JP2022047287A - Robot controller - Google Patents

Abstract

To provide a robot controller which enables reduction of welding failure caused by a welding position being dislocated even if a welding wire is bent.SOLUTION: A robot controller 100 according to the invention includes a search part 110, a bent amount estimation part 130, and a control unit 140. The search part 110 moves a welding torch 21 to place a welding wire, which protrudes from a tip of the welding torch 21, in contact with a workpiece W10, a welding object, to detect a contact position. The bent amount estimation part 130 estimates a bent amount of the welding wire bent by the contact. The control unit 140 controls the welding torch 21 so that the welding wire is located in a welding position in the workpiece W10 according to the detected contact position and the estimated bent amount of the welding wire.SELECTED DRAWING: Figure 2

Description

The present invention relates to a robot control device.

A welding robot that performs arc welding generally moves a welding torch along a welding line position preset by teaching to weld a workpiece to be joined. However, in this teaching operation, since the welding wire is operated in a state where the welding wire protrudes from the tip of the welding torch, the welding wire may be bent when the welding wire comes into contact with the work. If the welding wire is bent, the welding wire cannot be positioned at an appropriate welding position when performing welding, and as a result, welding defects may occur.

A method for controlling an industrial robot that prevents the welding wire from bending due to contact with the welding object or the like when the welding wire and the welding object or the like come into contact with each other during teaching has been proposed (for example, a patent). See Document 1).

Specifically, in the control method of the industrial robot disclosed in Patent Document 1, when the contact between the welding wire protruding from the welding torch and the work is detected, the welding wire is fed in the direction away from the work. By performing reverse feeding, the contact between the weld wire and the work is released. Then, after moving the welding torch in the direction away from the work, the welding wire is forward-fed by the same length as the back-fed length.

Japanese Unexamined Patent Publication No. 2014-223633

However, in the control method of the industrial robot disclosed in Patent Document 1, depending on the posture, state, speed, etc. at which the welding wire contacts the work, the welding wire is fed back to release the contact between the welding wire and the work. Even so, it may not be possible to prevent the welding wire from bending.

Therefore, an object of the present invention is to provide a robot control device that reduces welding defects due to displacement of the welding position even if the welding wire is bent.

The robot control device according to one aspect of the present invention has a search unit that detects a contact position by moving a welding torch and bringing a welding wire protruding from the tip of the welding torch into contact with a work to be welded. A bending amount estimation unit that estimates the bending amount of the welding wire due to contact, and a control that controls the welding torch so that the welding wire is positioned at the welding position in the work according to the detected contact position and the estimated bending amount of the welding wire. It has a part.

According to this aspect, the search unit detects the contact position between the welding wire and the work, and the bending amount estimation unit estimates the bending amount of the welding wire due to the contact. Then, the control unit controls the welding torch so that the welding wire is positioned at the welding position in the work according to the detected contact position and the estimated bending amount of the welding wire, so that the welding is performed by the contact between the welding wire and the work. Even if the wire is bent, it is possible to reduce welding defects due to the displacement of the welding position.

In the above aspect, the control unit may control at least one of the position, the posture and the direction of the welding torch.

According to this aspect, the control unit can position the welding wire protruding from the tip of the welding torch at an appropriate welding position on the work by controlling at least one of the position, the posture and the direction of the welding torch. ..

In the above aspect, the bending amount estimation unit may estimate the bending amount of the welding wire based on the search speed at which the search unit moves the welding torch.

According to this aspect, the bending amount of the welded wire tends to be larger when the search speed is high than when the search speed is low. Therefore, the bending amount estimation unit accurately determines the welding wire based on the search speed. Estimate the amount of songs. In this way, in order to accurately estimate the bending amount of the welding wire, the welding wire can be accurately positioned at an appropriate welding position on the work.

In the above embodiment, the bending amount estimation unit estimates the bending amount of the welding wire based on at least one of torch information regarding the welding torch, wire information regarding the welding wire, environmental information regarding the surrounding environment, and robot information regarding the welding robot. May be good.

According to this aspect, in addition to the search speed, torch information, wire information, environmental information, etc. also affect the bending amount of the welded wire. Therefore, the bending amount estimation unit considers these information and welds more accurately. Estimate the amount of wire bending. In this way, in order to estimate the bending amount of the welding wire more accurately, the welding wire can be positioned more accurately at an appropriate welding position on the work.

