JPH0424147B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an automatic weaving device for industrial robots, and particularly to an automatic weaving device that can easily weave even complex-shaped workpieces.

Among industrial robots, those that perform welding, painting, polishing, grinding, etc.
Weaving is necessary to prevent other processing irregularities.

Weaving involves adding to a trajectory without weaving a shift amount that changes periodically in the direction perpendicular to the trajectory without weaving (generally, a sine curve (single harmonic motion), trapezoidal, or triangular pattern). When welding the fillet part A of a ⊥-shaped workpiece H as shown in Fig. 1, the workpiece H is
When the workpiece H is placed at an angle as shown in Figure 1, it is preferable to do it in the horizontal direction (arrow α ₁ ), and when the workpiece H is placed as shown in Figure 1B, it is preferable to do it in the horizontal direction among the countless original right angle directions. The direction from the horizontal plane must be appropriately selected depending on the situation, as it is said that it is preferable to do it in the 45° direction (arrow α ₂ ) from the horizontal plane.

Conventionally, an operator instructs an industrial robot on the appropriate weaving direction, but while this may be fine if the workpiece shape is simple and there are few changes in the weaving direction, as the workpiece shape becomes complex and the weaving direction changes. If there are many angle changes from the horizontal plane, etc., it becomes difficult to give sufficient instructions. For example, in the case of workpiece N shown in Fig. 4, the angle of the weaving direction from the horizontal plane, etc., as shown by welding line R', must be changed from time to time, so detailed instructions must be given for each small section. Conventional methods of teaching the angular direction of weaving from a horizontal plane, etc., have been extremely inefficient.

The object of the present invention is to provide a device that automatically calculates a weaving direction that is perpendicular to the original locus and has an appropriate angle from a horizontal plane, etc., and creates a weaving pattern based on this weaving direction for weaving. This eliminates the burden on the operator of instructing the weaving direction and makes it possible to adequately handle weaving of workpieces with complex shapes.

That is, the present invention provides an industrial robot that controls the movement of a tool and the posture of a tool according to a predetermined procedure during playback based on the data at the time of teaching. a moving direction vector calculating means for calculating the original moving direction vector of the tool from processing data that has been subjected to predetermined processing such as processing, acceleration/deceleration processing, curving processing, etc.; an axial direction vector calculating means for calculating an axial direction vector of the tool from partial data; an outer product vector calculating means for calculating a cross product vector from the moving direction vector and the axial direction vector; and weaving the direction of the cross product vector. Weaving control means for controlling the movement and/or attitude of the tool to perform weaving in which the tool is shifted at an appropriate weaving frequency and weaving amplitude perpendicular to the original moving direction based on the weaving direction. The present invention provides an automatic weaving device for industrial robots having the following points in its configuration.

In the above configuration, the data or processing data at the time of teaching is, in other words, the data of the cutting edge trajectory of the tool when there is no weaving, for example, the data at the cutting edge of the tool when there is no weaving.
In the figure, the joining line R corresponds to the welding line when weaving is not performed. Further, the movement direction vector is a vector in the tangential direction of the locus. The calculation itself for obtaining the vector in the tangential direction from the trajectory data may be performed using a conventionally known tangential calculation method.

In addition, in the above configuration, the data at the time of teaching or the partial data regarding the tool posture of the processing data is, in other words, the angle data of the tool axis when there is no weaving, and the tool axis direction vector can be easily determined from that data. Calculated.

In addition, as a result of the automatic weaving device operating,
The obtained welding line becomes R' in FIG.

Furthermore, the teaching work for obtaining the data regarding the movement of the tool and the data regarding the posture may be performed by PTP teaching shown in the embodiments described later, or by so-called CP teaching.

Now, in Fig. 2, the trajectory U and each point P ₁ , P ₂ , P ₃
, i.e., movement direction vectors L ₁ →, L ₂ →, L ₃ →, and their respective points P ₁ , P ₂ ,
Preferred weaving direction at P ₃ (arrow β ₁ ,
β ₂ , β ₃ ), but as can be easily understood from this figure, weaving directions (arrows β ₁ ,
β ₂ , β ₃ ) are the moving direction vectors L ₁ →,
It is preferable that it is in a plane perpendicular to L ₂ → and L ₃ →.
On the other hand, as can be easily understood from Fig. 1 a and b, the weaving directions (arrows α ₁ and α ₂ ) are in a plane perpendicular to the axial direction vectors T ₁ → and T ₂ → of the tool K, respectively. is preferable. Therefore, in the end, it is preferable that the weaving direction be perpendicular to both the moving direction vector and the tool axis direction vector, but by calculating the cross product vector in the configuration of the present invention, it is possible to determine the weaving direction perpendicular to both. can be suitably determined.

