JP2013035054A - Welding robot control device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve such a problem that, when performing the position correction of a teaching point by a relative position method by using a laser sensor, it is necessary to take a long time for performing teaching correction if an attachment position relationship with a welding torch becomes abnormal due to the attachment/detachment of the laser sensor.SOLUTION: Calibration is performed between the laser sensor and the welding torch in an S10, and a new sensor-torch transition matrixTis obtained and stored in a memory in an S20. The operation of a differential transition matrixTis performed in an S30. A detection reference point stored in a sensing command is searched for with respect to a working program which is selected in advance in an S40. The differential transition matrixTis multiplied with each searched detection reference point in an S50, and the detection reference point is updated.

Description

  The present invention relates to a welding robot control device.

  Conventionally, in arc welding robots, the position of a workpiece is detected and calculated by a sensor attached to the welding robot during teaching before welding, and the welding teaching point at teaching is corrected based on the calculation result. ing. As the sensor, for example, a wire touch sensor is used in welding applications.

  The welding teaching point correction method includes an absolute position method and a relative position method. In the absolute position method, for example, a wire touch sensor detects and calculates a welding start point and a welding end point on the weld line as target positions, and the robot control device stores the target positions and performs the target when welding is performed. Welding is performed by directly using the position information of the position.

  On the other hand, in the relative position method, the robot controller stores the difference between the welding target position detected / calculated by the sensor and the previously taught target position as the “deviation amount”, and corrects the teaching position based on that information (parallel movement). ) And perform welding.

Specifically, the relative position method will be described with reference to FIGS.
As shown in FIG. 9, at the time of teaching, the positions P101 and P102 detected by the wire touch sensor 200 and the welding start point P103 and the end point P104 on the welding line Y of the master work W are taught. The taught point is called a taught point.

  Next, the welding operation is automatically performed on the workpiece W1 to be welded, and the difference between the position P101 ′ detected and calculated by the wire touch sensor 200 and the previous P101 as shown in FIG. And a displacement amount L2 which is a difference between the position P102 ′ detected and calculated by the wire touch sensor 200 and the position P102, and the robot control device stores the displacement amounts L1 and L2.

  Next, as shown in FIG. 11, when the welding operation is performed automatically on the workpiece W1 to be welded by automatic operation, the taught welding start point P103 is corrected by the deviation L1 and the welding start point is corrected. While correcting to P103 ′, the welding end point P104 is corrected to the welding end point P104 ′ by the shift amount L2.

  Since the relative position method can process the amount of deviation, it has a high ability to handle various actual workpieces, and fine adjustment of the target position is possible. Further, since the relative position method can be considered with the detection position and the welding position separately, it is possible to correct a point where direct sensing cannot be performed. For example, in FIG. 11, welding from the end of the workpiece W1 that cannot be sensed is possible. Furthermore, since the stored data can be reused, there is an advantage that the tact time is not wasted.

  For example, as shown in FIG. 11, when the workpiece W1 is displaced only in parallel to the master workpiece W at the time of teaching, the teaching of the position P102 and the detection of the position P102 ′ are unnecessary, and the welding end point is detected by the displacement amount L1. What is necessary is just to correct | amend P104 to welding end point P104 '.

  Further, an advantage when performing double-sided welding of a T-shaped workpiece by the relative position method will be described. FIG. 12 is a plan view of a T-shaped work, the dotted line indicates the master work W, and the solid line indicates the work W1 to be welded. In FIG. 12, Y10 and Y11 indicate the welding lines on the front side and the back side of the master work W, and Y12 and Y13 indicate the welding lines on the front side and the back side of the work W1, respectively. Further, the welding start point and the welding end point of the front welding line Y10 of the master work W are P110 and P111, and the welding start point and the welding end point of the back welding line Y11 are P112 and P113.

  In this case, with respect to the welding start point P112 and the welding end point P113 of the welding line Y11 on the back side of the master work W, the thickness t of the rising wall is known, the welding lines Y10 and Y11 are parallel, and the back side The positions of the welding start point P112 and the welding end point P113 are parallel to the welding start point P110 and the welding end point P111 by the thickness t.

  On the other hand, the welding start point and welding end point of the welding line Y12 on the front side of the workpiece W1 are P110 ′ and P111 ′, and the welding start point and welding end point of the backside welding line Y13 are P112 ′ and P113 ′. .

