WO1998039131A1

WO1998039131A1 - Welding robot system

Info

Publication number: WO1998039131A1
Application number: PCT/JP1998/000879
Authority: WO
Inventors: Akihiro Terada; Mitsuhiro Okuda
Original assignee: Fanuc Ltd
Priority date: 1997-03-03
Filing date: 1998-03-03
Publication date: 1998-09-11
Also published as: JPH10244367A

Abstract

A laser sensor (2) and a welding torch (3) are mounted to a robot terminal end (1) to perform welding on joints (A, B). The laser sensor (2) scans the surfaces of the joints to periodically perform detection of a weld line position and detection of a gap length. A robot periodically outputs to the sensor (2) current sensor position data with time stamp, and the weld line position is found as robot data. A weaving condition corresponding to a range of the gap length g (x) thus detected is selected, and a torch tip end (4) draws a path WV, in which tracking and weaving are superposed.

Description

Specification

Welding robot system

Technical field

The present invention relates to a welding robot system equipped with an arc welding torch and a laser sensor. More specifically, the present invention relates to a weaving operation and real-time tracking (welding line following) while adapting to a gap width of a welding joint. It relates to a welding robot system that is to be carried out in parallel.

Background technology

Arc welding A laser sensor is mounted on a robot with a torch attached to the hand (welding robot), and the robot is moved along the welding line while sensing the position of the welding line ahead of the welding point in the traveling direction. This technique is called real-time tracking in a welding robot. If this technology is applied, it is possible to move the mouth pot along the actual welding line even if the position of each joint to be welded is not accurate, so that the welding quality can be improved. Widely used as a means to increase

On the other hand, for a welding robot that performs welding to a thick plate or welding to a joint where the existence of a gap is predicted, weaving operation must be performed.ゥ The eving operation is an operation that periodically shifts the robot's movement path from the reference path (ゥ the path corresponding to an eaves amount of “0”) to the side. Therefore, the welding robot to which the real-time tracking is applied is When performing the weaving operation, it is necessary to realize a movement path in which the weaving amount is added to the position of the welding line detected by the laser sensor (the position of the weaving amount “0”).

If the gap length of the joint fluctuates along the welding path, weaving operation conditions (weaving amplitude, weaving cycle, weaving type (sine curve, figure 8, figure, zigzag, etc.), and the presence or absence of weaving , Etc.) is expected to be changed during the movement.

For example, a weaving operation is performed with a large weaving amplitude in a moving section with a large gap length, and a weaving operation is performed with a small weaving amplitude in a moving section with a small gap length (in some cases, no weaving operation is performed). Thus, the quality of welding can be improved.

As described above, the idea of switching the weaving conditions according to the gap length of the joint when tracking and weaving are used has been conventionally proposed. However, it is not easy to make a real welding robot perform such a complex operation, and the technology to realize this is not yet known.

In order to switch the weaving conditions according to the gap length while actually using both accurate tracking and weaving, objective profile information that serves as basic data for determining the welding line position and gap length (robot It is necessary to obtain the profile information expressed on the dot coordinate system) with high accuracy.

However, what is obtained from the detection output of the laser sensor is sensor data representing the profile information of the joint on the sensor coordinate system that moves integrally with the robot during weaving operation. It is not possible to determine the position of the weld line above (or on another objective coordinate system)-the gap. Therefore, it is necessary to determine the position of the welding line and the gap length based on objective profile information by properly associating the detected output of the laser sensor with the rod position at the time when the sensing corresponding to the detected output is performed. There is.

In addition, when switching the weaving conditions by detecting a change in the gap length, it is necessary to take care that the trajectory of the mouth bottle is not disturbed by the switching of the weaving conditions.

Disclosure of the invention

An object of the present invention is to provide a welding robot system capable of performing a weaving operation in parallel under a weaving condition corresponding to a gap length while accurately tracking a welding line.

Also, the present invention enables the position and gap length of the welding line to be accurately determined even during the weaving operation, and through that, weaving conditions according to the gap length are used together with weaving for accurate tracking. Welding robot system that can be switched It is to be offered.

