JPS63225106A - Method of preparing position of work surface and geometry - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for detecting the position and geometry of a workpiece surface, in particular a seam groove surface during automatic arc welding.

When using an automatic arc welder, even though the welder can follow a predetermined trajectory with great precision, the workpiece to be welded is often not as accurate as required. , there is a problem that it is not produced. This is often not economical (and indispensable in relation to the required accuracy of the final product. Therefore, for automatic welding machines to be used economically In such cases, a device is required to detect the actual position and geometry of the weld seam, so that the programmed trajectory data and welding parameters can be modified accordingly.

Devices for solving this problem are already known.

The optical solution consists exclusively in directing a sharp, bundled beam of light (in practice only a laser can be used) onto the workpiece at a specific angle to the axis of the sensing optics, and thus It is based on the fact that the movement of the impact point from a neutral position in the center is a measure of the distance of the workpiece point with respect to the sensor system. Even with a good light source (laser) and a suitable light source filter, this method has the following drawbacks: That is, the image receiving system is severely hampered by dirty surfaces and structures of the workpiece that cause scattering and reflection of the sensor beam and arc light, and the clarity control and scoring do not always proceed correctly. Without it, the function cannot be maintained accurately.

Another variant for measuring corresponding points is the parametric integration method. The method is, for example, a function of the square of failures that must admit a minimum for "detection". Directly from the difference measure, the position and distance with respect to the sensor coordination system can be determined using elementary laws of mathematics.

Correlation of individual structures can in principle be carried out by opto-mechanical movements, but electronic solutions are preferred due to their low cost. In addition to their correlation, differential methods are also used, since the surface structures should be noticeable in order to detect the entire geometry.

After the ascending method detects the geometry of the workpiece within the measuring range of the system, the device, when used for automatic arc welding, calculates the volume of the seam groove and adjusts the welding piston position and welding parameters for correction. Provide all the data to the computer placed above.

On the other hand, the computer placed on the stage can transmit the model structure to the sensor, and the sensor is used to identify the model structure of the a-eye. In this case, this identification is carried out by recognition methodological calculations for fibers, by variable calculations for high-level demands, and by calculating and storing correspondingly necessary data for other applications. , by evaluating the summation of the individual measurements for different positions, information about the entire surface of the workpiece can be known.

In particular, according to the invention, the workpiece surface is imaged using two separate optical machines on two separate image receptors or on two separate sectors of the same image receptor. Each optic is then assigned to an image receptor or to a specific section of a common image receptor.

Which of these two embodiments is used depends only on the essential solution required. Of course, only half the resolution is achieved if the images of the two optical machines are projected onto two sectors of the same image receptor.

Such an approach clearly requires less modification costs. This is because semiconductor devices are necessarily performed with a high degree of accuracy, and the calibration of the two images is relatively mutually unnecessary.

If a complete solution of technically acceptable image receptors must be exhausted, two image receptors are used. In that case, they must be closely coordinated with each other.

If it is possible to receive images close to the same cut in the workpiece, even if they are carried out sequentially in time with a correspondingly small forward speed, mirrors or translucent mirrors may be placed on the same receiver. Even under the aid of a suitable arrangement of auxiliary means, the photographing can be carried out sequentially from the two optical machines.

In this embodiment, mechanical modifications and dynamic expenditures are avoided.

Other embodiments of the method of the present invention are described in detail in the conventional claims, and those skilled in the art can easily embody these embodiments based on the above description of the method of the present invention. .

The object of the invention is to provide correct data on the entire geometry and position of a cut of the workpiece within the measurement range, independent of its actual structure, as a single measurement result, in which case: A method of the type mentioned at the outset, in which the method is small and requires that the mechanical parts of the lobster wind measuring device can be manufactured without moving parts and thus located close to the welding arc, It consists in providing.