  In this case, as shown in FIG. 12, when the workpiece W1 is deviated from the master workpiece W, the deviation amount of the welding start point and the deviation amount of the welding end point at the welding line Y10 on the front side are shown in FIG. By performing detection and calculation in the same manner as described with reference to FIG. 11, the welding start point P110 ′ and the welding end point P111 ′ at the welding line on the front side of the workpiece W1 can be corrected. In addition, the welding start point P112 'and the welding end point P113' of the back side welding line Y13 can also be corrected based on the thickness t and the shift amount.

  In the workpiece W1, by sensing only the welding start point P110 ′ and the welding end point P111 ′ of the front side welding line Y12, the welding start point P112 ′ and the welding end point P113 ′ of the back side welding line Y13 are positioned without sensing operation. Correction can be made.

  By the way, although the above demonstrated the relative position system of the wire touch sensor, the method of using the laser sensor attached to the welding torch like patent document 1, patent document 2, and patent document 3, for example is also generally known. ing.

  When a laser sensor is used, teaching steps can be reduced as compared with a wire touch sensor. For example, when the position P101 in the workpiece W1 of the fillet joint shown in FIG. 10 is detected, the wire touch sensor requires 10 steps (4 sensing points, 6 approach points and 6 retract points). As shown, the laser sensor LS has the advantage of a total of two steps: one approach and one sensing command. In FIG. 13, FOV indicates the field of view range of the laser sensor LS.

Further, the laser sensor has high detection accuracy and can detect a workpiece to which the touch sensor cannot be applied.
As shown in FIG. 14, the position detection of the workpiece by the laser sensor LS is performed based on a sensor coordinate system reference, for example, a camera coordinate system reference, and the detection reference point is stored in the camera coordinate system. FIG. 14 illustrates how the left overlap joint is detected by the laser sensor LS in the visual field range FOV. In the wire touch sensor, the wire tip is a control point at the time of sensing and welding, whereas in the laser sensor, the control point at the time of sensing is one point in the field of view of the sensor and is different from the control point at the time of welding. Therefore, as a means for obtaining the attachment relationship between the welding wire tip (control point at the time of welding) and the sensing control point of the laser sensor, the attachment relationship is often obtained by conversion (matrix) using software. Here, the operation for obtaining the attachment positional relationship between the tip of the welding wire and the sensing control point of the laser sensor is called sensor calibration, and the transformation matrix obtained at this time is called the transformation matrix of the sensor coordinate system-torch coordinate system.

Japanese Patent No. 3317101 Japanese Patent Laid-Open No. 7-104831 Japanese Patent No. 3608673

  As described above, in the relative position method, the detection reference point is stored as a camera coordinate value (also referred to as a sensor coordinate value), and at the time of sensing, the detected coordinate value obtained by sensing and the coordinate value of the stored detection reference point And a deviation amount of the robot coordinate system reference is obtained by a conversion matrix of the camera coordinate system-torch coordinate system.

  As described above, in the laser sensor, the detected coordinate value is obtained in the camera coordinate system. In order to reflect in the work program, the detection point is converted into robot coordinates. At this time, if the attachment of the laser sensor to the welding torch is changed due to a laser sensor hitting or the like, the attachment positional relationship between the laser sensor and the torch will be deviated. The positional relationship with is also shifted.

  If the master work can be installed at the original position, it is only necessary to update the detection reference point of the sensing command. However, when there is no master work or when the installation position cannot be restored, it is necessary to correct the welding point one by one in addition to updating the detection reference point.

  Further, if the number of workpieces is large, the number of welding locations is large, or a large workpiece is a target, the time required for teaching correction becomes enormous. As described above, the conventional relative position method has a problem in handling the already taught points.

  The object of the present invention is to correct the teaching correction time even when the mounting position relationship with the welding torch is incorrect due to the attachment / detachment of the laser sensor, etc., when the position correction of the teaching point is performed by the relative position method using the laser sensor. It is an object of the present invention to provide a welding robot control device that can be greatly reduced.

  In order to solve the above problem, the invention according to claim 1 stores a work program in which detection reference points on the groove detected by a sensor provided in a welding torch are stored in a robot coordinate system. 1 storage means, second storage means for storing a current sensor-torch conversion matrix before changing the mounting position of the sensor relative to the welding torch, and a new sensor-torch conversion matrix after changing the mounting position of the sensor relative to the welding torch. A third storage means for storing the new sensor-torch conversion matrix and a difference conversion matrix based on the current sensor-torch conversion matrix and the new sensor-torch conversion matrix. And calculating means for updating the detection reference point included in the work program based on the difference transformation matrix. It is summarized as welding robot controller according to.