Further, the present invention also intends to make it possible to smoothly change the trajectory of the welding robot system in accordance with the switching of the weaving conditions. And, through these, the present invention aims to improve the quality of welding for joints whose gap length is not constant.

The present invention relates to a robot equipped with a welding torch, a laser sensor having a main body attached to the robot so as to sense a forward region of the robot with respect to the welding torch, and a mouth port. to improve the welding robot system comprising a control unit is intended to achieve the object _c that is, the position data of the control means of the mouth pots laser sensors Raberi ring in Thailand Musuta pump laser sensor Means for outputting the data as the lopot data. On the other hand, the laser sensor is a means for detecting the profile of the joint to be welded based on the position data and the sensing result of the laser sensor thus labelled, and the position of the welding line based on the detected profile. There is a means for calculating and storing the gap length.

The robot control means further includes means for determining a correction amount for tracking the welding line based on the stored welding line position, and a weaving selected according to the stored gap length. The means for determining the weaving amount according to the conditions, the calculated correction amount for tracking and There is provided a means for determining a movement target position of the mouth-bottom based on the weaving amount and the interpolation position on the teaching path.

In a preferred embodiment of the present invention, the control means of the mouth pot further includes a weaving amount such that the position of the robot does not suddenly change when the selected weaving condition is switched according to the detected gap length. It has a means to determine

ADVANTAGE OF THE INVENTION According to the welding robot system of this invention, about the welding joint which a gap length variation is predicted, accurate real-time tracking along a welding line, and the weaving conditions controlled according to the gap length variation. Weaving can be performed in parallel. In addition, when the weaving conditions are switched, a smooth movement of the welding point can be guaranteed.

Brief explanation of drawings

FIG. 1 is a diagram for explaining an embodiment when welding is performed using the welding robot system according to the present invention.

FIG. 2 is a block diagram showing an outline of the welding robot system shown in FIG.

FIG. 3 is a diagram illustrating the measurement principle of the laser sensor used in the welding robot system of FIG.

FIG. 4 is a diagram showing a configuration of a laser sensor used in the welding robot system of FIG.

FIG. 5 is a block diagram showing the configuration of the robot control board shown in FIG. FIG. 6 shows an example of a screen for setting the weaving conditions.

FIG. 7 shows an example of a screen for setting the selection criteria for the weaving condition related to the gap length range.

8A, 8B, and 8C are diagrams illustrating weaving patterns.

FIG. 9 is a flowchart showing the content of processing 1 performed in the sensor control board of FIG.

FIG. 10 is a flowchart showing the content of the process 2 performed in the sensor control board of FIG.

FIG. 11 is a flowchart showing the content of the process 3 performed in the mouth port control board of FIG.

FIG. 12 is a flowchart showing the contents of a process 4 performed in the robot control board of FIG.

BEST MODE FOR CARRYING OUT THE INVENTION

An embodiment in which welding is performed using the welding robot system according to the present invention and a gap length is detected at the same time will be described with reference to FIG.

In Fig. 1, the workpieces A and B that make up the joint are positioned on the work table WT using a jig (not shown). This joint has a gap G extending substantially along the X-axis direction of a peak coordinate system set on the robot. The width of the gap G is the distance in the Y-axis direction orthogonal to the X-axis direction. The width of the gap G is not constant, but changes according to the X-axis coordinate value X. That is, some The width of the gap at the position x can be represented by g (X). _C g (X) = 0 indicates that there is no gap.

In Fig. 1, most of the body of the lopot is omitted, and only the tip 1 of the mouth pot is shown. A laser sensor 2 and a welding torch 3 are mounted on the robot hand 1 via an appropriate mounting mechanism (not shown). At the tip of the welding torch 3, a tool tip point (TCP) 4 of the robot, which is a welding point, is set. The TCP 4 draws a locus W V as shown in FIG. 1 in which weaving and tracking are superimposed.

The scanning beam 5 from the laser sensor 2 scans an area that precedes the welding point 4 in the robot traveling direction. The trajectories 6 A and 6 B of the bright spots formed on the surfaces of the workpieces A and B and the work table WT by the scanning beam 5 are detected by the light detecting section of the laser sensor 3, and based on the detected positions, the positions of the welding lines and the gaps are determined. The width g (X) is obtained.