In the above embodiment, the bending amount estimation unit is machine-learned using the search speed at which the search unit moves the welding torch and the bending amount of the welding wire generated when the welding wire comes into contact with the work at the search speed as teacher data. The predictive model may be used to estimate the amount of bending of the weld wire due to contact.

According to this aspect, the search speed and the bending amount of the welded wire are accumulated as teacher data, and the bending amount estimation unit estimates the bending amount of the welded wire due to contact using a prediction model optimized by machine learning. do. In this way, when estimating the bending amount of the welded wire, the prediction model optimized by machine learning is used, so that the optimum estimation can be performed at that time.

In the above embodiment, the teacher data may further include at least one of torch information regarding the welding torch, wire information regarding the welding wire, environmental information regarding the surrounding environment, and robot information regarding the welding robot.

According to this aspect, in addition to the search speed, torch information, wire information, environmental information, and the like also affect the bending amount of the welded wire. Therefore, if these information are used as teacher data, the accuracy of machine learning is further improved. In this way, when estimating the bending amount of the welded wire, a prediction model optimized by machine learning with improved accuracy is used, so that the optimum estimation can be performed at that time.

According to the present invention, it is possible to provide a robot control device that reduces welding defects due to a shift in the welding position even if the welding wire is bent.

It is a system schematic diagram which shows the welding robot system 10 which concerns on one Embodiment of this invention. It is a functional block diagram which shows each function of the robot control apparatus 100 which concerns on one Embodiment of this invention. It is a figure which shows the mode that the welding torch is controlled so that the welding wire is positioned at the welding position in the work according to the bending amount of the welding wire. It is a flowchart which shows the robot control method 400 executed by the robot control device 100 which concerns on one Embodiment of this invention. It is a figure which shows the relationship between the search speed which moves a welding torch and the bending amount of a welding wire. It is a flowchart which shows an example of the data storage method 600 for machine learning about the bending amount of the welding wire by a search speed. It is a figure which shows the state of moving a welding torch at the time of accumulating data for machine learning about the bending amount of a welding wire by a search speed. It is a figure which shows an example of the information used as the teacher data of machine learning.

Hereinafter, an embodiment of the present invention will be specifically described with reference to the accompanying drawings. It should be noted that the embodiments described below are merely specific examples for carrying out the present invention, and do not limit the interpretation of the present invention. Further, in order to facilitate understanding of the description, the same components are designated by the same reference numerals as possible in each drawing, and duplicate description is omitted.

[Overview of welding robot system]
FIG. 1 is a system schematic diagram showing a welding robot system 10 according to an embodiment of the present invention. In FIG. 1, the welding robot system 10 includes a welding robot 20, a teach pendant 30, a robot control device 100, and a power supply 40.

The welding robot 20 is connected to the robot control device 100 via a cable, and performs arc welding based on an operation command from the robot control device 100. The welding robot 20 is provided with a welding torch 21 at the tip of the arm, and by feeding a welding wire from the tip of the welding torch 21 and generating an arc with a metal material (work) to be welded. Perform arc welding.

The welding torch 21 is connected to the power supply 40 via a cable, and receives a welding voltage and a welding current supplied to the welding wire. In arc welding, when the welding wire is momentarily brought into contact with the metal material and energized, an arc discharge is generated between the welding wire and the metal material, and the heat of the generated arc melts the welding wire and the metal material. By doing so, welding is performed.

The teach pendant 30 receives input from a worker who performs welding work about welding-related teaching information of the welding robot 20. The operator inputs the optimum welding-related teaching information using the teach pendant 30 while checking the state of the arc.

Here, the welding-related teaching information is information about welding performed by the welding robot 20, and includes teaching information and welding conditions for teaching the operation of the welding robot 20. The teaching information of the welding robot 20 includes information on the operation of the arm of the welding robot 20, information on the position and posture of the welding robot 20, information on the protrusion length of the welding wire supplied from the tip of the welding torch 21, and the like. Will be welded. Further, the welding conditions include the welding voltage applied to the welding wire, the value of the welding current flowing through the welding wire, the welding speed indicating the moving speed of the welding torch 21 in the direction of the welding line during welding, and the like.

The robot control device 100 is a device that controls the welding robot 20. The robot control device 100 is connected to the teach pendant 30, and can acquire welding-related teaching information input to the teach pendant 30. The robot control device 100 controls the welding robot 20 and the power supply 40 based on the welding-related teaching information.

