JP3840973B2 - Robot teaching data correction method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method for correcting robot teaching data, and more particularly to a method for correcting robot teaching data created by simulation.
[0002]
[Prior art]
In recent years, simulation apparatuses have been used to create teaching data for operating a robot. For example, the robot model is moved on the simulation apparatus, and the teaching position is set, corrected, and changed so as to complete a series of operations without interference with peripheral equipment, and teaching data is created.
[0003]
The simulation for creating such teaching data uses design values for all peripheral equipment, workpieces, robot installation positions, tool shapes attached to robot end effectors, and robot models (mechanism models, control models). And simulate.
[0004]
However, in actual production equipment, the robot cannot be installed at the position as designed, the tool is difficult to install, the encoder zero point is misaligned, or the arm is bent. When the teaching data is reproduced, an error occurs in the teaching position, and the robot may interfere with peripheral equipment or a workpiece.
[0005]
In order to avoid such a problem, the teaching data created by simulation is corrected.
[0006]
The teaching data is corrected by performing a three-point touch-up operation using an actual robot, that is, an actual machine for which teaching data is to be created, and taking the teaching data at that time into the simulation apparatus. Corrects the position of the robot model.
[0007]
In addition, as a correction method for correcting the motion of the robot model itself, the actual motion of the actual robot such as deflection generated in the actual robot is parameterized and used as a parameter for defining the motion in the robot model. The model moves so that it matches the actual machine.
[0008]
Further, when the teaching data by simulation is reproduced by the actual machine, the teaching data is shifted by the function of the robot controller, or the troubled part is corrected while reproducing the actual robot.
[0009]
[Problems to be solved by the invention]
In such a conventional simulation, there is no particular problem in position correction because the position of the actual machine and the robot model in the simulation are matched.
[0010]
However, the method of correcting the error due to deflection or the like by combining the parameters of the robot model and the actual machine is very difficult to model with this parameter because the deflection is nonlinear. For this reason, in the past, an approximate expression is used for those that are difficult to completely parameterize, such as deflection, but because it is an approximate expression, it does not agree well with the well-matched part, and the error There was a problem that a part to be enlarged occurred.
[0011]
In addition, after the teaching data is created by simulation, the correction method using the controller of the actual robot only shifts the entire teaching point in the teaching data (X, Y, Z translation and axial rotation position are shifted). There is a problem that partial correction cannot be performed, such as deviation of the encoder zero point and deflection generated in the robot arm.
[0012]
Furthermore, although it is possible to correct the teaching points one by one with an actual robot, there is a problem that it takes a very long time if the number of teaching points is large.
[0013]
An object of the present invention is to provide a robot teaching data correction method capable of efficiently reducing errors at all teaching points when correcting teaching data created by robot simulation.
[0014]
[Means for Solving the Problems]
The object of the present invention is achieved by the following configurations.
[0015]
(1) A step of operating an actual robot with teaching data comprising a plurality of teaching points for determining an operation trajectory of the robot, a step of measuring the operation trajectory of the actual robot in the operation, and a specified axis among the axes of the robot Obtaining torque for each teaching point, grouping the teaching points where the obtained torque for each axis is in the same torque range, and corresponding the teaching point in the group on the measured motion trajectory A method for correcting robot teaching data, comprising: calculating a moving amount to be moved to a point to be corrected, and correcting the teaching data using the moving amount as a correction amount for all teaching points in the group.
[0016]
(2) The torque range is obtained by dividing the torque into a plurality of ranges based on a torque characteristic with respect to a deflection amount of the actual robot.
[0017]
(3) pre-Symbol specified were axes, when the robot is a six-axis robot, and wherein the second axis, the third axis, and a fifth axis.
[0018]
【The invention's effect】
The present invention has the following effects for each claim.
[0019]
According to the first aspect of the present invention, teaching points in which the torque of the specified axis among the robot axes is within the same torque range are grouped, and the robot teaching data is corrected in units of this group. Therefore, it is possible to reliably correct the partial deviation caused by the deflection of the actual robot, and by grouping multiple teaching points at once compared to correcting each teaching point. Since correction can be performed, the time required for correction can be reduced.
[0020]
According to the second aspect of the present invention, the torque range is obtained by dividing the torque into a plurality of ranges from the characteristic of the torque with respect to the deflection amount of the actual robot. The teaching points can be grouped depending on whether or not the torque characteristics are exhibited, and a plurality of teaching points can be grouped according to the degree of deflection regardless of the posture of the robot and the teaching point position.
[0021]
According to the third aspect of the present invention, when the robot is a 6-axis robot, the second axis, the third axis, and the fifth axis are the axes that are designated in the grouping, so that the greatest influence is exerted on the deflection. Since the grouping can be performed only by the axis that gives the grouping, the grouping process can be performed efficiently.