Next, the outline of the welding robot system shown in FIG. 1 will be described with reference to the block diagram of FIG.

As shown in Fig. 2, the welding robot system consists of a sensor body 10, a sensor control board 20, a robot control board 30, a robot body 40, and a welding part (power supply unit) 50. Is done. The sensor body 10 and the sensor control port 20 constitute the laser sensor 2 shown in FIG. The robot control board 30 controls the entire system. Sensor control board 20 and robot control board 30 connect bus 29 Connected through.

The configuration of the laser sensor 2 composed of the sensor body 10 and the sensor control board 20 will be described with reference to FIG.

The sensor main body 10 includes a laser oscillator 11 and an oscillating mirror (galvanometer) 12 for beam scanning. Further, a light detection unit including an optical system 13 for imaging and a light receiving element 14 is provided.

On the other hand, the sensor control board 20 includes a CPU 21 composed of a microprocessor, and the CPU 21 has a memory composed of an input / output device 28, a ROM, a RAM, and the like via a bus 29. 2 5 are connected.

The input / output device 28 has a laser driver 22 for driving a laser oscillator 11 to generate a laser beam, a mirror scanner 23 for oscillating a oscillating mirror 12, and a light receiving element 14 for receiving light. A signal detection section 24 for detecting the position of the reflection point S of the scanning beam from the detected position is connected.

The sensor control board 20 is connected to the robot control board 30 via a bus 29. When the sensor control port 20 receives a laser sensor start command from the robot control board 30 via the nozzle 29, the CPU 21 starts the laser sensor drive program stored in the memory 25. Then, a laser drive command is sent to the laser drive unit 22 and a mirror scan command is sent to the mirror scanning unit 23. Thus, the object to be inspected is scanned with the laser beam 5. The laser beam diffusely reflected at the reflection point S on the object surface forms an image on the light receiving element 14 by the optical system 13 according to the reflection position S. As the light receiving element, a CCD (Charge Coupled Device) as a split type element, a PSD (Position Sensitive Detector) as a non-split type and integrating type element, and the like are used. Here, a one-dimensional CCD array of a laser sensor is used as the light receiving element 14 (however, a two-dimensional array can also be used).

The light (image of the reflected light) incident on the light receiving surface of the light receiving element 14 is converted into photoelectrons and stored in the cell. The electric charge accumulated in the cell is output in order from the first end at predetermined intervals in accordance with the CCD scanning signal from the signal detection unit 24, and is transmitted through the signal detection unit 24 and the input / output device 28, such as for AD conversion. After being processed, the latest data is sequentially stored in memory 25.

The scanning cycle of the CCD is set to be sufficiently shorter than the scanning cycle of the oscillating mirror 12 (for example, 1/100), and the transition of the oscillating angle of the oscillating mirror 12 and the output state of the CCD element are changed. Changes can be monitored at any time. The output state of the CCD element is grasped by the cell position (cell number) with the maximum output, and the cell position where the light reflected from the reflection point S is detected is detected. <The position of the reflection point S is calculated from this cell position. .

Fig. 3 shows that when the light receiving element 14 of the laser sensor detects the reflection point S at the position Xa, the coordinate position (Xs, Ys) of the reflection point S can be obtained from the value of the position Xa. Is explained. Set the sensor origin O s (0, 0) in the optical system. Set on the line connecting the center and the center point of the light receiving element 14. This line is defined as the Y s axis, and a line extending from the origin O s in a direction orthogonal to the Y s axis is defined as the X s axis.

Also, the distance from the sensor origin O s to the center of the optical system 13 is L 1, the distance from the center of the optical system 13 to the center point of the light receiving element 14 is L 2, and the sensor origin O s to the X s axis D is the distance from the oscillating mirror 1 to the oscillating center of the direction 2 and L 0 is the Y s-axis distance from the sensor origin to the oscillating mirror's oscillating center. Is 0 with respect to the Y s axis direction.