According to the invention, the task is to image a cut of the workpiece to be measured from two different viewing positions on two image receptors, so that the same cut of the workpiece is displayed on these two image receptor images. This is achieved by projecting mutually offset images of the workpiece, and by measuring the geometry of the workpiece and its position with an observation system from the position of the two images of the workpiece and the distance between the prominent positions. ing.

In the method of the invention, dirty surfaces and structures of the workpiece do not pose any hindrance and even support the functioning of the system. Since this crop is being operated around weld seams, the presence of dirt is inevitable.

Furthermore, no equivalent special light source is required; adequate illumination of the workpiece, such as is usually sufficient by arc or room lighting, is sufficient. Additional lighting in dark rooms can be created without any high demands.

According to the invention, a cut in the workpiece surface (for example, a transverse groove in a weld seam) is created by means of two optical machines that can be arranged at a certain distance and at a certain angle to each other.
Images are received on two semiconductor image receptors. All noticeable surface positions on the corresponding image are then displaced relative to each other, independent of the spacing of the two optics. On each of the two images, noticeable incisions and locations of the workpiece surface can be seen resulting from the different structure of the surface, the intensity of the illumination as well as surface imperfections. 2
From the image positions of all the conspicuous points of the surface on the two image receptors, their positions are ■! The geometry of all cuts on the workpiece surface within the measuring range is determined from the positions of all calculated surface points.

The measurement of the position of points noticeable on the surfaces of the two image receptors can also be carried out in this method by means of a correlation method. In this case, the image of the structure detected on the image receptor is brought into compensation by means of cross-correlation, so that its position is determined in a co-invention system of measurement units. With a corresponding miniaturization of the correlation window, arbitrarily small structures can then be analyzed. Hereinafter, the method of the present invention will be further explained in the case of a linear imager with reference to the drawings.

For planar image receptors, the approach remains the same. The individual calculations then extend over two dimensions.

FIG. 1 is a schematic explanatory diagram of the basic arrangement when carrying out the method of the present invention. FIG. 2 shows various images obtained when implementing the method of the present invention. Figures 3 and 4 are various diagrams in the process of implementing cross-correlation using the light intensity function;
The figure shows a complete layout for carrying out the method of the invention.

From FIG. 1, it can be seen that the workpiece 5 having a joint 6 (joint groove) is
It can be seen that the two optics 9 and 10 are observed in mutually convergent directions (arrows 7 and 8). The optics 1 and 2 then observe a line 11 on the surface of the workpiece 5 running transversely to the seam 6, the cuts 1 to 4 being imaged side by side.

In the embodiment chosen in FIG. 1, for each cut 1 to 4,
It can be seen that there is a noticeable structure on the workpiece surface.

The two optical machines 9 and 10 are connected to the image receptor 1 attached to them.
On 3 and 14, images of cuts 0 to 4 are shown. In this case, the conspicuous structures 12 are imaged in a different manner. The spacing between the two images of structure 12 is a measure of the spacing of that structure from a previously chosen neutral plane ε. In image 2, it can be seen that the interval δ is 0. This is because the structures 12 in the cut 2 of the workpiece surface lie in the neutral plane ε.

Image receptors 13 and 14, shown schematically in FIG.
5 lead photodiode series. This image computer 15 sends the investigated geographic data via the surface of the workpiece 5 to a central computer 18 . The central computer 16 then evaluates it in order to modify the trajectory of the welding pistol 17 and the welding parameters.

3 and 4 illustrate how the light intensity function (b(x)) can be used to evaluate the data coming out of the image receptor in the example of cross-correlation. FIG. 3a shows the light intensity function b(x)
(x) or Geome) A diagram is shown in which the light intensity function is applied to the IJ hook position X. By calculating the number T, the copper number h (x) is obtained on FIG. 30. Thus, through the use of differentiation, plotting and thresholding, prominent structures on the workpiece surface are brought into relief. And in doing so, other structures are suppressed. This is done for the receivers 13 and 14 supplied by the optics 9 and 10. In this case, functions h (x) and h (x+8) that are mutually shifted by the distance between the two images δ are obtained. The correlation between these two functions h (x) and h (x+8) yields the cross-correlation function H(x) as shown in FIG. This cross-correlation (H(X)) is then determined by the largest extreme value, which corresponds to a noticeable structure, for example the edge of a workpiece.