  According to a second aspect of the present invention, in the first aspect of the present invention, in the work program, mark data is described for each detection reference point, and the update unit searches the mark data and updates the detection reference point. It is characterized by that.

According to a third aspect of the present invention, in the second aspect, the landmark data is a sensing command.
According to a fourth aspect of the present invention, there is provided the selection means for selecting a work program to be updated according to the second or third aspect, wherein the update means includes the detection reference point included in the work program selected by the selection means. Is updated based on the difference transformation matrix.

  According to a fifth aspect of the present invention, in any one of the first to fourth aspects, the sensor is a laser sensor.

According to the first aspect of the present invention, even if the attachment position relationship between the sensor and the welding torch changes due to attachment or detachment of the sensor, the time required for teaching correction can be greatly reduced.
According to the invention of claim 2, the updating means can efficiently perform the updating process by searching the mark data described in the work program and updating the detection reference point.

  According to the invention of claim 3, the updating means can efficiently perform the updating process by searching for the sensing command described in the work program and updating the detection reference point.

According to the invention of claim 4, the update process can be efficiently performed by updating the detection reference point only for the selected work program.
According to the invention of claim 5, when the sensor is constituted by a laser sensor in particular, the effects of claims 1 to 4 can be easily realized.

(A) is the block diagram of the welding robot control system of one Embodiment, (b) is the schematic of the teach pendant TP similarly. (A) is explanatory drawing of a tool coordinate system, (b) is a perspective view of the laser sensor LS which shows the attachment state of the laser sensor LS. The block diagram of robot controller RC. Explanatory drawing of the relationship with a tool coordinate system, a sensor coordinate system, and a robot coordinate system. Explanatory drawing of sensing teaching. Explanatory drawing of welding teaching. Explanatory drawing of the work program of a relative position system. 6 is a flowchart of detection reference point update executed by the CPU. Explanatory drawing of a relative position system. Explanatory drawing of a relative position system. Explanatory drawing of a relative position system. The top view in the case of double-sided welding of a T-shaped workpiece. Explanatory drawing which shows the relative positional relationship of a workpiece | work and the laser sensor LS. Explanatory drawing which shows the relative positional relationship of a workpiece | work and the laser sensor LS.

Hereinafter, an embodiment of a welding robot control apparatus of the present invention will be described with reference to FIGS. Hereinafter, for convenience of explanation, the arc welding robot is simply referred to as a welding robot.
The welding robot control system 10 includes a manipulator M1 that performs a welding operation, a robot control device RC that controls the manipulator M1, and a laser sensor LS as a position sensor that detects the position of the workpiece W1.

  In addition, a teach pendant TP as a portable operation unit is connected to the robot controller RC. As shown in FIG. 1A, the teach pendant TP is provided with a keyboard 41 including a numeric keypad and various keys, and a display 42 including a liquid crystal display device. The teach pendant TP corresponds to selection means. Various teaching data are input to the robot controller RC by the keyboard 41.

  The manipulator M1 includes a base member 12 fixed to a floor or the like, and a plurality of arms 13 connected via a plurality of shafts. In the present embodiment, the manipulator M1 is a six-axis robot having J1 to J6 axes (not shown). As shown in FIG. 4, a robot output flange 40 is provided at the distal end portion of the arm 13 located on the most distal side. The robot output flange 40 is provided with a welding torch 14 via a support arm 50. The welding torch 14 includes a wire 15 as a filler material, generates an arc between the tip of the wire 15 fed by a feeding device (not shown) and the workpiece W1, and welds the wire 15 with the heat. Thus, arc welding is performed on the workpiece W1. A plurality of motors (not shown) are disposed between the arms 13, and the welding torch 14 can be freely moved back and forth and left and right by driving the motors. Note that front and rear refer to the direction in which the welding torch 14 travels along the groove as the front, and the opposite direction 180 degrees as the rear. Left and right are referred to as left and right with reference to the direction in which the person travels.

