JP2019089201A - Teaching data creation device, method for controlling teaching data creation device, and robot system - Google Patents

Abstract

To create appropriate off-line teaching data even when there is a difference in shape between a structure in a virtual robot system and a structure in an actual robot system.SOLUTION: A teaching data creation device comprises: acquisition means that acquires an actually measured three-dimensional model representing a shape of a peripheral structure near an actual robot; and compensation means that, when a moving path of the robot according to teaching data showing the moving path of the robot can be shortened or interference occurs between the peripheral structure and the robot moving according to the teaching data in the acquired three-dimensional model, compensates the teaching data.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a teaching data creation device that teaches the operation of a robot off-line, a control method of the teaching data creation device, and a robot system.

  The off-line teaching device constructs a virtual robot system configured of a robot and a peripheral structure of the robot in a virtual space, creates an operation program of the robot, and teaches the operation of the robot off-line. There is usually an error between the virtual robot system and the real robot system, and when the robot operation program created by the off-line teaching device is supplied to the real robot system, between the robot and the surrounding structure Interference etc. may occur.

  Then, there exists a technique of patent document 1 as an off-line teaching device which considered the said difference | error. This technology detects an error in the placement position relative to the virtual robot system by measuring the placement position of the components of the actual robot system using a two-dimensional vision sensor, a three-dimensional vision sensor, a distance sensor, etc. The teaching point coordinates are shifted by the error to avoid the interference between the robot and the surrounding structure.

Japanese Patent Application Publication No. 2003-150219

By the way, in the off-line teaching device, a three-dimensional model of a robot or a peripheral structure is prepared and arranged in a virtual space to construct a virtual robot system. Then, using these three-dimensional models, the operation of the robot and the coordinates of the teaching point are set to perform teaching work.
When constructing a virtual robot system, a three-dimensional model prepared by a robot manufacturer can be used as a robot, but a three-dimensional model is often not prepared as a peripheral structure. In this case, a three-dimensional model of the peripheral structure is substituted by a simplified shape model whose external dimensions are substantially the same. Thus, in the virtual robot system and the real machine robot system, a difference may occur in the model shape of the structure.

However, the technology described in the above-mentioned Patent Document 1 only corrects an error in the arrangement position of the virtual robot system and the real machine robot system, and can not correct the error in the model shape of the structure.
Therefore, in the technology described in Patent Document 1 described above, there is a possibility that interference between the robot and the component may occur in the actual robot system due to the difference in the model shape. Also, if there is a difference in the model shape, the shortest path to the working point may be changed, but the teaching point coordinate shift for the error related to the arrangement position as in the technology described in Patent Document 1 No correction can be made to make the moving path the shortest path.
Therefore, it is an object of the present invention to create appropriate off-line teaching data even when there is a difference in shape between a virtual robot system component and an actual robot system component.

  In order to solve the above-mentioned subject, one mode of the teaching data creation device concerning the present invention is an acquisition means which acquires a measurement three-dimensional model which shows shape of a peripheral structure of a robot of a real machine, and the acquired said measurement 3D In the model, if the movement path of the robot can be shortened by teaching data indicating the movement path of the robot, or if the movement of the robot by the teaching data causes interference with the peripheral structure, the teaching data is corrected And correction means for

  According to the present invention, it is possible to create appropriate off-line teaching data even when there is a difference in shape between a virtual robot system component and a real robot system component.