[Brief explanation of the drawing]

FIG. 1 is a schematic explanatory diagram of the principle arrangement when implementing the method of the present invention. FIG. 2 shows various images obtained when carrying out the method of the present invention. 3 and 4 are various diagrams in the process of implementing cross-correlation using the light intensity function, and FIG. 5 shows all 5 workpieces for implementing the method of the present invention.
16. Central computer 6, groove 17. Welding Pistol Representative Patent Attorney (8334) Sunagawa Okabe 1 and (no change to the original) 13α calculation 4 fertilization h ζ n ζ ff ni htt now 9) Tekuhi phase l formal letter (self-motivated) March 1986 11E+ Commissioner of the Patent Office Kuro Ill Akio-dono 1, Indication of the case /)i'...7
,' □ 2 Patent application 2 filed on March 10, 1988, title of invention
Method of creating IJ-3, increasing number of claimed inventions Address: Austria, Ny 1140, Vienna, Minorgasose 60 Name: Franz Bokul Nationality: Austria 5. Address of agent: 2-2-39-417 Jingumae, Shibuya-ku, Tokyo Insert a column for detailed description of the invention in the specification. ``That is, the exposures on two offset images of prominent locations on the workpiece surface can be determined by calculations of obviousness and/or by correlation methods, and the positions of those exposures on the two image receptors From this, it is possible to calculate the equality of all the prominent surface points obtained by measurement. From the results, it is possible to calculate the equality of all surface points obtained by measurement. From the results, it is possible to determine whether the geometry of the entire surface can be used as a handicap for a particular surface, or by comparing it with various surface models. , can be easily investigated. If a correlation method is used to measure the parallel cut on two receptacles, it is especially carried out electronically (calculated) and , in which directly adjacent cuts in critical areas of the workpiece surface are correlated. The correlation window can be progressively reduced in order to elucidate the microstructure in the process of evaluating the measurement results. According to another variant of
)) and cross-correlation. In a further embodiment of the inventive method, the arrangement of the two optics can also be chosen for recording offset images. In that case, a projection of the two images onto a single image receptor takes place. During continuous mechanical movement or intermittent movement of the optical system, various image positions are brought to the image receptor in sequence for correlation. The computer detects the correlated image positions and at each instant, from the position of the known optical system, successively more prominent surface structures can be measured in the spatial correlation system of the measuring device. When carrying out the method of the invention in practice, only a linear cut (transverse to the seam groove) in the workpiece surface is made by two optical machines onto two separate linear image receptors (or It is suitable for use in automatic arc welding when the images are projected on two separate sectors for the evaluation of a common image receptor or on the same image receptor alternating in time. From the position above the shifted image, the position of the image receptor of the noticeable surface position is calculated at those points on the workpiece surface, and from every calculated individual point the profile and position of the seam groove are measured. Due to the intermittent repeated measurements and forward movement, the welding burner (and measuring system) can be located at the beginning of the seam groove by evaluating the individual measurement results and can perform the welding. During the welding process, the welding parameters are controlled according to the progression of the groove and adapted to the respective measured seam profile, with the technical The results (symension, position and shape of the welded seam) were also investigated,
In a further control, correction factors can optionally be taken into account. Such measurement and control configurations can also be advantageously applied to other tasks. Instead of the step-by-step method, the present invention allows important areas of the workpiece or large sections thereof to be imaged at once. However, it is also easily possible to measure the position of all noticeable points on the workpiece surface by calculation methods. An image that can be evaluated is collected through a ring-shaped lens on a sensor placed in a line pattern, so that a ray of light from a schematic line on the surface is obtained. The principle has the decisive advantage that, when used close to the welding arc, the optics can be placed behind a small slit in the container and protected against welding jets. Protected from contamination by blowing (or by intermittent pressurized air during the welding process). Obtains measurable measurement results regardless of the different irradiation intensities caused by room light, additional light beams or arcs. Therefore, different methods can be chosen depending on the difference in image reception.A sensor that integrates the light is the most suitable.It also allows the light intensity to vary depending on the extreme fluctuations due to different reception times. The average brightness is imaged. The reception time is then controlled linearly by the respective light intensity. With other sensors, a summation of many individual measurements is formed and used for evaluation." Procedural amendment (method) % formula % 2, name of invention (location and geometry of workpiece surface) +3- method of creating 3,
Person making the amendment Relationship to the case: Patent applicant address Austria, Ny 1140, Vienna Minorganose 60 Name Franz Pokliki Japanese language Austria 4, agent address 2-2-39-4176 Jingumae, Shibuya-ku, Tokyo, subject of amendment 1) Appropriate drawings 2) Document certifying authority of agency 7, contents of amendments 1) As shown in the attached document 2) Document certifying authority of agency is dated March 310, 1988, and has already been submitted.