  The robot controller RC is composed of a computer as shown in FIG. That is, the robot controller RC is a rewritable EEPROM 21 that stores a CPU (central processing unit) 20, various programs for controlling the manipulator M1, and a plurality of shape analysis programs prepared according to various groove shapes, A RAM 22 serving as a working memory and a storage unit 23 including a rewritable nonvolatile memory for storing various data are provided.

The CPU 20 corresponds to an update unit and a calculation unit. The storage unit 23 corresponds to a first storage unit, a second storage unit, and a third storage unit.
The storage unit 23 includes storage areas such as a first storage area 23a, a second storage area 23b, a third storage area 23c, a fourth storage area 23d, and a fifth storage area 23e.

  The first storage area 23a is an area for storing distance information (ranging data) obtained by measuring the visual field range FOV (see FIG. 4) with the laser sensor LS. In the second storage area 23b, teaching data such as a target angle, advancing / retreating angle, and a teaching point position input by the keyboard 41 of the teach pendant TP at the time of teaching is stored as a file. The third storage area 23c is an area for storing a work program in which a teaching instruction is described for each step. The work program will be described later.

The fourth storage area 23d is for storing the current sensor-torch conversion matrix _C T _1T before the sensor calibration when the sensor calibration is performed. When the first sensor calibration is performed, the conversion matrix obtained at that time is stored in the fourth storage area 23d and a fifth storage area 23e described later. When the second and subsequent sensor calibrations are performed, the fourth storage area 23d is rewritten with the contents of the fifth storage area 23e. In the fifth storage area 23e, a new sensor-torch conversion matrix _C T _2T acquired when the sensor calibration is performed after the second time is updated and stored.

  The robot controller RC controls the motor to drive the welding torch 14 along the main track of preset teaching data. The position and orientation of the welding torch 14 can be calculated by the CPU 20 based on the encoder provided in the motor and the link parameter of each arm 13.

  Further, the robot controller RC outputs welding conditions such as a welding current and a welding voltage to the welding power source WPS, and causes the welding operation to be performed by electric power supplied from the welding power source WPS through the power cable PK.

  The laser sensor LS is a scanning type laser displacement sensor that measures the distance to the workpiece W1 by emitting and receiving light of a laser, and is detachable from the welding torch 14 via a bracket 70 as shown in FIG. Installed. The laser sensor LS includes a light emitting unit that emits a laser beam toward the workpiece W1, a light receiving unit that receives the laser beam reflected by the workpiece W1, and the like (both not shown). The laser emitted from the light emitting unit is reflected by the workpiece W1 and received by the light receiving unit. The light receiving unit is constituted by, for example, a CCD line sensor (line laser sensor), and measures the distance from the laser sensor LS to the workpiece W1 in the visual field range FOV.

  The robot controller RC controls the laser sensor LS, and detects the groove position by analyzing the shape (distance information) between the laser sensor LS and the workpiece W1 according to the shape analysis program.

The sensor head LSa is attached to the laser sensor LS so that the laser irradiation direction is parallel to any axis of the tool coordinate system. FIG. 2A shows a welding torch 14 as a tool. Here, as shown in FIG. 2A, the tool coordinate system is expressed as the Z axis aligned with the axis of the welding torch 14. In this embodiment, as shown in FIG. 2B, the sensor head LSa of the laser sensor LS with respect to the welding torch 14 has a laser irradiation direction approximately in the Z-direction, and the welding torch 14 The welding progress direction is set to be approximately the X axis in the tool coordinate system, and the laser sensor LS is attached to be approximately parallel to the X axis. The laser sensor LS (that is, the sensor head LSa shown in FIG. 2B) is configured to irradiate a laser beam at a position spaced a predetermined distance from the tip of the welding torch 14 toward the welding direction. The laser sensor LS is not necessarily attached in the X axis + direction. Regardless of the mounting direction, the mounting relationship of the laser sensor LS with respect to the welding torch 14 is stored as the new sensor-torch conversion matrix _C T _{1T in the} storage area 23d.

  In addition, before the welding operation is performed, teaching data indicating the operation of the manipulator M1 when welding is performed, welding conditions, and the like are input to the robot controller RC via the teach pendant TP, and a work program is stored. It is created and stored in the third storage area 23 c of the storage unit 23. In the following, unless otherwise specified, “teaching” means inputting using the teach pendant TP.

(About work program)
First, an example of a relative position type work program stored in the third storage area 23c will be described with reference to the work program shown in FIG.

