JPH11123678A - Position detecting method for work - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a high speed and a high accuracy of position detecting method for a work. SOLUTION: The coordinates of a robot when the position of a tool 2 is fitted to a standard point 6; the coordinates of the robot 1 when a distance sensor 3 is at the position to detect the standard point 6; and the distance between the distance sensor 3 and the standard point 6, which is detected by the distance sensor 3; are found beforehand. And from the coordinates of the robot when the distance sensor 3 detects a random point, and the distance detected by the distance sensor 3, a calibration data to find the position of the above random point is produced. And the position of a feature point is found by an operation, depending on the distance between the distance sensor 3 and a work 4 when the distance sensor 3 is scanned to cross the feature point; the position and the posture of the distance sensor 3; and the above calibration data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a position of a work, and more particularly to a method for detecting a position of a work point to be processed prior to a processing operation performed by a robot.

[0002]

2. Description of the Related Art In a machining operation by a robot, for example, in fillet welding of a lap joint, the position of a joint is sensed in order to correct a deviation between a joint position previously taught to the robot and a joint position of an actual work. is there. In such a case, conventionally, a sensor is scanned in a direction intersecting the joint to detect an intersection point between the joint and the scanning locus. This intersection is called a feature point. In the lap joint, the feature point is detected as a step. The following method has been proposed as a method for detecting this feature point. A distance sensor for detecting a distance to the work surface is positioned on a feature point of the work, and the distance sensor detects a distance between the feature point and the distance sensor by the distance sensor.
When detecting the position of the feature point based on the detected distance and the position of the preceding distance sensor, the distance sensor teaches the distance sensor to scan across the feature point, and the distance sensor according to the teaching data. The position of the feature point is calculated and obtained based on the distance data of the distance sensor acquired when the sensor is scanned and the teaching data (for example, Japanese Patent Application Laid-Open No. 7-332927).

[0003]

However, in the prior art, the teaching data used in the calculation is only position coordinates that do not include the attitude data of the distance sensor. It had to be kept constant. Therefore, when the attitude of the distance sensor cannot be made constant due to interference with a work or a peripheral device, it cannot be performed. Further, even when the distance sensor cannot scan so as to cross over the feature point due to interference with a work or a peripheral device, it cannot be performed. Also, in order to make the robot scan at high speed and ensure the accuracy of the detection position, it is necessary to increase the sampling frequency and increase the number of distance data, but this depends on the performance of the CPU. There was a limit. For this reason, when the distance sensor scans at a high speed, there is a problem that the accuracy of position detection is deteriorated compared to the case where the distance sensor scans at a low speed. The present invention has been made in view of such circumstances, and has as its object to enable high-speed and high-accuracy work position detection.

[0004]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the present invention relates to a method for adjusting the coordinates of the robot when the position of the tool is adjusted to a reference point and the position at which the distance sensor detects the reference point. The coordinates of the robot at a certain time, the distance between the distance sensor and the reference point detected by the distance sensor are obtained, and the coordinates of the robot and the distance sensor when the distance sensor detects an arbitrary point From the distance detected, create calibration data to determine the position of the arbitrary point, the distance between the distance sensor and the workpiece obtained when scanning the distance sensor across the feature point And calculating the position of the feature point based on the position and orientation of the distance sensor and the calibration data.
Further, based on the distance between the distance sensor and the workpiece acquired when the distance sensor is scanned across the feature point, the position and orientation of the distance sensor, and the calibration data. It calculates surface position data, performs image processing on the position data, detects temporary feature points, and calculates true feature points by performing interpolation processing on the position data near the temporary feature points. is there.

[0005]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a system configuration diagram showing an embodiment of the present invention, FIG. 2 is a flowchart and an explanatory diagram showing a procedure of a calibration process in the first embodiment of the present invention, and FIG. 3 is a flowchart of a detection process in the first embodiment. FIG. 4 is a flowchart showing the procedure, FIG. 4 is an explanatory diagram of the detection process in the first embodiment, and FIG. 5 is a flowchart and an explanatory diagram showing the procedure of the interpolation process in the second embodiment.

In FIG. 1, reference numeral 1 denotes a robot, and a welding torch 2 and a one-dimensional laser sensor 3 are provided at the tip of the robot 1.
Is installed. The welding torch 2 and the one-dimensional laser sensor 3 can take any positions and postures within a predetermined operation range by the industrial robot 1 while maintaining the relative positions and postures of the two. A control point 2 a is set at the tip of the welding torch 2. The control point 2a is a point to which a control device (not shown) of the robot 1 refers, and the control device has the control point 2a at a predetermined position and the welding torch 2 takes a predetermined posture with respect to the control point 2a. Command to the robot 1 as described above. Position of control point 2a and control point 2
Here, the coordinates indicating the posture of the welding torch 2 with respect to a will be simply referred to as robot coordinates here. The one-dimensional laser sensor 3 is a distance sensor that measures a distance between the one-dimensional laser sensor 3 and a detection target. Reference numeral 4 denotes a workpiece to be processed, which is fixed to a jig stand 5.