Claims (1)

[Claims] (1) The workpiece surface, or in particular its cut, is imaged on an image receptor from two different observation points, and mutually discrepant images of the workpiece are produced on the image receptor; in each of the two images to determine the location of the noticeable image resulting from the observed surface structure of the workpiece or its defects, and each image of the noticeable point of the workpiece surface on one image receptor; to the corresponding image on the second image receptor, and thus the position of that point on the workpiece in the room coordination system of the image receptor from the position on the two image receptors of their respective salient image positions. Finally, from the position of each conspicuous point on the workpiece,
Method for detecting the position and geometry of a workpiece surface, in particular a seam groove, in an automatic arc welder, characterized in that the entire geometry of the observed workpiece and its position of the observation system are determined (2) 3. A method according to claim 1, characterized in that the workpiece surface is received on two separate image receptors or on separate sites on a common image receptor via two separate optical machines. 2. The method according to claim 1, wherein images are received intermittently on the same image receptor through separate optical machines and images taken one after the other in time are used for evaluation. (4) When only the cut edge of the workpiece surface is imaged on a linear image receptor, and the measurement system is used to measure a workpiece having a groove or groove-like surface, 4. A method as claimed in claim 1, characterized in that the method is arranged such that it is placed in a transverse direction. (5) The method according to claims 1 to 4, characterized in that the light intensity values occurring on the image receptor are pre-processed by differentiation and subsequent threshold comparison of salient image locations. Corresponding points on the image are found by an electrical computer using recognition methodological principles or mathematical methods of correlation or parameter integration; Claims characterized in that two mirrors, each directed onto an image receptor, are found by moving and from the position of the mirrors the position of the found point with respect to the sensor coordination system is determined. One of the methods of ranges 1 to 5 (7) From the calculated positions of the noticeable surface points, the structure of the entire surface at the measuring position is measured in an electronic computer, where the model structure is optionally determined. is given in advance, and the structure is identified by filtering and differentiation in order to measure the edge position and plane and calculate the change. (8) The geometry of a large cut on the surface of the workpiece or the entire workpiece is measured from the results of measurement cycles successively carried out in time and space. (9) When using this method in arc welding, it is necessary to ensure that the workpiece is illuminated by the arc and that, under operating conditions such that the arc does not burn, the light source is replaced by another light source that is used for illumination. The method (10) according to any one of claims 1 to 8, characterized in that the imaging time is controlled by irradiating the image receptor until it reaches a desired average amount of light; The method according to any one of claims 1 to 9, characterized in that subsequent imaging at the same time interval has accumulated up to an estimable light intensity value.