It is a block diagram which shows an example of the robot system of this embodiment. It is a figure showing the example of hardware constitutions of an offline teaching device. It is a flowchart which shows an example of a prior examination process procedure. It is a figure explaining the detail of a system construction part. It is a figure which shows the detail of screen component data. It is a figure which shows an example of a teaching point and a movement path | route. It is a flow chart which shows an example of a course amendment processing procedure. It is a figure for demonstrating the measuring method of a component. It is an example of the difference pattern of the shape of a three-dimensional model. It is a figure for demonstrating the comparison method of a model. It is a figure for demonstrating a path correction determination method (the shortest path correction). It is an example of shortest path correction determination. It is an example of shortest path correction determination. It is an example of shortest path correction determination. It is a figure which shows an example of the shape of an end effector. It is an example determined to perform the shortest path correction. It is an example determined to perform the shortest path correction. It is a figure for demonstrating a path correction determination method (interference avoidance path correction). This is an example in which it is determined that the interference avoidance path correction is performed. It is a figure for demonstrating the shortest path correction method. It is a figure which shows the method of setting the start point and end point of shortest path correction. It is a figure explaining a route search method. It is a figure explaining a route search method. It is a figure explaining a route search method. It is a figure explaining a route search method. It is a figure which shows the shortest path correction result. It is a figure which shows the interference avoidance path | route correction result.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
The embodiment described below is an example as a realization means of the present invention, and it should be appropriately corrected or changed according to the configuration of the apparatus to which the present invention is applied and various conditions. It is not limited to the embodiment of the invention.
(First embodiment)
FIG. 1 is a view showing a configuration example of a robot system 100 provided with an off-line teaching device in the present embodiment.
The robot system 100 includes a robot 10 and an off-line teaching device 20.
The robot 10 is, for example, an articulated robot, and a sensor unit 11 is attached to a tip end (hand tip) of an arm. The sensor unit 11 measures an object near the tip of the robot 10 and outputs the measurement result to the off-line teaching device 20. Here, the sensor unit 11 can be configured by, for example, a visual sensor or a distance sensor.

The sensor unit 11 does not have to be mounted on the robot 10. For example, the sensor unit 11 may be mounted on another work machine, or may be fixedly arranged above the predetermined imaging target space.
The robot 10 further includes a position and orientation change mechanism 12 capable of changing the position and orientation of the tip of the robot 10. The position and orientation change mechanism 12 changes the angle of each joint of the robot 10 to change the position and orientation of the tip of the robot 10. Here, the position and orientation change mechanism 12 may be driven by an electric motor, or may be driven by an actuator operated by fluid pressure such as oil pressure or air pressure. The position / posture change mechanism 12 is driven based on teaching data indicating the movement path and the like of the robot 10 created by the off-line teaching device 20.

Furthermore, an end effector 13 is attached to the tip of the robot 10. The end effector 13 is a tool for realizing the work according to the type of work of the robot 10, and is, for example, a robot hand or the like. As the robot hand, a hand having a chuck mechanism that can be driven by a motor and capable of gripping an object, a hand using a suction pad that sucks an object by air pressure, or the like can be used. The end effector 13 is detachably attached to the arm, and can be replaced according to the type of work.
The robot 10 is not limited to the articulated robot, and may be a movable machine capable of numerical control (NC).

The off-line teaching device 20 creates teaching data for the robot (for example, a movement path of the robot) in a virtual space, and performs off-line teaching (off-line teaching) for supplying this to the actual robot. Specifically, the off-line teaching device 20 arranges a virtual robot, a tool (end effector) attached to the virtual robot, a work as a work object, and a three-dimensional model such as a peripheral structure on the virtual space. Then, a virtual robot system is constructed, and preliminary examination processing for creating teaching data in the virtual robot system is executed.
Further, the off-line teaching device 20 executes path correction processing for correcting the teaching data created by the preliminary examination processing in accordance with the shape of the three-dimensional model of the component acquired after construction of the real machine robot system.

Hereafter, each part which comprises the off-line teaching device 20 is demonstrated in detail.
The off-line teaching device 20 is configured of, for example, a personal computer (PC), and as shown in FIG. 1, the system construction unit 201, a robot path creation unit 202, an operation check unit 203, and a measurement / model creation unit 204. , A shape difference determination unit 205, a path correction unit 206, an external IF unit 207, and an input / output control unit 208.
The system construction unit 201 constructs a virtual robot system on the virtual space.
The robot path generation unit 202 is a virtual robot system constructed by the system construction unit 201, and generates a movement path of the robot as teaching data.
The operation check unit 203 simulates the robot path created by the robot path creation unit 202 with animation.