  In the following description, “position the welding torch” or “position the robot” is detachably attached to the tip of the wire 15 protruding from the tip of the welding torch 14 or the tip of the welding torch 14. This means that the tip of the teaching jig (not shown) is positioned at each teaching point (original position, sensing point, welding start point, welding end point). For convenience of explanation, the case of a fillet joint in which a pair of iron plates are arranged in an L shape will be described as the master work W and the work W1 to be actually welded, but the types of joints are limited. It is not a thing.

  In the teaching mode, the operator uses the master work W to operate the teach pendant TP and drive and control the manipulator M1, thereby moving the welding torch 14 to the original position G, the master as shown in FIGS. The sensing points SN1 and SN2 for detecting the groove of the workpiece W, the welding start point R2, the welding end point R1 and the original position G are taught as teaching points in this order, and at each teaching point, necessary instructions are given. Alternatively, teaching data is input. The teaching data taught at each teaching point will be described later.

  The sensing point SN1 is the position of the welding torch 14 for detecting the welding end point R1 in the groove of the master work W. When the welding torch 14 is positioned at this position, the operator can issue a positioning command (or a straight line). An interpolation command) and a “sensing command & pause calculation command” are input by operating the keyboard 41 of the teach pendant TP. The sensing point SN2 is the position of the welding torch 14 for detecting the welding start point R2 in the groove of the master work W. When the welding torch 14 is positioned at this position, the operator can issue a positioning command (or linear interpolation). Command) and “sensing command & pause calculation command” are input by operating the keyboard 41 of the teach pendant TP.

  By operating the teach pendant TP, the operator teaches roughly at the sensing points SN1 and SN2 by positioning the robot so that there is a point where sensing is desired within the visual field range FOV of the laser sensor LS. do it. On the other hand, with respect to the welding end point R1 and the welding start point R2, the welding torch is positioned so as to accurately aim at a desired target position.

Next, a case where welding is performed on the workpiece W1 using the work program created in the teaching mode as described above will be described.
In step 1 shown in FIG. 7, the CPU 20 moves the welding torch 14 to the original position G by a “positioning command” (see FIG. 5). 5 and 6, the master work is indicated by W. However, for convenience of explanation, the reference sign W1 is used when used for the meaning of the work to be actually welded.

  In step 2, the CPU 20 causes the welding torch 14 to be positioned at the sensing point SN1 shown in FIG. 5 and to be in the posture taught in the teaching mode according to the “positioning command (or linear interpolation command)”. The sensing point SN1 is a sensing point on the welding end side.

  In step 3, a “sensing command” is taught at sensing point SN1. By this “sensing command”, the laser sensor LS detects a detection reference point Q1 of a welding end point R1 described later on the groove of the workpiece W1. That is, the CPU 20 detects the groove position by performing shape analysis on the distance measurement data detected in the range of the visual field range FOV according to the shape analysis program, and sets the detected groove position as the detection reference point Q1.

More specifically, the CPU 20 converts the sensing point SN1 into the robot coordinate system using the transformation matrix _R T _T between the robot coordinate system and the tool coordinate system as shown in FIG. A point obtained by shifting the detection point coordinate output by is set as a detection reference point Q1. In this case, a point obtained by shifting the detection point coordinate output from the laser sensor LS is used after being multiplied by the current sensor-torch conversion matrix _C T _1T stored in the storage area 23d between the welding torch 14 and the laser sensor LS. Is done. As a result, the detection reference point Q1 is represented by a coordinate value in the robot coordinate system.

For convenience of explanation, a description will be given collectively here, but a detection reference point Qn1 and a detection reference point Qn2 (both not shown), which will be described later, are similarly stored in the current sensor-torch conversion matrix _C T stored in the storage area 23d. _The value is multiplied by _1T .

  If the detection reference point acquired at the sensing point SN1 on the welding end side is Qn1 using the master workpiece W in the teaching mode, the detection reference point Qn1 is the coordinate value of the robot coordinate system as with the detection reference point Q1. In addition, the default value is stored in a predetermined area of the second storage area 23b of the storage unit 23 in association with the “sensing instruction” of step 2.

  Here, the tool coordinate system is a coordinate system in which the Z axis is aligned with the axis of the welding torch 14, and the robot coordinate system is a coordinate system with an arbitrary point of the base member 12 as the origin. The sensor coordinate system is a coordinate system with an arbitrary part of the sensor as the origin.