The reason why an appropriate weaving direction can be obtained from the tool movement direction vector and the tool axial direction vector is that both the tool movement direction vector and the tool posture direction vector depend on the shape of the workpiece such as the weld line. This is because the optimal direction is uniquely determined.

If the weaving direction can be obtained, weaving can be performed based on the given weaving frequency and weaving amplitude data in a conventional manner.

Next, embodiments embodying the present invention will be described with reference to the accompanying drawings from FIG. 3 onwards. Here, FIG. 3 is a configuration explanatory diagram of a welding robot according to an embodiment of the present invention, FIG. 4 is a perspective view of a workpiece for welding two pipes, and FIG. 5 is a diagram showing a moving direction vector and an axial direction vector. A vector diagram showing the relationship with the weaving direction vector, FIG. 6 is a flowchart of the main part of the control procedure in the apparatus shown in FIG. 3. In addition, the control procedure (step) number is
Represented by S1, S2, S3, ….

The apparatus shown in FIG. 3 is an arc welding robot apparatus 1, which consists of an articulated industrial robot 3 holding a welding torch 2 as a working tool, a control section 4 including a computer, and an operation panel 5. There is.

FIG. 4 shows an example of a workpiece, in which a pipe _n2 is joined to a pipe _n1 at a right angle, and the joining line R is a hollow curve. The welding line R' is obtained by adding weaving to this joining line R.

Work N by arc welding robot device 1
To weld, the operator drives the industrial robot 3 using the operation panel 5, guides the welding torch 2 to discrete welding positions (teaching positions), and
By storing the position data, the data (θ _1i , θ _2i , θ _3i , θ _4i , θ _5i ) as well as teaching
Regarding weaving, the weaving frequency and amplitude D are taught or stored in advance.
Further, the positioning time interval Δt during reproduction, the reproduction speed V, and an instruction to perform circular interpolation or linear interpolation are taught or stored.

The control unit 4 receives the taught joint angle data (θ _1i ,
Data (X _i , Y _i , Z _i , θ _4i , θ _5i ) is obtained by forward coordinate transformation from θ _2i , θ _3i , θ _4i , θ _5i ), and further coordinate transformation is performed to convert the data related to the tool posture into the direction. Obtain data (X _i , Y _i , Z _i , E _i , F _i , G _i ) contained in the form of cosines.
Next, interpolation point data (X _j , Y _j ,
Z _j , E _j , F _j , G _j ) are calculated. These coordinate transformations and calculations are similar to conventionally known processing.

Next, the procedure for calculating the weaving direction from each of the above data will be explained with reference to FIG.
Assume that the data of (0≦m, A ₀ is the first teaching point) is (Xm, Ym, Zm, Em, Fm, Gm),
The data of the next point An is (Xn, Yn, Zn,
En, Fn, Gn) (ie, n=m+1).

First, as shown by S1 in Fig. 6, the moving direction vector Ln→ at the point An is determined, but since the data regarding the movement indicates the coordinates of that point as described above, it is approximated by equation (i). A sufficiently practical movement direction vector Ln→ can be obtained.

Ln→=(Xn−Xm, Yn−Ym, Zn−Zm)
...(i) Next, as shown in S2, the axial direction vector Tn→ of the tool at the point An is determined. The data regarding the orientation of the tool is the welding torch 2 at that point.
Since it is the direction cosine of , it directly represents the axis direction vector, and is expressed by equation (ii).

Tn→=(En, Fn, Gn)...(ii) After that, as shown in S3, the cross product vector Wn→ of the moving direction vector Ln→ and the axial direction vector Tn→
get. The cross product vector Wn→ is obtained by equation (iii).

Wn→=Tn→×Ln→ =(Fn・(Zn−Zm)−Gn・(Yn−Ym), Gn・(Xn−Xm)−En・(Zn−Zm), En・(Yn−Ym)− Fn・(Xn−Xm)) …(iii) Furthermore, as shown in S4, weaving amount Cn
, which can be obtained by equation (iv) if weaving is performed in a sinusoidal manner.

Cn=D·sin(2πf·Δt·n) (iv) Next, as shown at S5 in FIG. 6, the point An' where the point An is displaced when weaving is applied is determined. To do this, first divide each component by the sum of the squares of each component of the cross product vector Wn→, and then multiply each component by Cn to obtain a weaving vector Wn→′ with the length of the cross product vector Wn→ being Cn. , then each component of the weaving vector Wn→' may be added to each component of the data (Xn, Yn, Zn) regarding the movement at the point An.