In addition, the measurement / model creation unit 204 controls the sensor unit 11 to measure information indicating the shape of the component of the real machine robot system, and creates a three-dimensional model (measured three-dimensional model) of the measured component. .
The shape difference determination unit 205 determines whether there is a difference in shape between the measured three-dimensional model created by the measurement / model creation unit 204 and the existing three-dimensional model (virtual three-dimensional model) existing in the virtual space corresponding thereto It is determined whether or not it is not, and information on the shape error of the existing three-dimensional model is acquired. When the shape difference determination unit 205 determines that there is the above difference, whether the difference affects the path of the robot created in the preliminary examination process, that is, corrects the path of the robot created in the preliminary examination process Determine if you need to.

The path correction unit 206 corrects the path of the robot created in the preliminary examination process according to the determination result by the shape difference determination unit 205. Here, when it is determined by the shape difference determination unit 205 that the difference in shape affects the robot path (the robot path needs to be corrected), the path of the robot is corrected.
The external IF unit 207 transfers the teaching data generated by the robot path generation unit 202 and the path correction unit 206 to the robot 10. In addition, the external IF unit 207 receives the measurement result of the sensor unit 11 and transmits this to the measurement / model creation unit 204.
The input / output control unit 208 inputs an operation performed by the user via a monitor screen using a pointing device such as a keyboard or a mouse, or displays a simulation result of the robot on the monitor screen.

(Hardware configuration of offline teaching device 20)
FIG. 2 is an example of a hardware configuration of the off-line teaching device 20. As shown in FIG.
The off-line teaching device 20 includes a CPU 21, a ROM 22, a RAM 23, an external memory 24, an input unit 25, a display unit 26, a communication I / F 27, and a system bus 28.
The CPU 21 centrally controls the operation of the off-line teaching device 20, and controls each component (22 to 27) via the system bus.
The ROM 22 is a non-volatile memory that stores control programs and the like necessary for the CPU 21 to execute processing. The program may be stored in the external memory 24 or a removable storage medium (not shown).
The RAM 23 functions as a main memory, a work area, and the like of the CPU 21. That is, the CPU 21 loads a necessary program and the like from the ROM 22 to the RAM 23 at the time of execution of processing, and realizes various functional operations by executing the program and the like.

The external memory 24 stores, for example, various data, various information, and the like necessary when the CPU 21 performs a process using a program. The external memory 24 stores, for example, various data, various information, and the like obtained by the CPU 21 performing processing using a program or the like.
The input unit 25 is configured of, for example, a keyboard, a mouse, etc., and an operator can give an instruction to the off-line teaching device 20 via the input unit 25.
The display unit 26 is configured of a monitor such as a liquid crystal display (LCD).
The communication I / F 27 is an interface for communicating with an external device.
The system bus 28 communicably connects the CPU 21, the ROM 22, the RAM 23, the external memory 24, the input unit 25, the display unit 26, and the communication I / F 27.

The function of each part of the off-line teaching device 20 is realized by the CPU 21 executing a program stored in the ROM 22 or the external memory 24. The process executed by the off-line teaching device 20 will be specifically described below.
(Preliminary processing)
FIG. 3 is a flowchart showing the preliminary examination process procedure executed by the off-line teaching device 20. This preliminary examination process is a process for conducting preliminary examination by simulation on a virtual space before constructing a real robot system.
First, in step S1, the off-line teaching device 20 constructs a virtual robot system on a virtual space. The virtual robot system is constructed by the system construction unit 201 described above.
The system construction unit 201 imports a three-dimensional model of each component constituting the robot system, and arranges the three-dimensional model in a virtual space to construct a virtual robot system. At this time, as shown in FIG. 4, for a component for which a three-dimensional model is prepared in the external CAD device 210, the three-dimensional model is imported from the CAD device 210, and the three-dimensional model is prepared in the CAD device 210. For the components that are not included, a simple three-dimensional model created on the screen by the user is imported via the input / output control unit 208. The CAD device 210 may be mounted on the off-line teaching device 20.

  When importing a three-dimensional model, the system construction unit 201 stores the screen component data 211 in a memory or the like. The screen component data 211 is composed of the fields shown in FIG. That is, in each component, the name 211a of the three-dimensional model, the coordinates 211b at which the three-dimensional model is arranged, the information 211c on the shape of the three-dimensional model, the reference destination CAD 211d of the three-dimensional model, and the shape correction execution flag 211e It is associated.