Then, the coordinate position (xs, ys) of the reflection point S of the laser beam 5 on the sensor coordinate system can be obtained by solving the following simultaneous equations (1) and (2).

x a Z L 2 = x s Z (L 1 — y s) ■ ■ ■ (1)

(L 0-ys) tan Θ = D-xs ■ (2) CPU 21 of sensor control board 20 stores in memory 25 according to the command from robot control board 30 The calculated position calculation program is started, and a process of obtaining sensor data xs, ys representing the position of the reflection point S from the above equations (1) and (2) is executed at a predetermined cycle. The obtained sensor data xs and ys are temporarily stored in the memory 25 with a time stamp. On the other hand, sensor position data with a time stamp is sent from the robot control board 30 via the bus 29 at a predetermined cycle. Therefore, the sensor data indicating the position of the reflection point S with the time stamp is The position of the reflection point S on the mouth-pot coordinate system is calculated by comparing the time stamps with xs, ys and the sensor position data with the time stamp. You. The obtained results are sequentially stored in memory 25.

Next, the configuration of the lopot control board 30 will be described with reference to the block diagram of FIG.

The robot control board 30 has a CPU 31 composed of a microprocessor, and the CPU 31 is connected to the CPU 21 of the sensor control board 20 via the bus 29. In addition, the CPU 31 is provided with a memory 32 consisting of ROM, a memory 33 consisting of RAM, a nonvolatile memory 34, a non-volatile memory 34, and a teaching operation panel 38 via a node 29. The axis control unit 35 and the general-purpose interface 39 are connected. The teaching operation panel 38 is equipped with a liquid crystal display section 37, the axis control section 35 is connected to a welding robot (body mechanism section) 40 via a servo circuit 36, and is welded to a general-purpose interface 39. Is connected to the power supply device 50.

The ROM 32 stores a system program for controlling the entire system including the sensor control board 20, the mouth port control board 30, the robot body 40, and the power supply 50. ing. The RAM 33 is a memory used for temporary storage of data and calculation. The non-volatile memory 34 stores programs for instructing the operation of the system including various parameter setting values and the robot body 40. You.

The parameter setting values include a parameter that describes a plurality of weaving conditions, a table that defines the relationship between the gap length range and the weaving conditions to be selected (weaving condition selection criteria), and the like. These are set on the LCD display 37 attached to the teaching operation panel 38 of the robot control board 30. Figures 6 and 7 show examples of this screen setting.

In the screen shown in FIG. 6, the numbers 1 to 11 and 99 in the leftmost row represent setting condition numbers, and the condition contents of each number are input by the operator. However, conditions No. ₉ 9 a shall in case the gap exceeds the allowable value, is accompanied by output and emergency stop of the alarm signal.

The settings from condition 1 to condition 11 in FIG. 6 include the weaving pattern, the weaving frequency f (the reciprocal of the period Tw, 1ZTW), and the weaving amplitude a. Also, the stop time at the left end and the right end of the weaving locus can be specified.

Various weaving patterns are used.Here, weaving pattern 1 shown in Fig. 8A, weaving pattern 2 shown in Fig. 8B, and weaving pattern 3 shown in Fig. 8C are used. A pattern is given. Figures 8A-8C show weaving patterns with linear movement superimposed on weaving. ing. In these patterns, the symbol λ represents the amount of movement of the rod per period Tw.

The weaving pattern 1 shown in FIG. 8A is as follows. That is, a displacement (weaving amount) of a sinusoidal waveform having an amplitude a and a frequency f (= 1 / Tw) is given in a direction perpendicular to the traveling direction programmed in the robot. However, at the right and left ends of the weaving locus, the weaving amount is maintained at a constant value (± a) only during the stop times Tsr and Tsi.

Considering the condition of time t = 0 and weaving amount w (0) = 0, for example, the weaving amount w (t) at time t in one cycle of weaving pattern 1 is as follows. Become. As described later, the time t for calculating the weaving amount is determined by a weaving timer set in the robot control board 30. The weaving timer is started or reset at the start of the weaving operation and at the time of switching.