  The CPU 20 calculates a shift amount L3 of each coordinate value between the detection reference point Qn1 obtained with the master work W and the detection reference point Q1 obtained with the work W1, and calculates the shift amount L3 as a shift correction file No. 2 and stored in the second storage area 23b of the storage unit 23.

  In step 4, the CPU 20 causes the welding torch 14 to be positioned at the sensing point SN <b> 2 shown in FIG. 5 and the posture taught in the teaching mode by the “positioning command (or linear interpolation command)”. The sensing point SN2 is a sensing point on the welding start side.

In step 5, a “sensing command” is taught at sensing point SN2.
By this “sensing command”, the laser sensor LS detects a detection reference point Q2 of a welding start point R2 described later on the groove of the workpiece W1. That is, the CPU 20 of the robot controller RC detects the groove position by performing shape analysis on the distance measurement data detected in the range of the visual field range FOV in accordance with the shape analysis program, and detects the detected groove position as the detection reference point Q2. To do.

More specifically, the CPU 20 converts the sensing point SN2 into the robot coordinate system using the transformation matrix _R T _T between the robot coordinate system and the tool coordinate system, and then the detected point coordinate output from the laser sensor LS. The point shifted from is set as a detection reference point Q2. In this case, a point obtained by shifting the detection point coordinate output from the laser sensor LS is used after being multiplied by the current sensor-torch conversion matrix _C T _1T stored in the storage area 23d between the welding torch 14 and the laser sensor LS. Is done. As a result, the detection reference point Q2 is represented by a coordinate value in the robot coordinate system.

  In the teaching mode, when the detection reference point acquired at the sensing point SN2 on the welding start side is Qn2 using the master work W, the detection reference point Qn2 is the coordinate of the robot coordinate system as with the detection reference point Q2. As a default value, it is stored in a predetermined area of the second storage area 23b of the storage unit 23 in association with the “sensing instruction” in step 4.

  The CPU 20 calculates a deviation amount L4 between the detection reference point Qn2 obtained with the master work W and the detection reference point Q2 obtained with the work W1, and the deviation amount L4 is calculated as a deviation correction file No. 1 and the deviation correction file No. 1 is stored in the second storage area 23 b of the storage unit 23.

In Step 6, the CPU 20 positions the welding torch 14 at the approach point P6 by a “positioning command” that moves to the approach point P6.
In step 7, in response to the “shift start command”, the CPU 20 reads the correction file No. stored in the third storage area 23c. 1 is read, and the deviation correction file No. 1 is read out. The welding torch 14 is shifted by the deviation amount L3 described in 1.

  In step 8, linear interpolation is performed by the “linear interpolation command” so as to be positioned at the taught welding start point R2 at the moving speed that is the teaching data stored in the second storage area 23b. The taught welding start point R2 is actually changed (shifted) by the shift amount L4. At this time, the CPU 20 reads the aim angle and the advance / retreat angle, which are teaching data stored in the second storage area 23b of the storage unit 23, from the second storage area 23b, so that the welding torch 14 assumes this posture, and The manipulator M1 is controlled to move so as to be positioned at the welding start point R2.

  In step 9, the CPU 20 reads the welding conditions such as the welding current, welding voltage, and welding speed as teaching data from the second storage area 23b in response to the “welding start command”, and these welding conditions are shown in FIG. Output to the wire feeder.

  The welding power source WPS controls the power so that the welding current and the welding voltage are obtained, and a wire feeding device (not shown) controls the wire 15 to be sent to the welding torch 14 at the feeding speed.

  In step 10, the CPU 20 calculates a shift amount by linear interpolation (to be described later) in order to change the shift amount according to the “shift amount change command”. That is, the change of the shift amount is to gradually change the shift amount when proceeding from the welding start point R2 based on the shift amount L4 and the shift amount L3. That is, the CPU 20 adds or subtracts the difference between the deviation amount L3 and the deviation amount L4 by an amount equally divided by the interpolation cycle of linear interpolation in step 11 to be described later to change the shift linearly. It is.

  In accordance with the “linear interpolation command” in step 11, the CPU 20 linearly interpolates the welding end point R 1 so that the shift changes linearly, and drives and controls the manipulator M 1 at the welding speed that is teaching data to control the welding torch 14. Move.

