JPH0426951B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to improvements to industrial robots, and particularly to robots that can easily simplify teaching operations.

When multiple work lines on a workpiece have the same pattern, teaching can be performed on that reference work line, and teaching information about the reference work line can be transferred to other work lines by coordinate transformation, which would make teaching work easier. This can be simplified. However, conventionally in such cases, it is necessary to accurately capture information on three reference points for these work lines, and when the three reference points for these work lines are superimposed, it is impossible to ensure that these points match perfectly. This is not always the case, and in such cases, the algorithm is not only inaccurate, but also complicated, such as overlapping the images so that the difference is as small as possible.

Therefore, in this invention, the reference points mentioned above do not necessarily have to coincide with each work line, and it is sufficient that these reference points can specify the local coordinates for the target work line, and the coordinates of these local coordinates are The above problem was solved by performing the conversion.
The aim is to provide industrial robots.

Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is an overall view including a rectangular coordinate welding robot RO, which forms the background of this invention and is effective in carrying out this invention, but this invention is not limited to this embodiment.

1 is a well-known rectangular coordinate (x,
y, z) is the vertical axis configured at the terminal of the robot RO. In this example, these rectangular coordinates (x,
y, z) is the absolute coordinate system.

Reference numeral 2 denotes a first arm supported at the lower end of the vertical shaft 1 so as to be able to rotate α around the shaft 1.

A second arm 3 is supported at the tip of the arm 2 by an oblique shaft 3a so as to be rotatable β. At the tip of the second arm 3 is a processing tool (in this embodiment) as an end effector.
A gripping tool 3b for gripping the MIG welding torch T) is provided.

and the central axis of shaft 1, shaft 3a and torch T
The TCs are constructed so that they can intersect at one point P.
Furthermore, the torch T is configured such that its welding operating point can coincide with the point P. Thus, by controlling the angles α and β, the attitude angle θ and the turning angle (the so-called Euler angle) of the torch T with respect to the vertical axis 1 can be controlled. It is assumed that the angle θ is variable up to 90 degrees, and that the angle can be rotated 360 degrees, that is, the angle can be rotated all around.

4 is a known welding power supply device. The device 4 includes a spool 4a that winds up the consumable electrode TW of the torch T, and a welding power source 4 between the electrode TW and the workpiece WK.
b. The device 4 also includes a detection power source 4c. The power source 4c has a voltage of about 100 to 200 V and a current limited to a small current, for example. A current sensor 4d is directly connected to the power source 4c. Further, the power source 4c and the sensor 4d are connected in parallel to the power source 4b via a switching means 4e.

Reference numeral 5 denotes a known computer as a control means for the entire embodiment. Computer 5 has
Including CPU and memory.

A power source 4b, a sensor 4d, and means 4e are connected to the bus line B of the computer 5.

The bus line B is further connected to the x-axis servo system Sx of the robot RO. x for servo system Sx
It includes the shaft power Mx and an encoder Ex that outputs its position information. Similarly, a similarly constructed y-axis servo system Sy, z-axis servo system Sz, α-axis servo system Sα, and β-axis servo system Sβ are connected to the bus line B.

The RE is a remote control panel and is equipped with a group of manually operated snap switches. If you flip the snap switch for each x, y, and z control axis to the "U" side, the position information for that control axis will increase (in the direction away from the origin), and if you flip it to the "D" side, it will move in the opposite direction. The end effector is configured to move. Further, the snap switches corresponding to the respective control angles θ and θ are also configured so that when turned to the "U" side, the torch T is rotated in a direction away from the origin, and when pushed to the "D" side, the torch T is rotated in the opposite direction.

In this case, it is necessary to perform coordinate transformation between the information on the angles α and β and the information on the angle θ, but this coordinate transformation is not particularly the gist of the present invention, so it will not be described in detail.

The operation panel RE is also provided with a speed command rotary switch SV. A mode changeover switch SM is also provided, and is configured to switch between manual mode M, test mode TE, and auto mode A. SE is a designation switch, and by switching up in the figure and operating the up-down switch SU, the sensor menu number (SM No.) is displayed and selected. Furthermore, by setting this switch SE to the left as shown in the figure and operating switch SU, linear interpolation "L", circular interpolation "C", and sensing "S" are selected and displayed in this order. There is. Furthermore, switch the switch SE to the right in the diagram, and
When SU is operated, the welding condition number (WNo.) is displayed and is to be selected. Furthermore, the operation panel RE is equipped with a start switch STA. The functions of the switch STA will be explained in detail in the explanation of the operation described later. These switches are then connected to bus line B.

In addition, when Switch SV is in manual mode,
By pressing the head, the moving speed of the torch T in the manual mode is stored as a constant value.