I: When time t is 0≤t≤TwZ4,

w (t) = a s i n (2 π t no T w)

---(1)

: When the time t is T w '4 <t ≤ (T w / 4) + T s r,

w (t) = a (2)

III: Time t is (T w / 4) + T s r <t

≤ (3 Τ w / 4) + T sr, w (t) = asin (2π [t-T sr] / Τ w) ■ ■ (3)

IV: Time t is (3TwZ4) + Τsr <t

≤ (3 Tw / A) + Tsr + TsI,

w (t) = — a ■ ■ ■ (4)

V: Time t is (3Tw4) + Tsr + Tsl <t

<T w + T sr + T s l

w (t) = a s inn (2π [t-one T sr—

T s i] Z T w)

■ ■ ■ (5) Also, in weaving pattern 2 (arc) and ゥ eving pattern 3 (polyline), which are represented by periodic functions u (t) other than the sine function (Fig. 8A), within one cycle In general, the weaving amount w (t) at time t is as follows. Note that the amplitude of the periodic function u (t) is a, and the period is Tw.

I: When time t is 0 ≤ t ≤ T w 4,

w (t) = u (t) • ■ · (6)

II: When time t is T w no 4 <t ≤ (T w 4) + T sr,

w (t) = a (7)

III: Time t is (T w no 4) + T sr <

When ≤ (3Tw / 4) + Tsr,

w t) = u (t — T sr) (8)

IV: The time t is (3TwZ4) + Tsr < ≤ (3 T w / 4) + Τ sr + Τ s I,

w (t) =-a---(9)

V: The time t is (3Tw / 4) + Tsr + Tsl <t

<T w + T sr + T sl,

w (t) = u (t-one T sr— T sl)

■ ■ ■ do) Based on the above, refer to the flowcharts in Fig. 9 to Fig. 12 for the processing (Process 1 to Process 4) for making the robot execute tracking with weaving operation. And each will be described. Process 1 (Figure 9) and Process 2

(Fig. 10) is repeatedly executed in the sensor control board 20 (see Figs. 2 and 4) with a task of a predetermined cycle. On the other hand, the processing 3 (FIG. 11) and the processing 4 (FIG. 12) are repeatedly executed in the robot control board 30 (see FIGS. 2 and 5) by the task having the same predetermined cycle.

Hereinafter, each process will be described. The “sensor current position data” referred to below is the position in the sensor coordinate system calculated from the robot's current position data at a certain point in time.

(Including posture).

[Process 1]

The sensor control board 20 waits for the time stamped sensor current position data to be sent from the robot control board 30 (step S101). The sensor current position data with the time stamp is calculated periodically in the robot control board 30 as described later. Lopot system When the sensor current position data with a time stamp is received from the control board 30, it is written to the buffer 1 set in the memory 25 (see FIG. 4) (step S102).

Thereafter, it is determined whether or not a system processing end command has been sent from the robot control board 30 (step S103), and if no end command has been sent yet, the process returns to step S101. Prepare for the output of the current sensor position data with the next time stamp. On the other hand, if a system processing end command is sent from the robot control board 30, this processing (processing 1) is ended.

Returning from step S103 (judgment: N) to step S101, the buffer 1 stores sensor current position data with a time stamp each time the processing is repeated. The buffer 1 stores data in a ring memory format, and newer data is preferentially stored within the capacity limit.

[Process 2]

The sensor control board 20 waits for a sensing start command to be sent from the robot control board 30 (step S201). When a sensing start command is received from the robot control board 30, the laser sensor 2 scans the laser beam and the data detected by the scan is stored in the memory 25 (step S2). 0 2).

Then, the sensor current position data with the latest time stamp is read from buffer 1 and stored in memory 2 5 Is transferred to the buffer 2 set in (step S203). Next, the profile data (sensor data) of the joint is obtained based on the data accumulated in the memory 25 in step S202 (step S204).

