METHOD AND DEVICE FOR CONTROLLING ROBOTS FOR WELDING WORKPIECES

• US 20100152870A1
• Filed: 02/19/2008
• Published: 06/17/2010
• Est. Priority Date: 02/19/2007
• Status: Active Grant

• Current Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V., Universitaet Rostock
• Original Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V., Universitaet Rostock
• Inventors: Zych, Alexander, Pfletscher, Uwe, Wanner, Martin Christoph
• Granted Patent: US 8,338,743 B2
• Time in Patent Office: Days
• US Class Current: 700/97
• CPC Class Codes: B23K 31/02 relating to soldering or welding dip or wave soldering in the manufacture of printed circuits H05K3/34
• CPC Class Codes: B23K 9/0256 for welding ribs on plates
• CPC Class Codes: G01B 11/2518 Projection by scanning of the object
• CPC Class Codes: G01B 21/042 Calibration or calibration artifacts G01B3/30, G01B9/02072 take precedence
• CPC Class Codes: G05B 19/4207 in which a model is traced or scanned and corresponding data recorded
• CPC Class Codes: G05B 2219/36248 Generate automatically machining, stitching points from scanned contour
• CPC Class Codes: G05B 2219/37009 Calibration of vision system, camera, adapt light level
• CPC Class Codes: G05B 2219/37555 Camera detects orientation, position workpiece, points of workpiece
• CPC Class Codes: G05B 2219/37567 3-D vision, stereo vision, with two cameras
• CPC Class Codes: G05B 2219/45104 Lasrobot, welding robot
• CPC Class Codes: G06T 7/521 from laser ranging, e.g. using interferometry; from the projection of structured light

Abstract

The invention relates to a method for controlling robots for welding three-dimensional workpieces, comprising the following steps: positioning and tacking profiles to a plate in order to form the workpiece to be welded, depicting the workpiece by means of a three-dimensional imaging system in the form of three-dimensional pixels, determining the geometric data of the plate and profiles, including the allocation of cutouts and final cutting shapes from the three-dimensional pixels, determining the weld seam data from the geometric data while considering the profile placement lines and the contact lines of profiles, allocating the weld seam data to parameterizable specifications for the weld plan into stored predefined movement patterns of the robot, and into commands for the control of the welding process.

Claims

1-20. -20. (canceled)

21. A method for controlling robots for welding three-dimensional workpieces,

having the following steps:

positioning and tack-welding profiles onto a plate in order to form the workpiece to be welded,

imaging the workpiece by means of a three-dimensional image detection system in the form of three-dimensional pixels,

determining the geometrical data of the plate and of the profiles including assigning cutouts and end cut edges from the three-dimensional pixels,

determining the weld seam data from the geometrical data taking into account the profile placement lines and the contact lines of profiles, and

assigning the weld seam data to parameterisable specifications for the welding plan to stored predefined movement patterns of the robot and also to commands for controlling and regulating the welding process.

22. The method according to claim 21, wherein the image of the workpiece is detected by means of a 3D laser scanner.

23. The method according to claim 21, wherein the image of the workpiece is detected by means of a projection of patterns and at least one camera with assigned evaluation.

24. The method according to claim 21, wherein the determination of the geometrical data is effected by 3D processing.

25. The method according to claim 21, wherein the plate is disposed on a work table, the position of the plate plane being determined during determination of the geometrical data and the measuring points of the workpiece being transformed into a local workpiece coordinate system, the x-y plane of which coincides with the plane of the plate.

26. The method according to claim 21, wherein the pixels of the plate plane and the pixels of the profiles are separated using height coordinates.

27. The method according to claim 26, wherein the contours of the plate are determined in order to delimit workpiece and background.

28. The method according to claim 26, wherein the pixels of the profiles are pre-sorted according to coherent structures.

29. The method according to claim 28, wherein linear profile courses are determined from the pixels of the coherent structures.

30. The method according to claim 29, wherein the profile placement lines and the contact lines of profiles are determined from the linear profile courses and also from the position of the plate plane.

31. The method according to claim 21, wherein the image detection system is calibrated comprising:

a) disposing reference balls in a plane with respect to the workpiece to be welded later, the reference values of the centres of the reference balls in an arbitrary coordinate system and at least one parameter for controlling the calibration method being prescribed,

b) producing a point cloud of the reference balls with the image detection system and determining the actual values of the ball centres in the coordinate system assigned to the actual values,

c) orientating the coordinate system of the reference values to the coordinate system of the actual values,

d) calculating the three-dimensional deviations between the reference and actual values of the ball centres, and

e) determining three-dimensional first correction values for coordinates of the image detection region by interpolation/extrapolation.

32. The method according to claim 31, wherein the point cloud photographed in step b) is corrected with the help of the first correction values and corrected actual values of the ball centres are determined, steps c) and d) are repeated and three-dimensional second correction values for the coordinates of the image detection system are determined by interpolation/extrapolation, the sum of the first and second correction values being formed in order to determine three-dimensional correction values for the image detection system.

33. The method according to claim 31, the image detection system comprising a laser scanner with a scanner matrix as image detection sensor, wherein line and column indices of the scanner matrix which are associated with the determined ball centres are determined between steps b) and c) and in that the three-dimensional first correction values are determined for all pixel elements of the scanner matrix.

34. The method according to claim 31, wherein in order to determine the second correction values, also the step of determining the column and line indices of the scanner matrix which are associated with the corrected actual values of the ball centres is repeated and in that three-dimensional correction matrices are formed for the scanner from the sum of the first and second correction values.

35. The method according to claim 31, wherein the actual values of the ball centres are determined by dividing the point cloud into individual partial point clouds, determining the coordinates of the surface points and calculating the centres of the individual balls with the help of the method of the smallest error squares.

36. The method according to claim 31, wherein the orientation according to step d) is implemented by orientating via the plane spanned by the ball centres and, in this plane, an orientation is undertaken by minimising the spacings between reference and actual values of the ball centres and also via a central ball placed in the centre of the coordinate system of the actual values.

37. A device for controlling robots for welding three-dimensional workpieces, comprising:

an image detection system for detecting a three-dimensional image of a workpiece to be welded which has profiles tack-welded onto a plate, and for delivering image data as pixels,

a geometry recognition module for determining the geometrical data of the plate and of the profiles from the three-dimensional pixels, for determining the weld seam data from the geometrical data taking into account the profile placement lines and the contact lines of profiles, and

a control device which contains a control algorithm for assigning the weld seam data to parameterisable specifications for the welding plan, to stored predefined movement patterns of the robot and also to commands for controlling/regulating the welding process.

38. The device according to claim 37, wherein the image detection system has a 3D laser scanner which delivers a spatial digital model of the workpiece to be welded from several individual photographs.

39. The method according to claim 37, wherein a calibration module for geometrical correlation of the coordinate systems of the robot, the image detection system and the workpiece to be welded is provided.

40. The method according to claim 38, wherein the calibration module contains a calibration algorithm wherein the profile placement lines and the contact lines of profiles are determined from the linear profile courses and also from the position of the plate plane.