  At this time, in step 5, the CPU 20 moves and controls the manipulator M1 so that the welding torch 14 is positioned up to the welding end point R1 with the attitude of each axis angle of the manipulator M1 stored in the second storage area 23b.

  In step 12, the CPU 20 reads the welding conditions such as the welding current and the welding voltage, which are teaching data, in accordance with the “welding end command”, and outputs these welding conditions to the welding power source WPS and a wire feeder (not shown). The welding power supply WPS controls the power supply so that the welding current and the welding voltage are obtained, and stops feeding the wire 15 by a wire feeding device (not shown).

In step 13, the CPU 20 stops the shift described above by “end of shift instruction”.
In step 14, in accordance with the “linear interpolation command”, the CPU 20 linearly interpolates to the retract point P14, and drives and controls the manipulator M1 at the operation speed that is the teaching data to move the welding torch 14.

In step 15, the CPU 20 drives and controls the manipulator M <b> 1 to the original position G by the “positioning command” to move the welding torch 14.
(Function)
The operation of the welding robot control system 10 configured as described above will be described.

  When the attachment of the laser sensor LS to the welding torch 14 has changed, for example, when the attachment position of the laser sensor LS with respect to the welding torch 14 is changed, the operator operates the teach pendant TP to detect it. Select the reference point correction mode. In the detection reference point correction mode, the worker displays the work program stored in the storage unit 23 of the robot controller RC on the display 42 by the input of the keyboard 41. Then, the operator selects in advance a work program whose detection reference point is to be updated by selection input of the keyboard 41 of the teach pendant TP, and then executes the detection reference point correction program as shown in FIG.

In S10, the CPU 20 performs known sensor calibration between the laser sensor LS and the welding torch 14, obtains a new sensor-torch conversion matrix _C T _2T, and stores it in the fifth storage area 23e. In S20, the CPU 20 stores the contents (current sensor-torch conversion matrix _C T _1T ) stored in the fifth storage area 23e before the sensor calibration is performed in the fourth storage area 23d.

In S30, the CPU 20 calculates the difference conversion matrix _O T _{N using} the following equation.
_O T _N = ( _C T _1T ) ⁻¹ · _C T _2T
In FIG. 4, the sensor coordinate system before reattachment is indicated by (X _C , Y _C , Z _C ), and the sensor coordinate system after reattachment is indicated by (X _CN , Y _CN , Z _CN ). The two sensor coordinate systems, the transformation matrices _C T _1T , _C T _2T , and the difference transformation matrix _O T _N are as shown in the figure.

  In S <b> 40, the CPU 20 searches for a step in which a “sensing instruction” is described and a detection reference point associated with the “sensing instruction” for the work program selected in advance. In the work program mentioned in the above example, detection reference points Qn1 and Qn2 (both not shown) associated with the “sensing command” in steps 3 and 5 are found by the search.

In S50, the CPU 20 multiplies the detected detection reference points (detection reference points Qn1 and Qn2 in the above example) by the difference transformation matrix _O T _N to update the detection reference points. In other words, the detection reference point associated with the “sensing command” is updated and stored in the second storage area 23b. As a result, the work program is an update of the detection reference point obtained by the master work W.

According to the welding position detection method and the welding position detection apparatus of this embodiment, there are the following features.
(1) In the robot controller RC of the present embodiment, detection reference points Qn1 and Qn2 (both not shown) on the groove detected by the laser sensor LS provided on the welding torch 14 are described in the robot coordinate system. A storage unit 23 (first storage means) for storing the work program is provided. The storage unit 23 stores the current sensor-torch conversion matrix _C T _1T before changing the mounting position of the laser sensor LS with respect to the welding torch 14 as second storage means. Further, when the sensor calibration for obtaining the new sensor-torch conversion matrix _C T _2T after changing the mounting position of the laser sensor LS with respect to the welding torch 14 is performed as the third storage means, the storage unit 23 receives the new _sensor- A torch transformation matrix _C T _2T is stored. Furthermore, the robot controller RC includes a CPU 20 (calculation means) that calculates a difference conversion matrix _O T _N based on the current sensor-torch conversion matrix _C T _1T and the new sensor-torch conversion matrix _C T _2T . Further, the CPU 20 updates the detection reference points Qn1 and Qn2 included in the work program based on the difference transformation matrix _O T _N as an update unit.