Here, when a three-dimensional model does not exist in the CAD device 210 and a three-dimensional model created by the user is imported, nothing is described in the field of the reference destination CAD 211 d. The shape correction execution flag 211e is a flag indicating whether or not there is a possibility that a difference in shape with the actual machine has occurred, and corresponds to attribute information indicating the reliability of the model shape. In the field of the shape correction execution flag 211e, for example, a three-dimensional model does not exist in the CAD device 210, and the configuration created and imported by the user causes a shape difference with the real machine robot system. A flag indicating that there is a risk of being The shape correction execution flag 211e is used to determine whether or not the route correction is necessary in the route correction process described later.
Returning to FIG. 3, in step S2, after the off-line teaching device 20 constructs a virtual robot system in step S1 described above, the teaching point of the robot and the movement path of the robot obtained by linear interpolation between the teaching points are create. Here, the teaching point and the robot path are created so that the work tact becomes short and the robot path does not interfere with the peripheral devices.

Next, in step S3, simulation is performed on the virtual space based on the teaching points and the path created in step S2.
FIG. 6 illustrates an example of a robot path obtained by the simulation result. A point 401 is a teaching point, and a robot path 402 is a line connecting the teaching points in the order of teaching point a → b → c → d → e → f. Thus, the teaching point 401 and the robot path 402 are created so as not to interfere with the peripheral device 301.
In FIG. 3, the process of step S1 is the process executed by the system construction unit 201 as described above, the process of step S2 is the process executed by the robot path creation unit 202, and the process of step S3 is the operation check This is processing executed by the unit 203.

(Path correction processing)
Next, the path correction process performed by the off-line teaching device 20 will be specifically described.
FIG. 7 is a flowchart showing the path correction processing procedure. The path correction process is a process for correcting the robot path created in the above-described preliminary examination process as necessary, and is executed after the actual robot system is built after the preliminary examination process is performed.
First, in step S11, the off-line teaching device 20 measures the components of the real machine robot system by the sensor unit 11. Here, the component to be measured is determined based on the screen component data 211 created when constructing the virtual robot system.

  First, referring to the field of the shape correction execution flag 211e of the screen component data 211 shown in FIG. 5, the component which may cause a shape difference with the actual robot system is determined. Next, with reference to the field of the arrangement 211b of the screen component data 211 shown in FIG. 5, the arrangement position in the virtual space of the component that may cause the shape difference is acquired. As a result, the components to be measured by the sensor in the real robot system are determined.

In the case where the component is measured by the sensor unit 11, the component to be measured is measured from a plurality of directions as long as it is possible to create a three-dimensional model. For example, as shown in FIG. 8, the measurement direction is one of the X direction (the width direction of the component 310), the Y direction (the depth direction of the component 310), and the Z direction (the height direction of the component 310). Then, based on the field information of the shape 211c of the component.
Thus, when the component to be measured is determined, the off-line teaching device 20 outputs a measurement command to the sensor unit 11 via the external IF unit 207, and acquires the measurement result by the sensor unit 11. Then, based on the acquired measurement result, a three-dimensional model (measured three-dimensional model) of the component is created, and the process proceeds to step S12.

In step S12, the off-line teaching device 20 compares the actually measured three-dimensional model created in step S11 with the existing three-dimensional model in the virtual space corresponding to the actually measured three-dimensional model. When the existing three-dimensional model is a simplified shape model in which the outside diameter dimension is made equal to the actual component, as in the three-dimensional model 301 of FIG. 9A, the actually measured three-dimensional model is as shown in FIG. As in the three-dimensional models 311 to 313 of (d), there are cases where the existing three-dimensional model 301 and the outer diameter size are substantially the same, but the shape differs from the existing three-dimensional model 301.
Therefore, for example, as shown in FIGS. 10A and 10B, the existing three-dimensional model 301 and the actually measured three-dimensional model 311 are respectively represented by voxels, and their shapes are compared. In this case, the difference in shape between the existing three-dimensional model 301 and the measured three-dimensional model 311 is confirmed using a method of extracting the Fourier spectrum of the voxelized shape.