The power supply 40 is connected to the welding robot 20 via a cable, and supplies a welding voltage and a welding current to the welding torch 21 in the welding robot 20 based on a command from the robot control device 100.

Although the teach pendant 30 is connected to the robot control device 100 via a cable in FIG. 1, it may be connected wirelessly. That is, the teach pendant 30 and the robot control device 100 may include a communication unit that performs wireless communication. By wirelessly connecting the robot control device 100 and the teach pendant 30, the operator can move freely without being bothered by the existence of the cable or being restricted by the length of the cable. While it is possible to input welding-related teaching information.

[Robot control device configuration]
In the welding robot system 10 described above, when the welding robot 20 performs arc welding on a work, the welding position on the work is important. The robot control device 100 confirms the position of the work by bringing the contact sensor into contact with the work by using the contact sensor, and detects the welding position in the work. Hereinafter, the robot control device 100 for detecting the welding position using the contact sensor and the method thereof will be described in detail.

FIG. 2 is a functional block diagram showing each function of the robot control device 100 according to the embodiment of the present invention. In FIG. 2, the welding robot 20 includes a contact sensor 22, and the robot control device 100 includes a search unit 110 including a contact detection unit 111 and a voltage application command unit 112, a search condition storage unit 120, and a bending amount estimation unit. It includes a 130 and a control unit 140.

As shown in FIG. 1, the robot control device 100 is connected to the welding robot 20, the teach pendant 30, and the power supply 40, and has many functions for performing various controls to perform processing. Is going. Here, the robot control device 100 is shown with respect to a configuration and a function of detecting a welding position by using a contact type sensor 22, but other configurations and functions may be provided.

The contact type sensor 22 in the welding robot 20 may typically use a wire touch sensor. In a state where a sensing voltage is applied between the welding wire protruding from the tip of the welding torch 21 in the welding robot 20 and the work to be welded, the welding torch 21 in the welding robot 20 is subjected to an operation command of the control unit 140. Move. When the welding wire protruding from the tip of the welding torch 21 comes into contact with the work, the voltage changes at that time, and the contact between the welding wire and the work is recognized by the change in the voltage. That is, the welding torch 21 and the welding wire protruding from the tip of the welding torch 21 bear one end of the contact type sensor 22.

The search unit 110 moves the welding torch 21 based on the search conditions stored in the search condition storage unit 120. The search unit 110 detects the contact position by moving the welding torch 21 (contact type sensor 22) and bringing the welding wire protruding from the tip of the welding torch 21 into contact with the work to be welded. Specifically, the voltage application command unit 112 commands the power supply 40 to apply a voltage for sensing between the welding wire and the work, and in a state where the voltage is applied, an operation command of the control unit 140. Moves the welding torch 21 in the welding robot 20. The contact detection unit 111 monitors the voltage and detects the contact between the welding wire and the work by the change of the voltage.

The search condition storage unit 120 stores search conditions when the search unit 110 moves the welding torch 21 to search for a welding position on the work. The search condition stored in the search condition storage unit 120 may be input and set by the operator via the teach pendant 30. The search condition includes, for example, a search speed at which the search unit 110 moves the welding torch 21.

The bending amount estimation unit 130 estimates the bending amount of the welded wire caused by the contact between the welded wire and the work. Specifically, as described above, the search unit 110 detects the contact position by moving the welding torch 21 and bringing the welding wire protruding from the tip of the welding torch 21 into contact with the work. The welding wire may be bent due to the impact caused by the torch. The bending amount estimation unit 130 estimates the bending amount of the welded wire as to how much the welded wire is bent by the contact. The details of the method of estimating the bending amount of the welded wire will be described later.

The control unit 140 sends and receives commands and data to the search unit 110 and the music amount estimation unit 130, controls the processing executed by the robot control device 100, and stores them in a storage unit (not shown) such as a memory. The welding robot 20 and the power supply 40 are controlled based on the welding-related teaching information provided. Here, the control unit 140 is a welding torch so that the welding wire is positioned at the welding position in the work according to the contact position detected by the search unit 110 and the bending amount of the welding wire estimated by the bending amount estimation unit 130. The control of 21 will be described in detail.