[0022]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings.
[0023]
FIG. 1 is a diagram for explaining a system configuration for carrying out the present invention, and FIG. 2 is a May routine flowchart showing a processing procedure of a correction method according to the present invention.
[0024]
As shown in FIG. 1, the system configuration includes a simulation apparatus 1 for creating teaching data, an actual robot 2 operated by the teaching data, and a three-dimensional position measuring device 3 for measuring the movement of the hand of the actual robot.
[0025]
The simulation apparatus 1 is a general computer as hardware, and a program for simulating a robot is executed by this computer. Then, the simulation apparatus 1 creates and corrects teaching data for operating the robot along a predetermined route.
[0026]
The actual robot 2 is a robot that is operated by teaching data created by simulation. The illustrated real robot 2 includes a first axis 21 for turning the entire robot, a second axis 22 for tilting the entire robot, a third axis 23 for tilting the arm 27, a fourth axis 24 for rotating the arm 20, and an end effector 28. Is a six-axis robot comprising a fifth axis 25 for inclining and a sixth axis 26 for rotating the end effector 28.
[0027]
The actual robot 2 is not necessarily the same as the robot that is actually operated by the teaching data created by the simulation apparatus 1. For example, a robot of the same type and shape as that of an actual robot actually used at a production site may be used.
[0028]
The three-dimensional measuring instrument 3 measures three-dimensionally the movement of the hand of the actual robot 2 (tip of the tool attached to the end effector). As such a three-dimensional measuring instrument, for example, an infrared light emitting diode (infrared marker) is attached to the hand of a robot, and the infrared marker is measured using at least three CCD line sensors, and the obtained three distances are measured. A device that measures the three-dimensional position of the infrared marker from the sensor and outputs it as a three-dimensional coordinate value. Also, it measures the position of the robot's hand by emitting sound waves, infrared rays, lasers, etc. and receiving the reflected sound and reflected light. Such devices are generally commercially available.
[0029]
The simulation apparatus 1, the actual robot 2, and the three-dimensional measuring instrument 3 are connected to each other via a LAN line or the like so that data can be exchanged. Such data exchange is a well-known computer technology and will not be described.
[0030]
Next, the procedure for correcting teaching data will be described with reference to FIG.
[0031]
First, teaching data prepared in advance is read by the simulation apparatus 1 (S1). The teaching data is data for operating the robot along a predetermined movement locus, and includes a plurality of teaching points. The teaching data itself is created by operating the robot model drawn by the design data by the simulation apparatus 1.
[0032]
The teaching points P1D to P8D of the teaching data read at this stage are shown in FIG. The teaching data that is read first is data indicating the design motion trajectory (this is referred to as design motion trajectory data). The teaching data that is first created so that the actual robot moves along the design motion trajectory data. The data is corrected.
[0033]
Next, the current teaching data (in this step, corrected teaching data may come as will be described later) is transferred to the controller (not shown) of the actual robot 2 and played back, and the actual robot 2 is actually (S2).
[0034]
Then, as shown in FIG. 4, the position of the hand of the operating robot is measured by a three-dimensional measuring instrument (S3).
[0035]
Subsequently, the measured data is taken into the simulation apparatus 1 (S4). Points at positions corresponding to teaching points in the acquired measurement data are P1m to P8m (this is referred to as measurement points).
[0036]
Then, the position of each teaching point is compared with the position of the measurement point in the design motion trajectory data (initially acquired teaching data) to determine whether or not the deviation is within the allowable range (S5). If it is within (S5: Yes), the correction process is terminated. In this determination, the coordinate system and the measurement coordinate system in the simulation apparatus 1 are matched in advance. Here, the coordinate system of the measurement reference is used.
[0037]
In the first stage, as shown in FIG. 5, the measured measurement points P1m to P8m are deviated from the teaching points P1D to P8D of the design motion trajectory due to the deflection of the robot arm or the like.
[0038]
Subsequently, the torque of each axis of the robot at each teaching point is calculated by the simulation apparatus 1, and the teaching points in which the torque of the arbitrarily designated axis of the robot is in the same torque range are grouped (S6).
[0039]
Here, the grouping process will be described with reference to a subroutine flowchart of the grouping process shown in FIG.
[0040]
First, the teaching point number variable i is initialized to 0 (S61).
[0041]
Subsequently, the joint angle of each axis at the i-th teaching point is calculated (S62). The calculated shaft joint angles are stored as JV1 to JVn for each axis. Note that n is the maximum axis number. Individually, the joint angle to be calculated is a joint angle of each axis when moving from a certain teaching point (for example, teaching point P1D) to the next teaching point (P2D).