  Next, in step S13, the off-line teaching device 20 affects the robot route created in the pre-study process based on the information of the component determined to have the difference in shape in step S12. It is determined whether or not. Here, when the robot route created in the preliminary examination process can be corrected to the shortest route with a shorter work tact due to the difference in the shape, it is determined that the difference in the shape affects the robot route. In addition, since the robot route created in the preliminary examination processing causes interference with the components of the actual machine, the robot route differs in shape even if it is necessary to correct the robot route to a route avoiding the interference. To influence the

First, a method of determining whether or not correction to the shortest path is possible will be described.
In this case, as shown in FIG. 11, the robot 10 and the three-dimensional model 310 of the components of the actual machine are crawled from both sides in the X direction. At this time, the size of the height direction (Z direction) of the surface of the existing three-dimensional model facing the robot 10 and the size of the height direction of the surface of the measured three-dimensional model facing the robot 10 (Z direction) Focusing on the size in the height direction (Z direction) and the depth direction (Y direction) of the end effector 13 of the robot 10.
For example, when the measured three-dimensional model is the three-dimensional model 311 shown in FIG. 9B, as shown in FIG. 12, for the existing three-dimensional model 301, attention is focused on the size H1. Also, for the measured three-dimensional model 311, attention is focused on the sizes H2 and H3.

When the measured three-dimensional model is the three-dimensional model 312 shown in FIG. 9C, as shown in FIG. 13, for the existing three-dimensional model 301, attention is focused on the size H1 as in FIG. Also, for the measured three-dimensional model 312, attention is focused on the size H4.
Similarly, when the measured three-dimensional model is the three-dimensional model 313 shown in FIG. 9D, as shown in FIG. 14, the size H1 is focused on the existing three-dimensional model 301. Further, for the measured three-dimensional model 313, attention is focused on the size H5 (= H1).

Then, based on the above noted size, it is determined whether the difference in shape between the existing three-dimensional model and the measured three-dimensional model affects the robot route created in the preliminary examination process.
In the present embodiment, when the end effector 13 of the robot 10 has a shape of height H, width W, and depth D as shown in FIG. 15, the height direction in the plane facing the robot 10 of the existing three-dimensional model The robot path is affected when the value obtained by subtracting the sum of the size in the height direction on the surface of the measured three-dimensional model facing the robot 10 from the size of the is greater than the size of the height H or width W of the end effector 13 Judge to give.
For example, in the case of the example shown in FIG. 12, when H1- (H2 + H3)> H or H1- (H2 + H3)> D, as shown in FIG. 16, the end effector 13 of the robot 10 is different from the existing three-dimensional model 301. There is a possibility that it can pass through the opening part which is a part. Therefore, it is determined that the shortest path may be newly found, and in this case, it is determined that the existing robot path is affected.

For example, also in the case of the example shown in FIG. 13, when H1-H4> H or H1-H4> D, as shown in FIG. 17, the end effector 13 of the robot 10 differs from the existing three-dimensional model 301. There is a possibility that it can pass through the opening part which is a part. Therefore, also in this case, it is determined that the existing robot path is affected (the shortest path can be corrected).
On the other hand, in the case of the example shown in FIG. 14, for example, H1-H5 = 0, and the above conditions are not satisfied, so the existing robot route is not affected (corrected to the shortest route) Can not do).
From the above, it can be determined whether or not the correction to the shortest path is possible. This determination is made on both sides in the X direction of FIG.

Next, a method of determining whether a correction to the interference avoidance path is necessary will be described.
FIG. 18 is a diagram showing an example of a robot path created in the preliminary examination process. The robot path 403 is a path through which the robot 10 passes the teaching point a → b → c → d → e → f in the vicinity of the existing three-dimensional model 301 in the virtual space.
As shown in FIG. 19, the existing three-dimensional model 301 in FIG. 18 is replaced with the actually measured three-dimensional model 314 to determine whether a correction to the interference avoidance path is necessary. Judging based on the positional relationship.