A first embodiment of the present invention will be described. First,
Create calibration data. The calibration data is generated based on the coordinates indicating the position and orientation of the control point 2a of the industrial robot 1, that is, the robot coordinates and the distance to the point detected by the one-dimensional laser sensor 3 on the work 4, and the above-described detection is performed. To calculate the coordinates of the point
This is the task of finding the parameters. Here, the calibration data includes a unit vector e in the laser optical axis direction of the one-dimensional laser sensor 3 and an offset vector D from the robot control point 2a to the laser emission point. The procedure for creating the calibration data will be described with reference to FIG. First, in step ST11, an arbitrary point fixed on the work 4 is selected, and the reference point 6 is determined. Next, in step ST12, the robot 1
And the robot 1 is moved to the position A so that the tip of the welding torch 2 coincides with the reference point 6. Next, in step ST13, the robot 1 is taught and the position B
Then, the robot 1 is moved so that the laser light of the one-dimensional laser sensor 3 hits the reference point 6. Next, in step ST14, the robot 1 is taught and the position C
Then, the robot 1 is moved so that the laser light of the one-dimensional laser sensor 3 hits the reference point 6. However, this position C is different from the position B taught in step ST13. Further, in step ST15, the robot coordinate value P0 of the point taught in step ST2 is obtained. Next, in step ST16, the previous step ST
13. The robot coordinate values P1 and P2 of the point taught in step ST14 and the one-dimensional laser sensor 3 at that point.
And the distance d1, d2 between the reference point 6 and. Finally, in step ST17, a vector e and a vector D are calculated by the following equations based on the data obtained in steps ST15 and ST16 before. e = (P2 P0 -P1 P0) / (d2 -d1) D = P1 P0 -d1 × e As described above, steps ST11 to ST17 are performed.
By performing the above operations in order, calibration data is created.

Next, the procedure for finding the characteristic points of the work 4 will be described.
This will be described with reference to FIGS. First, step ST18
In FIG. 4, the robot 1 is taught so that the laser beam from the one-dimensional laser sensor 3 hits a scanning start point 7 for detecting a feature point, as shown in FIG. The coordinates of the robot 1 at this time are P3. Next, step ST
At 19, the robot 1 is taught so that the laser beam from the one-dimensional laser sensor 3 hits the scanning end point 7, as shown in FIG. The coordinates of the robot 1 at this time are P4. Next, in step ST20, the robot 1 is moved to the position P3 taught in step ST18. Next, in step ST21, the scanning of the one-dimensional laser sensor 3 is started. That is, the robot 1
Is moved from the position P3 to the position P4 taught in step ST19, and at the same time, the robot coordinate values (position and posture)
Then, sampling of distance data between the one-dimensional laser sensor 3 and the work 4 is started. Next, step ST22
In, sampling is continued until the movement of the robot 1 is completed. Next, in step ST23, the distance data of the point sequence on the workpiece 4 obtained by the sampling in the previous step is analyzed, and a point at which the distance data changes discontinuously is determined. . FIG. 4C is a diagram showing a state where the one-dimensional laser sensor 3 has detected the characteristic point 9. Next, in step ST24, distance data du when the feature point 9 is detected and robot coordinate values Pu (position and posture) are acquired. Finally, in step ST25, the position P of the feature point 9 is obtained from the distance data du when the feature point 9 is detected, the robot coordinate value Pu (position and orientation), and the previously acquired calibration data according to the following equation. Calculate based on P = Pu + D + du × e

