum brightness points of each line. The shoulder of the welding groove on the imaging surface is detected by the , and two-dimensional data of the welding groove shape on the imaging surface is obtained from the obtained shoulder signal, and this two-dimensional data is set according to predetermined settings. A three-dimensional distance between the imaging device and the welding groove is determined using two-dimensional data of a real image obtained under the conditions and a predetermined focal length, and this three-dimensional distance and each of the two
A groove welding tracing control method characterized by controlling a welding torch to trace using three-dimensional data obtained from dimensional data.