In the case of the example shown in FIG. 19, it can be understood that interference occurs in the path passing the teaching point b → c → d due to the difference in shape. That is, the path passing through the teaching point b → c → d is a path that requires correction to the interference avoidance path.
In this way, it can be determined whether correction to the interference avoidance path is possible.
If the off-line teaching device 20 determines that the shape difference affects the existing robot route in step S13 of FIG. 7, the off-line teaching device 20 determines that the robot route is to be corrected, proceeds to step S14, and affects the existing robot route. If it is determined that the robot path is not given, it is determined that the robot path is not to be corrected, and the process proceeds to step S16 described later.
In step S14, the off-line teaching device 20 corrects the robot path. Here, the off-line teaching device 20 partially corrects the robot path created in the preliminary examination process only for the path that needs to be corrected.

First, the method of correction to the shortest path will be described.
FIG. 20 is a diagram showing an example of a robot path created in the preliminary examination process. The robot path 403 is a path through which the robot 10 passes the teaching point a → b → c → d → e → f in the vicinity of the existing three-dimensional model 301 in the virtual space. The robot path 404 is a path through which the robot 10 passes the teaching point x → y → z in the vicinity of the existing three-dimensional model 301 in the virtual space.
When searching for the shortest path, first, as shown in FIG. 21, the existing three-dimensional model 301 in FIG. 20 is replaced with the actually measured three-dimensional model 311.

Next, a near area 501 at a certain distance from the measured three-dimensional model 311 is set, and a teaching point in the near area 501 is searched. In the example shown in FIG. 21, the teaching points b, c, d, e and y correspond to this. Here, the near region 501 is a cuboid space, but the near region 501 may be any shape as long as it can define a region near the measured three-dimensional model, and may be, for example, a spherical shape.
Further, a route from the outside of the near region 501 to a teaching point in the near region 501 (that is, a route coming in from the outside of the near region 501 into the near region 501) is searched. In the example shown in FIG. 21, the path 411 of a → b and the path 412 of x → y correspond to this.
Similarly, a path from inside the near area 501 to outside the near area 501 is searched. In the example shown in FIG. 21, the path 413 of e → f and the path 414 of y → z correspond to this.

Next, the end point of the route (route 411 and route 412 in FIG. 21) entering from the outside of the near region 501 into the near region 501, and the route from the inside of the near region 501 to outside the near region 501 (in FIG. The teaching point where the start point of the path 414) is not at the same coordinates is searched. In the example shown in FIG. 21, the teaching point b and the teaching point e correspond to this.
In FIG. 21, in the path of teaching point x → y → z, the teaching point y has elements of both the start point and the end point, and the three-dimensional model is obtained only by the end effector 13 entering into the near area 501 for a moment. It can be determined that the unrolled portion of 311 is irrelevant to the passage of the end effector.
Therefore, it is understood that the shortest path search may be performed with the teaching point b as the start point and the teaching point e as the end point.

For example, a search tree method is used for the shortest path search.
FIG. 22 is a diagram for explaining a path search method by the graph search method.
As shown in FIG. 22, the case of searching for the shortest path from the departure point 601 to the target point 602 in the space 600 will be considered. Here, it is assumed that an obstacle, that is, a region 603 which can not pass through exists in the space 600. In addition, although it represents with two-dimensional space for description here, it is the same also in three-dimensional space.

First, as shown in FIG. 22, the space 600 is meshed.
Next, as shown in FIG. 23, it is considered to move from the departure point 601 and the target point 602 toward the adjacent meshes. At this time, if there is an obstacle 603 on the adjacent mesh, it does not proceed. For example, in the case of the starting point 601 in FIG. 23, since the obstacle 603 exists in the directions of C and D among the eight directions A to H, the obstacle 603 is prevented from advancing. On the other hand, since there is no obstacle 603 in the adjacent mesh from the target point 602, it is possible to move in all directions A to H.

Then, as shown in FIG. 24, the mesh is advanced to the adjacent mesh where there is no obstacle in the same way as the mesh that is advanced (for example, the meshes 604 and 605). This is repeated at both the departure point 601 and the target point 602, and the shortest path that travels from both sides and enters the same mesh is the shortest path sought. In this example, as shown in FIG. 25, the route 610 is the shortest route.
With the teaching point b in FIG. 21 as the start point and the teaching point e as the end point, the shortest path searched by the above method is as shown by the path 415 in FIG. In this way, the existing robot path can be corrected to the shortest path.