FIG. 3 is a diagram showing how the welding torch is controlled so that the welding wire is positioned at the welding position in the work according to the bending amount of the welding wire. As shown in FIG. 3A, the welding wire protruding from the tip of the welding torch 21 comes into contact with the work W10, so that the welding position Q10 is detected at the position P10 of the welding torch 21.

Due to the contact between the welding wire and the work W10, the welding wire is bent at the end of the search. Therefore, when the welding torch 21 is moved to the position P10 and the welding is actually performed, the welding wire protruding from the tip of the welding torch 21 is arranged at the position Q11 deviated from the welding position Q10 ( FIG. 3 (B).

Therefore, the control unit 140 adjusts, for example, at least one of the position, posture, and direction of the welding torch 21 so that the welding wire protruding from the tip of the welding torch 21 is appropriately arranged at the welding position Q10. The welding torch 21 is controlled to the position P11 (FIG. 3 (C)).

[Robot control method]
Next, the method of controlling the welding torch so that the bending amount of the welding wire is estimated and the welding wire is appropriately positioned at the welding position in the work after moving the welding torch to detect the welding position is specified. Will be explained in detail.

FIG. 4 is a flowchart showing a robot control method 400 executed by the robot control device 100 according to the embodiment of the present invention. In FIG. 4, the robot control method 400 includes steps S401 to S405, and each step is executed by a processor included in the robot control device 100.

In step S401, the search unit 110 moves the welding torch 21 (contact type sensor 22) based on the search conditions stored in the search condition storage unit 120 to search for the work W10 to be welded.

In step S402, the contact detection unit 111 in the search unit 110 detects the contact between the contact sensor 22 (welding wire protruding from the tip of the welding torch 21) and the work W10. As a result, the search unit 110 grasps the welding position Q10 in the work W10 and the position P10 of the welding torch 21 (FIG. 3A).

Here, since the welding robot 20 usually welds not only one point or one place but also a continuous range or a plurality of places, the welding start position and the welding end position are detected and the welding line position is confirmed, respectively. Check the welding position at multiple locations.

In step S403, the control unit 140 determines whether all the welding positions have been searched and grasped. The search unit 110 repeats the processes of steps S401 and S402 until the search for all the welding positions is completed (No in step S403). When the search for all the welding positions is completed, the process proceeds to the process of step S404 (Yes in step S403).

When it is determined in step S403 that the search for all the welding positions is completed, the welding wire protruding from the tip of the welding torch 21 may come into contact with the work W10 multiple times, and may be bent. Is high (Fig. 3 (B)).

In step S404, the bending amount estimation unit 130 estimates the bending amount of the welded wire caused by the contact between the welded wire and the work W10. Here, the bending amount of the welding wire may be affected by the search speed at which the search unit 110 moves the welding torch 21 when the welding torch is moved to search for a welding position in the work W10.

FIG. 5 is a diagram showing the relationship between the search speed for moving the welding torch and the bending amount of the welding wire. As shown in FIG. 5, as the search speed increases, the bending amount of the weld wire also increases. As the search speed increases, the impact when the weld wire comes into contact with the work W10 also increases, and the bending amount of the weld wire also increases. For example, if the search speed is about 100 [cm / min], the bending amount of the weld wire is less than 0.1 [mm], there is almost no bending, and the search speed is about 400 [cm / min]. If so, the bending amount of the welding wire is about 0.4 [mm].

The search unit 110 moves the welding torch 21 to search the work W10 based on the search conditions stored in the search condition storage unit 120 (step S401), but the search unit 110 searches for the work W10 according to the search speed included in the search conditions. , The bending amount estimation unit 130 can estimate the bending amount of the welded wire (step S404).

In step S405, the control unit 140 determines the welding position Q10 in the work W10 according to the contact position detected by the search unit 110 in step S402 and the bending amount of the welding wire estimated by the bending amount estimation unit 130 in step S404. The welding torch 21 is controlled so that the welding wire is located at (FIG. 3 (C)).

For example, when the search unit 110 moves the welding torch 21 at a search speed of 300 [cm / min] to search for a work (step S401), the bending amount estimation unit 130 performs welding with the search speed shown in FIG. The bending amount of the welded wire is estimated to be 0.25 [mm] based on the relationship with the bending amount of the wire (step S404). Then, in step S405, the control unit 140 assumes that the welding wire is bent by 0.25 [mm], and arranges the welding wire protruding from the tip of the welding torch 21 at an appropriate welding position Q10. In addition, the welding torch 21 is controlled to the position P11 (FIG. 3 (C)).