To carry out weaving, the data at point An′ is used as data regarding movement, and the data regarding posture at point An is treated as data regarding posture.
From these, industrial robot 3 is created by inverse coordinate transformation.
joint angle data (θ _1o ′, θ _2o ′,
θ _3o ′, θ _4o ′, θ _5o ′) and control the industrial robot 3 based on these values. Note that weaving is performed by moving the welding torch 2 in parallel without changing the attitude of the welding torch 2.

As another example, if the interpolation point data is calculated by circular interpolation, instead of calculating the movement direction vector approximately using equation (i), data on the center and radius of the circle obtained by circular interpolation is used. An example of this method is to accurately obtain a vector in the tangential direction and use this as the moving direction vector.

Still other embodiments include one in which not only the data related to movement is corrected using the data at point An', but also the data related to posture, or only the data related to posture. In these cases, weaving is performed by the deflection of the axis of the welding torch 2 or only by the deflection, and the weaving is performed by the deflection of the axis of the welding torch 2.
Weaving is performed by specifying the deflection direction and deflection angle.

Furthermore, as another embodiment, a welding torch 2
Alternatively, a coating gun, a grinder, or the like may be held by the industrial robot 3, and painting, polishing, etc. may be performed while weaving.

In the above embodiment, only so-called PTP teaching was explained, but it is also possible to adopt a so-called CP type teaching method, and in this case, the above-mentioned interpolation calculation procedure is unnecessary. It goes without saying that the industrial robot to be used is not limited to the above-mentioned multi-joint type, but can also be applied to rectangular coordinate type, horizontal multi-joint type (SCARA type), cylindrical coordinate type, and others.

As described above, the present invention provides an industrial robot that controls the movement of a tool and the posture of a tool according to a predetermined procedure during playback based on the data at the time of teaching. a moving direction vector calculating means for calculating the original moving direction vector of the tool from processed data obtained by subjecting the data to predetermined processing such as interpolation processing, acceleration/deceleration processing, curving processing, etc.; an axial direction vector calculation means for calculating an axial direction vector of the tool from partial data regarding the orientation of the tool, a cross product vector calculation means for calculating a cross product vector from the movement direction vector and the axial direction vector, and the cross product vector. Weaving control means for controlling the movement and/or posture of the tool to perform weaving in which the tool is shifted at an appropriate weaving frequency and weaving amplitude perpendicular to the original movement direction based on the weaving direction in the direction of the weaving direction. This invention provides an automatic weaving device for industrial robots characterized by the following: According to this, it is possible to perform weaving by automatically selecting an appropriate weaving direction, so that each teaching It is no longer necessary to teach the weaving direction at each point, and it is now possible to work on workpieces with complex shapes that require the weaving direction to be changed from time to time (for example, welding workpieces with a hollow-shaped weld line as shown in Figure 4). ,
This has the effect of making it possible to quickly perform weaving teaching work for welding workpieces with changing groove shapes, etc., or for workpieces that are placed on a positioner and continuously positioned at the same time as the robot.

[Brief explanation of drawings]

Fig. 1 is a side view showing the relationship between the direction of the workpiece and the weaving direction, Fig. 2 is a plan view showing the relationship between the moving trajectory of the tool and the weaving direction, and Fig. 3 is a welding robot according to an embodiment of the present invention. Configuration diagram,
Fig. 4 is a perspective view of a workpiece for welding two pipes, Fig. 5 is a vector diagram showing the relationship between the moving direction vector, the axial direction vector, and the weaving direction vector, and Fig. 6 is a perspective view of the workpiece in which two pipes are welded. This is a flowchart of the main parts of the control procedure. Explanation of symbols, 1... Robotic device for arc welding,
2...Welding arch, 3...Industrial robot, 4...Control unit, 5...Operation panel, N...Workpiece, R...Joining line,
R′...Welding line, Ln→...Movement direction vector, Tn→...
Axial direction vector, Wn→…cross product vector.

Claims (1)

[Scope of Claims] 1. In an industrial robot that controls the movement of a tool and the posture of a tool according to a predetermined procedure during reproduction based on data at the time of teaching, the data at the time of teaching or the data a moving direction vector calculation means for calculating the original moving direction vector of the tool from processed data that has been subjected to predetermined processing; an axial direction vector calculation means for calculating a cross product vector from the moving direction vector and the axial direction vector; An automated industrial robot characterized by comprising a weaving control means for controlling the movement and/or posture of a tool to perform weaving that shifts the tool at an appropriate weaving frequency and weaving amplitude perpendicular to the direction of movement. Weaving device. 2. The automatic weaving device for an industrial robot according to claim 1, wherein the teaching data or processing data is a welding line, and the cutting edge trajectory of the weaving tool is a welding line.