As described above, since the above method is used for the path search algorithm, if the neighborhood area 501 is made wide, path correction of a long distance is possible and the correction path can be smoothed. However, in this case, it can be easily inferred that the processing time will be long, so the proximity region 501 may be set depending on whether to prioritize the processing time or the smoothness of the route.
In addition, although the case where a shortest path was calculated | required using a graph search method was demonstrated here, you may use another algorithm (sampling base search methods, such as RRT and PRM, etc.).

Next, the correction method to the interference avoidance path will be described.
First, the start point and the end point of the interference avoidance path are searched. In the case of the example shown in FIG. 19 described above, the path passing through the teaching point b → c → d is a path that requires correction to the interference avoidance path. Therefore, in this case, a search for an interference avoidance path is performed with the teaching point b as the start point and the teaching point d as the end point.
The path search for interference avoidance is performed in the same manner as the shortest path search described above.
As a result, as shown in FIG. 27, with the teaching point b as the departure point and the teaching point d as the target point, a corrected path 416 in which interference with the measured three-dimensional model 314 is avoided is obtained. The correction path 416 passes, for example, the teaching points g and h newly set in the path search.
As described above, after correcting the path through the shortest path or avoiding the interference, the off-line teaching device 20 proceeds to step S15 in FIG. 7 and updates the virtual robot system. That is, the off-line teaching device 20 exchanges the three-dimensional model in the virtual space with the measured three-dimensional model to update the robot system in the virtual space.

Next, in step S16, the off-line teaching device 20 transfers the robot path data after correction to the actual robot system. Thus, the operation check by the real robot system is performed.
As described above, in the present embodiment, the movement path of the robot taught by the virtual robot system in the virtual space is corrected according to the shape error of the virtual three-dimensional model constituting the virtual robot system in advance for examination. .
If there is a difference in the shape of the components of the robot system between the virtual robot system and the real robot system, if the robot is operated as it is with the operation program created on the virtual robot system, the robot and peripheral devices in the real robot system Interference may occur or a desired work tact may not be realized.

  Therefore, in the conventional off-line teaching device, the created operation program is transferred to the real machine robot system, adjustment work is performed by the real machine robot system, and the movement program is corrected. Furthermore, if the operation program corrected by the actual robot system is returned to the off-line teaching device, the robot and peripheral devices will generate interference etc. in the simulation on the off-line teaching device, so it will be the same as the actual robot system. In this way, the robot system in the virtual space is rebuilt and the re-teaching operation is performed.

On the other hand, in the present embodiment, when a shape error of the virtual three-dimensional model occurs, the operation program created on the virtual robot system is automatically corrected and transferred to the real robot system. Therefore, there is no need for on-site adjustment or re-teaching work using a real robot system. As described above, even when there is a shape difference between components in the virtual space and the actual robot system, appropriate off-line teaching can be performed.
In addition, it is necessary to determine whether the difference in shape of the three-dimensional model affects the robot path created in the preliminary examination process, that is, whether the robot path needs to be corrected or not. If it is determined that the robot route is corrected. Therefore, the correction process can be performed efficiently.

  Furthermore, whether or not a shape error of the virtual three-dimensional model is generated is determined by measuring the shape of the component of the actual robot system using a visual sensor, a distance sensor, or the like. Specifically, based on the measurement result by the sensor, an actual measurement three-dimensional model of the component is created, the actual measurement three-dimensional model and the virtual three-dimensional model are compared, and the shape error of the virtual three-dimensional model is calculated. . As described above, since the difference in shape is determined based on the measurement result, it is possible to reliably grasp that there is a difference in the shape of the component of the robot system between the virtual robot system and the actual robot system.

Further, at this time, the shape measurement by the sensor is performed only on the component on the real machine robot system corresponding to the virtual three-dimensional model which is determined that there is a possibility that a shape error has occurred. Here, whether or not there is a possibility that a shape error has occurred is determined, for example, according to whether or not the virtual three-dimensional model is a simplified shape model created by the user.
For robots, three-dimensional data prepared by the robot manufacturer can be used, but for peripheral equipment of robots, there are many items such as tools, etc. There are two-dimensional drawings but three-dimensional models are prepared Often not. In particular, since peripheral devices are procured at a production site, creating a new three-dimensional model is a very time-consuming task for workers at a production site not designed with a three-dimensional CAD apparatus. Therefore, with regard to peripheral devices, a simplified shape model having almost the same external dimensions as a three-dimensional model is often created and substituted. Thus, in the virtual three-dimensional model constituting the virtual robot system, data having different model shape reliability is mixed.