Specifically, as a method of controlling the welding torch 21 to the position P11 so that the welding wire protruding from the tip of the welding torch 21 is arranged at an appropriate welding position Q10, the position of the welding torch 21 is adjusted. Further, the posture may be adjusted by changing the angle, or the direction may be adjusted by changing the direction. In addition, if the welding wire protruding from the tip of the welding torch 21 is placed more accurately at the appropriate welding position Q10, the arm and rotation axis of the welding robot 20 can be adjusted, or from the tip of the welding torch 21. The protruding length of the protruding welding wire may be adjusted.

As described above, if the teaching information for teaching the operation of the welding robot 20 is generated so as to control the welding torch 21 in consideration of the bending amount of the welding wire estimated by the bending amount estimation unit 130, the welding wire and the work are generated. Even if the welding wire is bent due to contact with W10, it is possible to reduce welding defects due to the displacement of the welding position.

Further, in the present embodiment, in step S404, the bending amount estimation unit 130 welds according to the search speed based on the relationship between the search speed for moving the welding torch and the bending amount of the welding wire shown in FIG. The amount of bending of the wire was estimated, but it is not limited to this. For example, the bending amount of the welding wire with respect to the search speed may be calculated by using a calculation formula for the relationship between the search speed and the bending amount of the welding wire, or by using a look-up table stored in advance.

Further, the bending amount estimation unit 130 estimates the bending amount of the welded wire according to the search speed, but the present invention is not limited to this. For example, the bending amount of the welding wire may be estimated more accurately in consideration of the torch information regarding the welding torch 21, the wire information regarding the welding wire, the environmental information regarding the surrounding environment, and other robot information regarding the welding robot. ..

The torch information includes teaching information that teaches the operation of the welding robot 20, for example, the position, posture, direction, search and number of contacts of the welding torch 21 when searching for the work W10 and when contacting the work W10. Information such as search and contact order, search and contact time, etc. is included. The wire information includes, for example, information such as the protrusion length, material, diameter, and state of the welded wire. The environmental information includes, for example, information such as climate, time zone, place, temperature, and humidity. The robot information includes, for example, performance information such as the type and size of the welding robot, arm length, mass, and moment of inertia, and motion information such as each axis angle, angular velocity, and acceleration.

Such information as various factors (factors) for bending the welding wire is calculated in advance by adding weighting to the relationship with the bending amount of the welding wire according to the degree of bending of the welding wire and the degree of influence. It may be set and calculated using a lookup table.

Further, these pieces of information may be regarded as search conditions, may be stored in the search condition storage unit 120, or may be input and set by the operator via the teach pendant 30.

[Learning method for estimating the bending amount of weld wire]
Further, the bending amount of the welded wire with respect to the above-mentioned search speed, torch information, wire information, environmental information, robot information and the like can be estimated more accurately by accumulating data. The so-called machine learning will be described below with an example.

FIG. 6 is a flowchart showing an example of a data storage method 600 for machine learning about the bending amount of the weld wire depending on the search speed, and FIG. 7 is a flowchart for machine learning about the bending amount of the weld wire depending on the search speed. It is a figure which shows the state of moving a welding torch when accumulating data. In FIG. 6, the data storage method 600 includes steps S601 to S607, and each step is executed by a processor included in the robot control device 100.

In step S601, the search unit 110 moves the welding torch 21 at a low speed to search the work W20. The low speed is a search speed at which the bending amount of the welding wire is small and the welding wire hardly bends due to the impact when the welding wire and the work W20 come into contact with each other. For example, it is shown in FIG. Based on the example, it may be about 100 [cm / min].

In step S602, the contact detection unit 111 in the search unit 110 detects the contact between the contact sensor 22 (welding wire protruding from the tip of the welding torch 21) and the work W20. At this time, the search unit 110 detects the position P20 of the welding torch 21 (FIG. 7A).

In step S603, the search unit 110 moves the welding torch 21 at high speed to search the work W20. The high speed is higher than the search speed in step S601, and is a search speed to which at least the weld wire is bent due to the impact when the weld wire and the work W20 come into contact with each other. From the viewpoint of accumulating data in machine learning, it is preferable to accumulate a large amount of data at various search speeds.

