JPS63232986A - Control method of robot - 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 (Industrial Application Field) The present invention provides a robot control method that uses a sensor to automatically correct tool movement trajectory errors.
The time required for trajectory teaching has been significantly reduced.
This relates to a method of controlling a robot.

(Prior art) When a tool such as a welding torch is attached to a robot arm and the tool is moved by the robot along a pre-taught trajectory, a so-called teaching process is performed in which the robot is taught the trajectory in advance. . When the robot starts operating, the sensor detects whether or not the tool is moving along the taught trajectory, and if a deviation occurs between the actual trajectory and the previously taught trajectory, the deviation is detected. Depending on the amount, the tool will be stopped or the trajectory will be corrected while it is still moving.

(Problem to be solved by the invention) In this way, in a device that uses a sensor to correct the movement trajectory, if the error detection range of the sensor is within the error detection range, even if the tool posture is out of alignment, Even if there is (for example, when it hits a tool or something and bends it), the work that occurs because of this! Since the correction function on the robot side based on sensor detection compensates for the deviation, there was no problem with welding or other operations unless a sensor malfunction occurred. However, if the sensor itself or the sensor's arithmetic circuit breaks down, it will disrupt the work that had previously been done without any problems due to detection by the sensor, resulting in water leakage, such as welding a part that is significantly misaligned. This may cause problems such as production being completely stopped. This problem is solved by Utility Model Application No. 59-16, which is the prior art of the present invention.
Even the one disclosed in Japanese Patent No. 5005 cannot solve the problem.

The present invention has been made in view of this point, and in the above-mentioned cases, it automatically calculates data based on the data detected by the sensor, whereas conventionally the operator had to create correction data. The present invention aims to provide a control method for a robot in which the robot itself automatically corrects itself based on the maximum distribution value determined.

(Means for Solving the Problems) The present invention, as a means for solving the above problems, attaches a tool to the arm of the robot and also attaches a sensor for detecting the movement trajectory of one nuclear tool, and uses the sensor to detect the In a control method for a robot that follows a tool movement trajectory, the difference between the initial teaching data and the actual tool movement trajectory is found from the detection data of the sensor, and this is stored for a set number of sampling times, and the distribution is Calculate the maximum distribution value to find the maximum distribution value, calculate the error value of the maximum distribution value with respect to the initially taught trajectory based on the maximum distribution value, and calculate the error value between the calculation result and the initially taught trajectory data. A robot control method has been established that is characterized by modifying and rewriting the robot to correct it.

(Function) With this configuration, when the work is proceeding with the initially taught trajectory data, the teaching data is automatically rewritten to the one with the error corrected, so the teaching data is always new. Therefore, even if the sensor malfunctions and the automatic correction at@ does not work, the robot's control trajectory is corrected to the maximum error distribution value, so the same product can be continuously manufactured using this new data. It will be possible to manufacture it by doing so.

(Example) Next, an application example of the method of the present invention will be explained with reference to the drawings. In Fig. 1, l is a steel plate as a workpiece, and the edge portion is overlapped on another steel plate 2 as a workpiece, and a torch is applied. It is to be welded. 3 is a welding torch as a type of tool, and 1 is attached to an arm of a robot (not shown). A sensor 4 is attached near the welding torch 3 by a bracket 5. In this case, the robot approaches by moving in the direction of the arrow 6 from the upper steel plate 1 to the lower steel plate 2 and toward the arc start point 7. 8 is the arc end point, and 9 is the initial teaching locus. However, due to some reason such as a change in the material, the mating portion of the two steel plates 1.2 has shifted as shown in the figure. Conventionally, in such cases, if the trajectory is corrected based on the data detected by the sensor, the corrected data is erased, and the process of detection → correction → deletion is repeated until the work is finished. In the method of the present invention, the data is corrected as described below, and in the process of moving the welding torch 3 as shown by the arrow 6,
The step 11 between the two steel plates 1.2 is detected, and from this detection, the teaching trajectory 9 between the arc start point 7 and the arc end point 8 taught in advance and the teaching from the actual arc start point 11 to the arc end point 12 are determined. The deviation from the trajectory 13 is corrected, and welding is performed using the corrected teaching trajectory.

This will be explained using the flowchart shown in FIG. First, the sensor 3 detects data (step 14), then calculates a corrected trajectory value based on the detected data (step 1).
5) The trajectory is corrected (step 16), this correction data is stored in the memory circuit (step 18), and this correction and storage continues until a preset sampling number is reached or until the similar operation is completed. is performed (step 18). After the storage has been performed the set number of times of sampling, the distribution of the corrected data is calculated (step 19), and the maximum distribution value is determined by this calculation (step 2G).

The method for determining the maximum distribution value is as shown in Figure 3.
The distribution of the corrected direction and amount is taken with reference to ', and the trajectory is calculated so as to correct the error between the maximum value 24 of the distribution and the taught trajectory 9'.

The obtained locus data is re-stored as new teaching data (steps 21 to 23). By controlling in this way, the teaching tjk trace will be the same as that of the tool within the number of samples even if the sensor 4 fails. Since maintenance is always performed to match the maximum distribution value of the movement locus, work such as welding can be continued.Also, even when switching to work using only the teaching locus, it is less susceptible to changes in the funnel of the workpiece, etc. It comes.

(Effects of the Invention) Since the present invention is a robot control method as described above, even if the sensor fails during operation, it will not be recorded as long as it is within the initial teaching range for the workpiece or the detection force of the sensor. Subsequent control will be performed based on the stored maximum distribution value. Therefore, re-teaching is no longer necessary, and the number of man-hours required for teaching can be significantly reduced. In addition, when teaching data is created offline without using the actual workpiece, errors between the input data and the actual workpiece can be automatically corrected, so there is almost no need to check the data after it is transferred to the production line. It turns out.

With these effects, it is possible to reduce the time and man-hours required for switching over to a new line.

[Brief explanation of the drawing]

Figure 2 is a perspective view illustrating the robot DJ control method according to the present invention, Figure 2 is a flowchart explaining the flow of the method of the present invention, and Figure 3 is a graph showing the distribution of the direction and amount of the corrected trajectory. . 1.2... Steel plate 3... Welding torch 4... Sensor 7.11... Arc start point 8.12... Arc end point 9.13... Teaching locus lO... Level difference characteristic Applicant: Toyota Motor Corporation Figure 1
Figure 2 1.2 Goban 3 5 Tsuke Torch 4... Gombu 7.11 Arc start, east. 8.12 Ark End, 9.13-, Gajiki Mushroom- 10 p2 difference

Claims (1)

[Claims] (1) In a robot control method in which a tool is attached to a robot arm and a sensor is attached to detect the movement trajectory of the tool, and the movement trajectory of the tool is followed by the sensor, from the detection data of the sensor, Find the difference between the initial teaching data and the actual tool movement trajectory, store this for the set number of samplings, calculate the distribution to find the maximum distribution value, and calculate the maximum distribution value based on the maximum distribution value. 1. A method for controlling a robot, comprising: calculating an error value with respect to an initially taught trajectory; and correcting and rewriting the calculated result to correct the error between the initially taught trajectory data.