The effects of the above-mentioned embodiments will be described below. Please also refer to Figure 2 et seq.

Now, the workpiece WK is placed on the table Wo as shown in the figure.
Assume that blocks W ₁ and W ₂ of the same shape and size, which are obtained by obliquely cutting a part of a rectangular parallelepiped, are conductors that are temporarily attached to each other with the right and left sides reversed, that is, reversed. And now block W ₁ point P ₁₁ to P ₁₂
Fillet welding is performed in a straight line from point P ₁₃ to point P ₁₄ , and similarly for block W ₂ , the corresponding points P ₂₁ to P ₂₂ and from point P ₂₃ to P ₂₄ are applied.
The space between them shall be welded. The aforementioned fillet weld line for block _W1 is defined as the "reference work line" in this embodiment. In order to specify the position of this reference work line, the required number of points on the surface of block W ₁ are set as feature points, SP ₁₁ , SP ₂₁ , SP ₃₁ ,
Detect and set a total of 6 points: SP ₄₁ , SP ₅₁ , and SP ₆₁ . However, points SP ₂₁ , SP ₃₁ , and SP ₄₁ are not on a straight line;
Furthermore, points SP ₅₁ and SP ₆₁ are not included on a straight line perpendicular to the plane determined by points SP ₂₁ , SP ₃₁ , and SP ₄₁ . In other words, it should be understood that once these six points are determined, the position and orientation of block _W1 are determined, and therefore the position of the reference work line is also determined.

(1) First, the operator operates switches SE and SU to select sensor menu number SM. Now set this menu number to "10",
The contents of this menu are as follows: ``In order to determine a local rectangular coordinate system using three mutually orthogonal planes of a rectangular parallelepiped, one point on each of the three planes,
The positions of a total of six points, two points and three points, are sensed, and a local rectangular coordinate system fixed to a rectangular parallelepiped is specified from these positional information. ", and furthermore, "point position sensing on each plane is based on the direction of the torch located in front of that point (torch axis).
This is executed by moving the torch T in the direction of TC (toward the operating point P of the torch T) and capturing the information of each axis at that time by inputting a signal from the sensor 4d. ”. Note that the direction in which the torch is moved during sensing may be modified in addition to the above-mentioned direction, such as in a predetermined direction, and the direction may be the direction of the torch projected onto a plane.

(2) Next, in order to sense the positions of the six points on block _W1 , the operator, for example,
In order to sense SP ₁₁ , operate switch SM to "M" or manual mode,
Furthermore, operate the switch SW to manually control the position and orientation of the torch T, and move the operating point P to the sensing start position of SP ₁ and change the orientation to the point on block W ₁ , as shown in Figure 2 T 1 _. Make it almost perpendicular to the plane where SP ₁₁ is located. Then, operate the switches SE and SU to select "S" and then operate the switch STA. The computer 5 uses menu number "10" to output the position (or angle) information of each control axis of the robot RO at this time based on the output of each encoder of each control axis.
Import menu number "10" and sensing command "S". At the same time, the computer 5 switches the means 4e (the state indicated by the two-dot chain line in the figure) and executes the sensing operation. That is, move the torch T in the sensing direction,
The tip of the consumable electrode TW of the torch T is blocked W ₁
The computer 5 stops the movement of the torch T in response to a signal from the sensor _4d corresponding to the energization due to proximity to the point SP 11 on the surface of the torch T, and captures the position information at that time as that of the point SP ₁₁ .

In the same manner, the operator points the torch T at points SP ₂₁ to SP ₆₁ and points SP ₂ to SP _6.
position, and switch at each position.
By operating STA, computer 5
takes in the respective pieces of information similar to those described above (steps ST ₁ and ST ₂ in FIG. 4).

(3) Next, the operator uses the same manual operation to determine the position of welding point _P11 , the speed of movement up to that point, the position of welding point _P12 , the welding conditions and speed of movement up to that point, linear interpolation commands, and the welding The position of point P ₁₃ ,
The moving speed up to that point, the position of welding point _P14 , the welding conditions and moving speed up to that point, direct interpolation commands, etc. are taught for each step by operating the switch STA. The computer 5 takes in these as part of the program (step _ST3 ).
Of course, these series of teachings can be performed without selecting the sensing command "S".

(4) If the work line is not the standard work line or the work line with the same pattern as described above, perform teaching using a conventionally known method (Steps ST ₄ and
_ST5 ).