In step S604, the contact detection unit 111 in the search unit 110 detects the contact between the contact sensor 22 (welding wire protruding from the tip of the welding torch 21) and the work W20. At this time, the search unit 110 detects the position P21 of the welding torch 21 (FIG. 7 (B)). At this time, as shown in FIG. 7B, the welding wire is bent due to the impact caused by the contact between the welding wire and the work W20.

Here, in step S601, although the welding wire is not bent, the position P21 of the welding torch 21 detected in step S604 is deviated from the position P20 of the welding torch 21 detected in step S602. This is a detection error (P21-P20) by high-speed search caused by the timing and time lag of detecting contact with the work W20 when the welding torch 21 is moved at high speed.

In step S605, the search unit 110 again moves the welding torch 21 at a low speed to search the work W20.

In step S606, the contact detection unit 111 in the search unit 110 detects the contact between the contact sensor 22 (welding wire protruding from the tip of the welding torch 21) and the work W20. At this time, the search unit 110 detects the position P22 of the welding torch 21 (FIG. 7 (C)).

In step S607, the bending amount estimation unit 130 estimates the bending amount of the welding wire based on the position P20 of the welding torch 21 detected in step S602 and the position P22 of the welding torch 21 detected in step S606.

More specifically, in step S604, the welding wire is bent, and in step S606, the contact with the work W20 is detected by the welding wire in the bent state, and the position P22 of the welding torch 21 at that time is determined. It is being detected. In other words, the difference in the position of the welding torch 21 (P22) with respect to the position of the welding torch 21 when the contact between the welding wire and the work W20 is detected by searching at the same search speed (search condition) called low speed search. It can be seen that −P20) is the bending amount of the welding wire. The bending amount estimation unit 130 may consider the detection error (P21-P20) by the high-speed search generated in step S604 when estimating the bending amount of the welded wire, and considers the detection error. Thereby, the bending amount of the welded wire can be estimated more accurately.

By repeating the low-speed search, the high-speed search, and the low-speed search in this way, it is possible to estimate the bending amount of the weld wire with respect to how much bending is generated in the welding wire at the search speed used in the high-speed search. Can be done.

Further, if data on the bending amount of the welded wire is accumulated using various search speeds in step S603, the bending amount of the welded wire can be estimated more accurately by machine learning.

That is, the bending amount estimation unit 130 stores the search speed at which the search unit 110 moves the welding torch 21 and the bending amount of the welding wire generated when the welding wire comes into contact with the work at the search speed as teacher data. .. Then, based on the accumulated teacher data, the bending amount estimation unit 130 can estimate the bending amount of the welded wire due to the contact with the work by using the machine-learned prediction model. Since machine learning is improved in accuracy by accumulating teacher data and an optimized prediction model can be derived at that time, the bending amount estimation unit 130 can estimate the bending amount of the optimum weld wire. can.

The timing for executing the data storage method 600 and storing data for machine learning is, for example, welding every predetermined period such as every week or every month, every 10 times or every 100 times. It may be performed every predetermined number of times, every time a welding defect occurs, when the incidence of welding defects exceeds a predetermined value, or at any timing determined by the operator.

Further, although the search speed is used here as the teacher data for machine learning, the search speed is not limited to this, and for example, torch information, wire information, environment information, robot information, and the like may be used.

FIG. 8 is a diagram showing an example of information used as teacher data for machine learning. For example, if the welding torch 21 contacts the work many times or the contact time with the work is long, the bending amount of the welding wire becomes large, and further, the bending amount of the welding wire also depends on the search and the contact order. It is possible that it will affect. Further, the bending amount of the welding wire is affected by the protruding length, material, diameter, etc. of the welding wire, but the bending amount of the welding wire may also change depending on the temperature and humidity environment. .. In addition, depending on the type of welding robot and the length of the arm, the combination and compatibility with the welding wire may affect the bending amount of the welding wire. If more information about these factors and conditions can be accumulated as teacher data, an optimized prediction model can be derived by better machine learning, so that the music amount estimation unit 130 is optimal. The bending amount of the welded wire can be estimated.

As described above, according to the robot control device 100 and the robot control method 400 according to the embodiment of the present invention, the bending amount estimation unit 130 estimates the bending amount of the welded wire, and the control unit 140 bends the welded wire. Since the welding torch 21 is controlled so that the welding wire is positioned at an appropriate welding position Q10 in the work W10 according to the amount, even if the welding wire is bent due to the contact between the welding wire and the work W10, the welding position. Welding defects due to misalignment can be reduced.