The obtained profile data (sensor data) of the joint is transferred to the buffer 2 in step S203 using the data transferred to the buffer 2 on the robot coordinate system (that is, on the objective spatial coordinate system). ) To the profile data (robot data) (step S205). Next, the position Q (Xq, yq, zq) of the welding line and the gap length g (X) are obtained based on the converted profile data (mouth data) of the joint ( Step S206). In the example shown in Fig. 4, the position Q of the weld line is calculated, for example, as the midpoint between the edges EA and EB of the joints A and B. The gap length g (X) is calculated as the distance between the edges EA and EB.

The position Q (Xq, yq, zq) of the welding line and the gap length g (x) determined in step S206 are time stamped (the same as that read in step S203). (Step S207) in association with the object (labeling) and in the buffer 3 (Step S207).

Next, it is determined whether or not a system processing end command has been sent from the robot control board 30 (step S208). If the end command has not been sent yet, step S200 is executed. Return to step 2 and execute the processing following step S202 again. Run. On the other hand, if a system processing end command is sent from the robot control board 30, this processing (processing 2) is ended.

Returning to step S202 from step S208 (judgment: N), each time the processing is repeated, the buffer 3 stores the data of the position Q of the welding line labeled by the time stamp in the buffer 3. The gap length g (X) is accumulated. The buffer 3 also stores data in a ring memory format, and stores newer data preferentially within the capacity limit.

[Process 3]

This process is started by the mouth pot control board 30 outputting a sensing start command to the sensor control board 20 (step S301).

Then, the current position data of the robot is converted into the current position data of the sensor (step S302). As is well known, this transformation can be performed using homogeneous transformation matrix data representing the relative position / posture relationship between the sensor coordinate system and the robot coordinate system. Such homogeneous transformation matrix data is acquired by sensor calibration and stored in the non-volatile memory 34 of the robot control board 30.

The converted sensor current position data is appended to the time stamp and transferred to the sensor control board 20 (step S303).

Then, it is determined whether or not a system end command has been output. If not, step S302 is executed. Return to and execute the process in the next cycle. On the other hand, if the system end command is output, this process (process 3) is ended.

Every time the process returns from step S304 (judgment: N) to step S302, and is repeated, the sensor current position data labeled by the time stamp is output to the sensor control port 20 one after another. Is done.

[Process 4]

This processing is started when the CPU 31 of the robot control board 30 reads the first one block of the operation program from the non-volatile memory 34 (step S401). Then, a timer (weaving timer) for controlling the phase of the weaving operation is turned on (step S402). The initial value of the weaving timer is zero (t = 0).

Then, the next interpolation point position r on the teaching path is calculated based on the trajectory plan established according to the known processing (step S403). The interpolation point position r is a point that becomes the robot's movement target position if both the correction for tracking and the weaving amount are 0.

Next, data of the position Q of the welding line and the gap length g (X) are read from the buffer 3 (step S404). However, it is not the latest data that is generally read out, but rather at the time that goes back by the time corresponding to the distance between the scanning position of the laser sensor and the tool tip 4 (see Fig. 1). Is the data labeled with a close time stamp. Scanning position of laser sensor and tool tip 4 (Figure

Assuming that the distance to d) is constant (fixed value) and the command speed (specified in the program) is V, the time to go back is t v = d Z v. Therefore, in step S404, the time t V (= d no V) is traced back from the time when the data of the position Q of the welding line and the data of the gap length g (X) are read from the buffer 3 in the step S404. The data of the position Q and the gap length g (X) labeled with the timestamp closest to are selected.

Then, the position Q of the welding line and the position r of the interpolation point are compared, and a correction amount Atr for tracking is calculated (step S405). Further, a weaving condition suitable for the gap length g (X) read from the buffer 3 in step S404 is determined based on the setting data (see Fig. 7) of the weaving condition selection criterion (see FIG. 7). 4 6).

Then, it is determined whether or not the determined weaving conditions are to change the current weaving conditions (step S407). Note that the initially set driving condition is condition 1 (no weaving). Therefore, if it is determined that the current weaving conditions do not need to be changed (step S407, determination N), the correction A tr for tracking and the weaving amount w (without changing the current ゥ weaving conditions) The position (r + Atr + w (t)) is calculated by applying the shift obtained by adding t) to the corrected position r on the teaching route. The weaving amount w (t) is calculated as It is calculated from the setting data (see Fig. 6) corresponding to the bing condition number and the current value t of the ゥ eving timer.