[0042]
Subsequently, each axis torque is calculated from the joint angle (S63). The torque of each axis calculated here is the torque of each axis when moving from one teaching point to the next teaching point based on the relationship between the joints calculated above. For example, the torque of the teaching point P2D is the teaching point P1D. Is the torque of each axis output when moving from the teaching point to the next teaching point P2D. The calculated torque of each axis is stored as TR1 to TRn.
[0043]
Subsequently, the axis number variable j is set to 1 (S64), and the torque characteristic division number of the j-th axis is set (S65).
[0044]
The number of torque characteristic divisions depends on the length and rotation direction of the robot arm, the weight of the arm and end effector, etc., and in order to divide the magnitude of the torque output with respect to the amount of deflection, The number of divisions is set.
[0045]
FIG. 7 is a diagram illustrating an example in which the torque characteristics with respect to the deflection amount are graphed and the torque is divided into ranges.
[0046]
For example, as shown in FIG. 7A, in the case of an axis driving an arm with a relatively small amount of deflection, the torque amount may be divided into two, that is, the number of torque characteristic divisions may be two. Further, as shown in FIG. 7B, in the case of an axis driving an arm with a large amount of deflection, the torque amount may be divided into four, that is, the torque characteristic division number may be four.
[0047]
The number of divisions of this torque characteristic depends on the shape of the robot and the arm weight, and when using a hand that grabs or lifts the workpiece on the end effector, the amount of deflection of the arm varies depending on the weight of the workpiece. It is set appropriately according to them.
[0048]
After setting the number of torque characteristic divisions for the j-th axis, it is subsequently determined which torque range the j-th axis torque is in (S66). Here, the torque range refers to each of the divided areas shown in FIG. 7, and “a” or “b” in FIG. 7A, or “a”, “b”, “c”, or “ d ".
[0049]
Subsequently, the axis number variable j is incremented (S67), and it is determined whether or not the variable j exceeds the maximum axis number n (S68). If it does not exceed (S68: No), the process returns to step S66 and the incremented j Determine the torque range of the torque of the second axis.
[0050]
On the other hand, if the variable j exceeds the maximum axis number n (S68: Yes), then the teaching point number variable i is incremented (S69), and it is determined whether or not the variable i exceeds the maximum teaching point number m. (S70) If not exceeded (S70: No), the process returns to step S62 and the subsequent processing is repeated.
[0051]
When grouping by torque is completed for all axes at all teaching points (S70: Yes), teaching points in which all the axes set in advance for each teaching point belong to the same torque range are grouped together. (S71).
[0052]
For example, as shown in FIG. 8, it is easy to understand if the torque ranges to which the torques of the respective axes belong for each teaching point are tabulated. This grouping is performed in the simulation apparatus 1, that is, in the computer. It is not necessary to create such a chart in the actual processing in the computer, and FIG. 8 explains the concept of grouping. It is a chart for doing.
[0053]
In the example shown in FIG. 8, the teaching points 1, 3, and 4 have the same torque range for all the axes, so that these are grouped together.
[0054]
At this time, in grouping, an axis to be focused on is set in advance. This is because, for example, the movement of the first axis for rotating the entire robot, the fourth axis for rotating the arm, the sixth axis for rotating the end effector, and the like hardly affect the deflection of the arm. Therefore, for the axes that do not affect the deflection of the arm, these are excluded as axes that do not affect the correction of the teaching point position, and conversely, the second axis that greatly affects the deflection of the arm, The third axis and the fifth axis are the axes that need to be corrected, and are used when grouping these.
[0055]
If there are teaching points that do not belong to any group due to this grouping, the number of axes to be focused on when grouping is reduced and another attempt is made to group teaching points that do not belong to any of those groups. . Eventually, if there are teaching points that do not belong to any group even if grouping is performed with the number of axes focused on when grouping, one teaching point is one group. To do.
[0056]
When grouping of teaching points to be corrected is completed as described above, the process returns to the main routine.
[0057]
In the main routine, the correction amount is calculated for each group (S7). The calculation of the correction amount will be described.
[0058]
Here, for example, as shown in FIG. 9B, as a result of grouping for the first teaching points P1D to P3D as shown in FIG. 9A, the teaching points P1D, P3D, P5D, and P7D are the first group, P2D, A case where P4D and P6D are in the second group will be described.
[0059]
First, for the first group, as shown in FIG. 9C, three points P1D, P3D, and P7D are selected from the teaching points in the first group, and these are connected by lines to form a triangle. The three points selected at this time are selected so that the area of the triangle is the largest.
[0060]
Then, a triangle is created by the measurement points P1m, P3m, and P7m corresponding to the teaching point that created the triangle, and the triangle by the teaching point is moved so as to overlap the triangle by the measurement point. At this time, if the triangles do not overlap only by parallel movement, the entire triangles may be rotated. In addition, when the triangles do not completely overlap, the shift is minimized.