Further, in general, if the search speed is increased, the bending amount of the welded wire is increased due to the impact due to the contact between the welded wire and the work W10, and as a result, the welding position is displaced and welding defects are likely to occur. There is a concern. However, since the bending amount estimation unit 130 in the robot control device 100 according to the embodiment of the present invention estimates the bending amount of the welding wire based on the search speed at which the search unit 110 moves the welding torch 21, the search speed. Even when is large, the bending amount of the welded wire is estimated accurately. As a result, welding defects due to the displacement of the welding position can be reduced, and the search speed can be increased, so that the tact time can be shortened, which leads to improvement of productivity and cost reduction.

Further, in general, if the search speed is increased, the detection error caused by the timing for detecting the contact between the welding wire and the work W10, the time lag, etc. becomes large, and as a result, the welding position shifts and welding defects occur. There is a concern that it will be easier. However, as described with reference to FIGS. 6 and 7, the detection error can be taken into consideration in the process of machine learning by repeating the low-speed search, the high-speed search, and the low-speed search, and the song amount estimation unit 130 can be used. The bending amount of the welded wire can be estimated accurately. As a result, welding defects due to the displacement of the welding position can be reduced, and the search speed can be increased, so that the tact time can be shortened, which leads to improvement of productivity and cost reduction.

Further, if the bending amount of the welded wire is estimated by machine learning, the accuracy is improved by accumulating the teacher data, and an optimized prediction model can be derived at that time. Therefore, the bending amount estimation unit 130 Can estimate the optimum amount of bending of the welded wire. Since the accuracy of the bending amount of the welding wire estimated by the bending amount estimation unit 130 is improved, it is possible to further reduce welding defects due to the deviation of the welding position.

Further, according to the present invention, since the bending amount of the welding wire can be accurately estimated by the bending amount estimation unit 130, the welding torch 21 when the welding wire protruding from the tip of the welding torch 21 comes into contact with the work. The bent weld wire can be straightened by considering the position, attitude and direction, search speed, and other materials of the weld wire.

The embodiments described above are for facilitating the understanding of the present invention, and are not for limiting the interpretation of the present invention. Each element included in the embodiment and its arrangement, material, condition, shape, size, and the like are not limited to those exemplified, and can be appropriately changed. Further, it is possible to partially replace or combine the configurations shown in different embodiments.

10 ... Welding robot system, 20 ... Welding robot, 21 ... Welding torch, 22 ... Contact sensor, 30 ... Teach pendant, 40 ... Power supply, 100 ... Robot control device, 110 ... Search unit, 111 ... Contact detection unit, 112 ... Voltage application command unit, 120 ... Search condition storage unit, 130 ... Music amount estimation unit, 140 ... Control unit, 400 ... Robot control method, 600 ... Data storage method, S401 to S405 ... Robot control method 400 steps, S601 to S607 ... Each step of the data storage method 600, W10, W20 ... Work, P10, P11, P20, P21, P22, Q10, Q11 ... Position

Claims (6)

A search unit that detects the contact position by moving the welding torch and bringing the welding wire protruding from the tip of the welding torch into contact with the work to be welded.
A bending amount estimation unit that estimates the bending amount of the welding wire due to the contact,
A control unit that controls the welding torch so that the welding wire is positioned at the welding position in the work according to the detected contact position and the estimated bending amount of the welding wire.
A robot control device. The control unit controls at least one of the position, orientation and direction of the welding torch.
The robot control device according to claim 1. The bending amount estimation unit estimates the bending amount of the welding wire based on the search speed at which the search unit moves the welding torch.
The robot control device according to claim 1 or 2. The bending amount estimation unit estimates the bending amount of the welding wire based on at least one of torch information regarding the welding torch, wire information regarding the welding wire, environmental information regarding the surrounding environment, and robot information regarding the welding robot.
The robot control device according to any one of claims 1 to 3. The bending amount estimation unit is machine-learned using the search speed at which the search unit moves the welding torch and the bending amount of the welding wire generated when the welding wire comes into contact with the work at the search speed as teacher data. Using the prediction model, the amount of bending of the welded wire due to the contact is estimated.
The robot control device according to any one of claims 1 to 4. The teacher data further includes at least one of torch information about the welding torch, wire information about the welding wire, environmental information about the surrounding environment, and robot information about the welding robot.
The robot control device according to claim 5.

Images