(5) If the work line has the same pattern as the standard work line,
That is, regarding the working line of block _W2 in this embodiment, teaching is performed together with sensing commands at sensing command points corresponding to the six sensing command points taught in block _W1 . That is, block
SP 201 , SP ₂₀₁ on the top surface of block W ₂ , corresponding to the three locations SP ₂₁ , SP ₃₁ and SP ₄₁ on the top surface of W ₁
3 locations SP ₃₀₁ and SP ₄₀₁ , 2 locations SP ₅₁ on the front of block W ₁ (looking in the direction of the arrow)
and SP ₆₁ , SP ₅₀₁ and SP ₆₀₁ in the corresponding front of block W ₂ , block
In order to sense a total of six points, one point SP ₁₁ on the right side of block W 1 and _one point SP ₁₀₁ on the left side of block W ₂ (because the left and right sides are reversed), blocks W ₁ and Similarly, a sensing command at a sensing command point (not shown) further away from that point is taught, but with the menu number "11" (that is, the number corresponding to "10" reversed left and right in the sense of FIG. 2), and the computer 5 takes them in (step ST ₆ ). In these cases, each point on _{the 2nd} side of the block W does not need to correspond to an exact position to correspond to each point on _{the 1st} side of the block W; in short, it is sufficient to find the points on that side (however, Point SP ₂₀₁ ,
It should be noted that SP ₃₀₁ and SP ₄₀₁ are not on a straight line, etc.).

(6) Although the teaching in the embodiment illustrated above ends, if there is any other conventional teaching that is normally performed, it goes without saying that it is also performed. When the teaching is finished in this way, the operator turns on the switch SM.
, select the test mode, execute a conventionally known test, correct any mistakes, and complete the teaching of the user program.

(7) Next, in order to continuously execute the user program taught in this way, the operator sets the switch SM to auto mode and operates the switch STA.

(8) The computer determines whether the sensing command "S" is included in each step of the user program (step _ST7 ). If not included, a conventionally known normal command is output (step _ST8 ).

(9) The following explanation assumes that the sensing command "S" is included. That is, by sensing command "S" and menu number "10",
The computer 5 calculates the points SP ₁₁ , SP ₂₁ , SP ₃₁ , which were obtained and captured in the teaching described above.
From the point position information of SP ₄₁ , SP ₅₁ , and SP ₆₁ , determine a local rectangular coordinate system O ₀ −ξ ₀ η ₀ ζ ₀ along the intersection line of each orthogonal plane determined by these points, and further The position of the torch T is controlled at the taught sensing start point SP ₁ , and the means 4e
After switching, the torch T is moved in the same direction as the teaching described above, and the position information of the point _SP'11 on the block _W1 is taken in. In the same manner, the position information of each point SP' ₂₁ to SP' ₆₁ is taken in (step ST ₉ ). Furthermore, the local coordinate system O′ ₀ −
Define ξ′ ₀ η′ ₀ ζ′ ₀ .

Then, the computer 5 uses the coordinate system O ₀ −
A coordinate transformation matrix 〓 ₀ between ξ ₀ η ₀ ζ ₀ and O′ ₀ −ξ′ ₀ η′ζ′ ₀ is calculated (step ST ₁₀ ). The coordinate transformation algorithms for industrial robots, including the algorithm for calculating this coordinate transformation matrix, are well known (for example, Kyoritsu Shuppan's bit,
(Published July 1976, Volume 97, p. 76 onwards), so it will not be discussed in detail.

Note that if the reference local coordinate system O ₀ −ξ ₀ η ₀ ζ ₀ is already calculated and determined at the time of teaching, the calculation time during playback can be shortened.

Furthermore, the computer 5 obtains position information P ₁₁ (ξ ₀₁₁ , η ₀₁₁ , ζ ₀₁₁ ,
θ ₀₁₁ , ₀₁₁ ) or P ₁₄ (ξ ₀₁₄ , η ₀₁₄ , ζ ₀₁₄ , θ _01
Four ,
₎ to the coordinate system O′ ₀ −ξ′ ₀ η′ ₀ ζ′ ₀ , the coordinate transformation matrix 〓 ₀ is used to calculate the coordinate transformation (step ST ₁₁ ). In the above explanation, if the workpiece WK is the same workpiece used for teaching, the matrix ₀ becomes the unit matrix. Generally, the matrix ₀ is calculated as a certain value depending on the individual differences between the workpieces WK and the mounting position error.
Furthermore, in order to determine the position information on the work line that undergoes coordinate transformation, although the details are not shown in the figure, this can be done by providing an operation switch on the operation panel RE that allows the information to be input into the computer. This shall be done by known means.

(10) The position information in the absolute coordinate system of points _P11 to _P14 taught and captured in this way is corrected, and this corrected position information is output as sequential command position information (step
_ST12 ).