On the other hand, if it is determined that the current weaving conditions should be changed (step S407, determination Y), it is determined whether the weaving amount calculated in the previous calculation processing cycle is smaller than the reference value ε. (Step S409). The reference value ε is a parameter to prevent the robot trajectory (the trajectory at the tip of the torch) from suddenly changing (flighting discontinuously) due to the switching of weaving conditions. O mm).

As a result of the determination in step S409, if the previous weaving amount is equal to or larger than the reference value ε, there is a possibility that the robot trajectory (trajectory at the tip of the torch) may be suddenly changed by switching the weaving conditions. And the weaving conditions are not changed (step S408).

On the other hand, if the state where the weaving amount is small (below ε) has arrived, the weaving condition can be changed (step S408, judgment Υ), and the weaving phase is changed according to the switching of the weaving condition. In order to adjust the time, reset the weaving timer, that is, set t = 0 (step S410). Then, the shift (the sum of the correction amount tr for tracking and the weaving amount w (t)) applied to the correction position r on the teaching path is calculated (that is, r + △ tr + w (t)). (Step S411). The weaving amount w (t) is calculated in step S4. It is calculated from the setting data (see Fig. 6) corresponding to the weaving condition number determined in 06 and the current value of the weaving timer (reset in step S410).

When the position r + Atr + w (t) is obtained in step S408 or step S411, the position is set as the movement target point, and the movement command for each axis of the robot is obtained by a well-known calculation. Is created and passed to the servo of each axis (step S4 12).

Then, it is determined whether or not the movement processing for one block read in step S401 has been completed, and if not completed, the processing in step S403 and the subsequent steps is performed again. On the other hand, when the movement processing for one block is completed, it is further determined whether the movement processing for all blocks has been completed (step S414). When completed, end this process. On the other hand, if the movement processing for all the blocks has not been completed yet, the flow returns to step 401 to read the next one block of the operation program, and thereafter executes the same processing.

By the above-described processes 1 to 4, real-time tracking along the welding line and weaving are performed in parallel. Then, the weaving conditions are automatically adapted to changes in the gap length. In addition, when the weaving conditions are switched, a smooth movement of the welding point can be guaranteed.

In this embodiment, welding of a butt joint is taken as an example. However, the present invention is not limited to the case where the gap length is expected to vary. It is clear that the present invention can be generally applied to a welding robot system for welding a joint in a normal state (for example, a corner welding joint having a gap).

According to the present invention, the position of the welding line and the gap length can be accurately determined even during the weaving operation, so that the weaving conditions can be switched in accordance with the gap length while using the weaving for accurate tracking. . Also, it is possible to prevent a sudden change in the trajectory (jump or disturbance) due to the switching of the weaving conditions. Therefore, by using the welding robot shim of the present invention, it is possible to improve the quality room for welding a joint having a variable gap length.

Claims

The scope of the claims

A robot to which a welding torch is attached, a laser sensor having a main body attached to the robot so as to sense a region in front of the welding torch in a direction in which the robot travels, and a control means for the robot. Welding robot system with

The robot control means includes means for outputting, as robot data, position data of the laser sensor labeled with a time stamp to the laser sensor, and the laser sensor includes: A means for detecting a profile of the joint to be welded based on the position data of the first sensor and the sensing result, and calculating and storing a welding line position and a gap length based on the detected profile. Means for obtaining a correction amount for tracking a welding line based on the stored welding line position, and a control means for controlling the robot in accordance with the stored gap length. Means for determining the weaving amount in accordance with the weaving condition selected by the user, and the correction amount for tracking determined as described above. Means for determining a movement target position of the robot based on the determined amount of e-iving and the interpolation position on the teaching path, the welding robot system.

The control means for the mouth pot further includes a weaving condition selected according to the detected gap length. The welding robot system according to claim 1, further comprising: means for determining a weaving amount so that the position of the robot does not suddenly change when the robot is replaced.