In the first embodiment, feature points are selected from the sampled points. Therefore, even if there is a true feature point between the sampled points, it cannot be detected. Therefore, in order to improve the detection accuracy of the feature points, it is necessary to obtain a large amount of data by shortening the sampling period, and there is a problem that the detection and calculation time is long. The solution to this point is the second aspect of the present invention described below.
This is an embodiment of the invention. This second embodiment is different from step ST1 in step ST1.
The procedure from step 1 to step ST17 is performed to create calibration data, and the procedure from step ST18 to step ST22 is performed to sample robot coordinate values and distance data. The description is omitted because it is the same as the example. In the second embodiment of the present invention, a procedure for analyzing a robot coordinate value and distance data to obtain a true feature point will be described with reference to FIG. First, in step ST26, coordinate data of a detected point sequence on the work surface is calculated from the robot coordinate values (position and posture), sensor distance data, and calibration data. Next, in step ST27, the coordinate data obtained in step ST26 is analyzed, and as shown in FIG. 5B, a point Pu at which the slope of the curve obtained by connecting the point sequence changes abruptly is determined. It is a feature point. Finally, in step ST28, the curve 31 connecting the point sequences Pu-n, .., Pu-3, Pu-2, Pu-1 arranged near the left side of the temporary feature point Pu is approximated by the least squares method. Ask. Similarly, a point sequence Pu + arranged near the right side of the provisional feature point Pu
A curve 32 connecting n, .., Pu + 3, Pu + 2, and Pu + 1 is obtained. The intersection Q of the curve 31 and the curve 32 is calculated, and this is set as a true feature point. The scanning by the one-dimensional laser sensor 3 is performed by 1
Although the example in which the position and the posture of the three-dimensional laser sensor 3 are changed at the same time has been described, it is also possible to move only the position while keeping the posture with respect to the work 4 constant. Alternatively, scanning may be performed while the position of the one-dimensional laser sensor 3 is kept constant and only the posture is changed. The object to be detected in the present invention is not limited to a step, and it goes without saying that any point such as a gap, a ridgeline, or a groove may be used as long as the shape of the detection object changes discontinuously.

[0010]

As described above, the present invention has the following effects. (1) Since the calibration is performed in advance so that the coordinates of the point detected on the work can be obtained from the robot coordinates and the distance data obtained by the one-dimensional laser sensor, the posture of the one-dimensional laser sensor with respect to the work Can be rubbed while changing the distance, so that it is easy to avoid interference with the work and peripheral devices. (2) Since the feature points are obtained by interpolating the coordinates of a small number of point sequences, the measurement time and the calculation time are short, but the detection accuracy is high nonetheless.

[Brief description of the drawings]

FIG. 1 is a system configuration diagram of an embodiment of the present invention.

FIGS. 2A and 2B are diagrams illustrating a procedure of a calibration process according to the first embodiment of the present invention, wherein FIG. 2A is a flowchart and FIG.

FIG. 3 is a flowchart illustrating a procedure of a detection process according to the first embodiment of the present invention.

FIG. 4 is an explanatory diagram of a detection process according to the first embodiment of the present invention, wherein (a) is a diagram showing a scanning start point,
(B) is a diagram showing a scanning end point, and (c) is a diagram showing a state in which a feature point is detected.

5A and 5B are diagrams showing a procedure of an interpolation process in a second embodiment of the present invention, wherein FIG. 5A is a flowchart and FIG. 5B is an explanatory diagram.

[Explanation of symbols]

 1: robot 2: welding torch 2a: control point 3: one-dimensional laser displacement sensor 4: workpiece (work) 5: jig stand 6: reference point 7: scanning start point 8: scanning end point 9: feature point Pn: Point sequence detected by scanning Pu: Provisional feature point Q: True feature point 31: Curve connecting point sequence on the left of provisional feature point 32: Curve connecting point sequence on the right of provisional feature point

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Inventor Haruhiko Sato 2-1 Kurosaki Castle Stone, Yawatanishi-ku, Kitakyushu-shi, Fukuoka Prefecture Yasukawa Electric Co., Ltd.

Claims (3)

[Claims] 1. A distance sensor for detecting a distance to a surface of a work, and a tool mounted on a robot so as to maintain a relative position and posture of the distance sensor and the tool. A work position detection method for detecting a distance between the distance sensor and the feature point by the distance sensor and detecting a position of the feature point based on the distance and the position of the distance sensor. In advance, the coordinates of the robot when the position of the tool is adjusted to a reference point, the coordinates of the robot when the distance sensor is at a position where the reference point is detected, and the coordinates detected by the distance sensor. Obtain the distance between the distance sensor and the reference point, from the coordinates of the robot when the distance sensor detects an arbitrary point and the distance detected by the distance sensor Calibration data for obtaining the position of the arbitrary point is created, and the distance between the distance sensor and the workpiece obtained when the distance sensor is scanned across the feature point, and the distance sensor Based on the position and orientation, and the calibration data,
A method for detecting a position of a workpiece, wherein the position of the feature point is calculated and obtained. 2. The position detection of a work according to claim 1, wherein the position of the distance sensor is fixed, and only the posture of the distance sensor is changed so as to scan across the feature point. Method. 3. A distance between the distance sensor and the workpiece, which is acquired when the distance sensor is scanned across the feature point, a position and a posture of the distance sensor, and the calibration data. Calculating a true feature point by calculating position data of the work surface based on the calculated position data, obtaining a temporary feature point, and performing interpolation processing on the position data in the vicinity of the temporary feature point. 3. The method for detecting a position of a work according to claim 1 or claim 2.