(11) In the next step of the user program, in the program for block W ₂ , the sensor menu number becomes "11", and since it is reversed with respect to the work line in block W ₁ , the computer 5 corresponds to it. Therefore, points SP ₁₀₁ to SP ₆₀₁ on the surface of block W ₂
The point position information is sensed in the same manner as described above. Then, the local coordinate system O ₂ as shown in Figure 3
−ξ ₂ η ₂ ζ ₂ is obtained, and the coordinate transformation matrix 〓 ₂ between O ₀ −ξ ₀ η ₀ ζ ₀ is calculated. Then, in the same manner as described above, by converting the coordinates of the position information of points P ₁₁ to P ₁₄ using matrix 〓 ₂ ,
Point position information of the corresponding points P ₂₁ , P ₂₂ , P ₂₃ , and P ₂₄ can be obtained.

(12) The above calculations and commands are executed to the end of the user program step, and the process ends.

As is clear from the above explanation, when teaching a plurality of blocks W ₁ and W ₂ having the same pattern of work lines to automatically weld, even if the placement accuracy of the blocks W ₁ and W ₂ is poor, each work If the local coordinate system specific to the line is specified, the actual work trajectory can be created by simply teaching the reference work line without having to teach the work lines of uncertain blocks W ₁ and W ₂ individually. is possible,
Teaching work is simplified.

The present invention is not limited to the above-mentioned embodiments, and the following modifications are also possible.

(b) The mechanical configuration of the robot RO may be in a coordinate system other than the rectangular coordinate system. However, as a coordinate system that can specify the position of a work line on a workpiece and perform coordinate transformation, it is practically desirable to use rectangular coordinates or oblique coordinates in which the transformation matrix is linear. It is necessary to add a coordinate transformation algorithm to the system.

(b) In addition to the welding line as in this embodiment, the work line on the workpiece may correspond to the movement line of the end effector of the robot, such as the cutting line or the movement trajectory of the gun for painting. can.

(c) Six sensing positions on the workpiece are required for specifying the work position if the workpiece is a rectangular parallelepiped as in the above embodiment.
If the workpiece has a different shape, a feature point that can specify the position of the workpiece may be used as the sensing position.

(d) In this embodiment, the welding torch itself is used as a sensor to sense the position on the workpiece, but other known contact or non-contact sensors may be provided at the end of the robot. Good too. If the sensing position is different from the working point of the end effector, it is sufficient to correct it accordingly.

(E) The program for executing the above-mentioned operations of the computer 5 may be such that, for example, teaching data is divided into several blocks and these blocks are edited to create an actual work program.

(f) In the above embodiment, block W ₂ was inverted with respect to W ₁ , but if this is not inverted,
Coordinate transformation may be performed using the same sensor menu.

(g) In addition, it is also possible to replace each component with an equivalent within the scope of the technical idea of this invention.

As described above, the present invention has the following unique and remarkable effects.

() In order to specify the position of the work line of the same pattern, the feature points on the workpiece that are sufficient to specify the local coordinates are detected, so there is no need to make the correlation between the detection points of the workpieces strict. Not only does the teaching work become easier, but also the coordinate transformation between the workpieces becomes more accurate.

() Since coordinate transformation between local coordinates was used to transfer the work line position information, it can be easily carried out even when the work lines are reversed.

() Errors in the work itself are automatically corrected by the sensing operation.

() Easy to configure and implement.

[Brief explanation of drawings]

The drawings all show one embodiment of this invention, and the first embodiment is shown in the drawings.
The figure is an overall block diagram including a perspective view, and Figure 2 is the first
3 is a perspective view for explaining the operation, and FIG. 4 is a flowchart. x, y, z... Absolute coordinate system, ξ ₀ , η ₀ , ζ ₀ and ξ ₂ , η ₂ , ζ ₀ ... Local coordinate system, T... Welding torch (end effector), WK... Workpiece, RO...
Robot, P ₁₁ , P ₁₂ , P ₁₃ , P ₁₄ , P ₂₁ , P ₂₂ , P ₂₃
and P ₂₄ ...start and end points of welding line (work line), 5
...computer.

Claims (1)

[Scope of Claims] 1. In an industrial robot equipped with means for detecting workpiece position information using absolute coordinate system information using the end effector itself or a sensor replacing it, a plurality of workpieces having the same pattern of work lines are provided. Work position command information with respect to the reference work line of one of the workpieces, and necessary points on the workpiece surface in order to specify the position of the reference work line of the workpiece in a local coordinate system specific to the reference work line. A means for incorporating sensing command information for detecting a point and recording its position into one step of a user program; and means for detecting necessary points on the surface of each of the other workpieces as the feature points and importing sensing command information for recording the positions into other steps of the user program, in order to specify the posture. The industrial robot further comprises means for converting coordinates of work position command information for the reference work line to the work line of the other workpiece based on the captured information.