  As a result, according to the present embodiment, in a system that reflects the amount of relative displacement with respect to the detection reference point of the sensor in the welding teaching point, the welding position is accurately taught when a work program including a sensing command is newly taught. If so, the time required for teaching correction by attaching and detaching the sensor head can be greatly reduced.

  (2) In the robot controller RC of the present embodiment, the work program describes “sensing instructions” as mark data for each of the detection reference points Qn1 and Qn2 obtained by using the master work W. The CPU 20 searches for a “sensing command” and updates the detection reference points Qn1 and Qn2.

As a result, in this embodiment, the CPU 20 searches for a sensing command described in the work program and updates the detection reference point, so that the update process can be performed efficiently.
(3) The robot controller RC of the present embodiment includes a teach pendant TP (selection unit) that selects a work program to be updated, and the CPU 20 (update unit) is included in the work program selected by the teach pendant TP. The detection reference points Qn1 and Qn2 are updated based on the difference conversion matrix.

As a result, according to the present embodiment, the CPU 20 updates the detection reference point only for the selected work program, so that the update process can be performed efficiently.
(4) The robot control device RC of the present embodiment uses a laser sensor LS as a sensor. For this reason, the effects (1) to (3) can be easily realized in the welding robot including the laser sensor LS.

In addition, this invention is not limited to the said embodiment, You may comprise as follows.
In the embodiment described above, the detection reference point associated with the “sensing command” is updated by searching for the “sensing command” as the mark data. However, the mark data is not limited to the sensing command. In the step where the sensing command of the work program is described, in order to search for the detection reference point, mark data other than the “sensing command” is described, the mark data is searched, and the detection reference point is updated. It may be.

  In the embodiment, the work program for updating the detection reference point is selected, but instead, all the work programs stored in the storage unit 23 are searched for and updated in a batch. You may do it.

  -Although the line laser sensor was used for the laser sensor LS of the said embodiment, you may replace with a scanning type laser displacement sensor or a spot laser sensor which scans with a laser hitting a mirror.

In place of the laser sensor LS of the embodiment, the touch sensor described in the related art may be changed.
In the above embodiment, the manipulator M1 is a 6-axis robot, but may be configured by a multi-axis robot such as a 7-axis robot.

  In the above-described embodiment, the robot coordinate system is a coordinate system having an arbitrary part of the base member 12 as the origin, but the robot coordinate system is, for example, a robot coordinate system, as shown in FIG. A mechanical interface coordinate system having an origin on the J6 axis of the robot output flange 40 may be used.

In this case, in step 3, the CPU 20 uses the transformation matrix _J6 T _T between the robot coordinate system (mechanical interface coordinate system) and the tool coordinate system as shown in FIG. It is assumed that a point obtained by shifting the detection point coordinates output by the laser sensor LS after the conversion to is a detection reference point Q1.

  Thus, the robot coordinate system may be a coordinate system provided on the manipulator M1 side.

14 ... welding torch,
20 ... CPU (calculation means, update means),
23 ... Storage section (first storage means, second storage means, third storage means),
M1 ... manipulator, TP ... teach pendant (selection means),
RC: Robot control device.

Claims (5)

First storage means for storing a work program in which detection reference points on a groove detected by a sensor provided in a welding torch are stored in a robot coordinate system;
Second storage means for storing a current sensor-torch conversion matrix before a change in the mounting position of the sensor with respect to the welding torch;
A third storage means for storing the new sensor-torch conversion matrix when sensor calibration is performed to obtain a new sensor-torch conversion matrix after changing the mounting position of the sensor relative to the welding torch;
Calculating means for calculating a difference conversion matrix based on the current sensor-torch conversion matrix and the new sensor-torch conversion matrix;
A welding robot control apparatus comprising: an updating unit configured to update the detection reference point included in the work program based on the difference conversion matrix. In the work program, landmark data is described for each detection reference point,
The welding robot control apparatus according to claim 1, wherein the update unit searches the mark data and updates the detection reference point.   The welding robot control device according to claim 2, wherein the mark data is a sensing command. A selection means for selecting a work program to be updated;
The welding robot control device according to claim 2, wherein the update unit updates the detection reference point included in the work program selected by the selection unit based on the difference transformation matrix. .   The welding robot control device according to any one of claims 1 to 4, wherein the sensor is a laser sensor.

Images