[0061]
Thus, the amount of movement when the triangle is moved by the teaching point becomes the correction amount of each teaching point in the first group.
[0062]
Similarly, the correction amount is obtained for the second group. That is, as shown in FIG. 9E, a triangle is formed using the teaching points P2D, P4D, and P6D of the second group, and as shown in FIG. 9F, the triangle is formed into a triangle by the corresponding measurement points P2m, P4m, and P6m. It moves so that it may overlap, and let that amount of movement be a correction amount.
[0063]
In the above calculation of the correction amount, when there are no more than three teaching points in one group, that is, when a triangle cannot be formed, the line segments connecting the teaching points are used. The movement amount (parallel movement amount and rotation amount) is obtained so as to overlap the line segment connecting the corresponding measurement points, and this movement amount is set as the correction amount of each teaching point in this group.
[0064]
As described above, the correction amount of each teaching point in each group is obtained. Therefore, the teaching data is corrected by shifting the position of each original teaching point by this correction amount (S8). FIG. 10 shows teaching points P1c to P8c in the corrected teaching data.
[0065]
Then, returning to step S2, the actual robot is replayed with the corrected teaching data (S2), the movement of the hand is again measured by the three-dimensional measuring instrument (S3), and the measurement data is taken into the simulation apparatus 1. (S4) As shown in FIG. 11, it is confirmed whether or not each measurement point and the teaching point of the design motion trajectory are within an allowable range (S5). The processes in and after step 6 are repeated until the deviation of the design movement locus from the teaching point is within the allowable range. If it is within the allowable range, the correction process is terminated.
[0066]
After the correction process is completed, an interference check with peripheral equipment and workpieces is performed, and the result is input to the actual robot as final teaching data.
[0067]
As described above, according to the present embodiment, when the actual robot is operated based on the design motion trajectory data and the teaching data, the deviation of the motion trajectory is grouped for each teaching point using the torque as an index. This makes it possible to correct a plurality of teaching points that are misaligned at the same time, so that it is possible to reliably correct misalignment caused by deflection of the robot arm, which has been difficult to parameterize. It becomes possible. In addition, since the plurality of teaching points can be corrected at one time by grouping, it is possible to correct the teaching points more quickly and easily than by manually correcting each teaching point one by one. This shows the case where the number of teaching points is 8 in the present embodiment, so that it does not take much time to correct each one, but the actual number of teaching points is from several tens to one hundred. As the number of teaching points increases, the effect of applying the present invention is greater.
[0068]
Although the embodiment to which the present invention is applied has been described above, the present invention is not limited to such an embodiment. For example, in the above-described embodiment, the example of the six-axis robot has been described, but it goes without saying that the present invention can also be applied to a three-axis robot and other robots.
[Brief description of the drawings]
FIG. 1 is a diagram for explaining a system configuration for carrying out the present invention.
FIG. 2 is a May routine flowchart showing a processing procedure of a correction method according to the present invention.
FIG. 3 is a diagram showing teaching points.
FIG. 4 is a diagram showing a state in which the position of the hand of a robot in operation is measured by a three-dimensional measuring instrument.
FIG. 5 is a drawing showing design operation locus data teaching points and actually measured measurement points;
FIG. 6 is a subroutine flowchart showing a grouping procedure;
FIG. 7 is a graph showing an example of torque characteristics with respect to a deflection amount.
FIG. 8 is a diagram showing an example in which torque ranges of respective axes are summarized in a tabular form for each teaching point.
FIG. 9 is a view for explaining grouping of teaching points;
FIG. 10 is a diagram showing teaching points in corrected teaching data.
FIG. 11 is a diagram showing a deviation between a measurement point and a teaching point of design operation trajectory data.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Simulation apparatus 2 ... Real robot 3 ... Three-dimensional position measuring device

Claims (3)

A stage of operating the actual robot with teaching data consisting of a plurality of teaching points that determine the robot's movement trajectory;
Measuring the movement trajectory of the actual robot in the movement;
Obtaining the torque of the designated axis among the axes of the robot for each teaching point, and grouping the teaching points where the torque for each obtained axis is in the same torque range;
A movement amount for moving the teaching point in the group to a corresponding point on the measured motion locus is calculated, and the teaching data is corrected using the movement amount as a correction amount for all teaching points in the group. Stages,
A method for correcting robot teaching data, comprising: The torque range is
2. The robot teaching data correction method according to claim 1, wherein the torque is divided into a plurality of ranges based on a characteristic of torque with respect to a deflection amount of the actual robot. Before SL specified were axes, when the robot is a six-axis robot, the second axis, the third axis, and a method of correcting the robot teaching data according to claim 1 or 2, wherein the a fifth axis .

Images