  In the present embodiment, when constructing a virtual robot system, whether each virtual three-dimensional model has high reliability data prepared by a manufacturer or the like, or low reliability data easily generated by the user. Attribute information (shape correction implementation flag 211e) indicating whether or not Then, referring to the attribute information, a virtual three-dimensional model which may have a shape error is determined. Therefore, it is possible to relatively easily and appropriately determine the components that require shape measurement by the sensor.

Further, when comparing the measured three-dimensional model and the virtual three-dimensional model, both are represented by voxel data to determine the difference in shape. Therefore, shape comparison can be performed appropriately.
Furthermore, whether the shape difference of the three-dimensional model affects the robot route created in the preliminary examination process, that is, whether the robot route needs to be corrected, the robot route is the shortest movement route. It is determined according to whether or not correction is possible, or whether the robot path is a path where interference occurs.

  In the pre-study process, the robot and the peripheral device do not interfere with each other on the virtual robot system, and off-line teaching is performed so as to minimize the work tact, and an operation program is created. However, if there is a difference between the shape of the virtual three-dimensional model and the shape of the component of the actual machine, the shortest route to the work point may be changed (a shortest route may be found). In this embodiment, if it is determined that the robot path can be corrected to the shortest movement path by the difference in shape of the three-dimensional model, the robot path can be corrected to the shortest movement path. Thereby, the shortest path to the work point can be properly taught, and the work tact time can be further shortened.

  Also, at this time, whether or not the robot path can be corrected to the shortest movement path based on the shape of the virtual three-dimensional model, the shape of the measured three-dimensional model, and the shape of the three-dimensional model of the end effector of the robot Determine Specifically, from the length of the side in the height direction that constitutes the face of the virtual three-dimensional model facing the robot, the length of the side in the height direction that constitutes the face of the measured three-dimensional model facing the robot When the subtracted length is longer than the length corresponding to the height direction of the three-dimensional model of the end effector, it is determined that the robot path can be corrected to the shortest movement path. Therefore, when the component of the real robot system has a cutout and the size is such that the end effector can pass, the path through which the end effector passes through the cutout is the shortest movement path. It can be determined that the robot path can be corrected.

Furthermore, in the present embodiment, when it is determined that the robot path is a path where interference occurs in the real robot system due to the difference in shape of the three-dimensional model, the robot path can be corrected to the interference avoidance path. Therefore, it is possible to eliminate the adjustment work after transferring the operation program to the actual robot system.
At this time, based on the shape of the measured three-dimensional model and the robot path, it is determined whether the robot path is a path where interference occurs. As described above, it is possible to easily determine whether or not interference occurs from the positional relationship between the measured three-dimensional model and the robot path.
In addition, when correcting to the shortest movement path or the interference avoidance path, among the robot paths created in the preliminary examination process, only the movement path that needs correction in the vicinity of the virtual three-dimensional model is partially corrected. Therefore, the correction processing time can be shortened.

When correcting to the shortest movement path, the virtual three-dimensional model is replaced with the actual measurement three-dimensional model in the virtual space, and a near region at a predetermined distance from the actual measurement three-dimensional model is set. Then, the start point and the end point of the movement path to be partially corrected are selected from the teaching points existing in the near region, and the shortest movement path connecting the selected teaching points is searched. At this time, a teaching point that is the end point of the movement path from the teaching point outside the near area to the teaching point inside the near area is selected as the start point, and the moving path from the teaching point within the near area toward the teaching point outside the near area Select the teaching point that is the start point of as the end point.
Since the path search is performed by selecting the start point and the end point of the movement path to be partially corrected as described above, a path where the end effector passes through the unrolled portion of the three-dimensional model may be searched It becomes possible. Therefore, it is possible to search for a path that moves to the work point at the shortest distance than the robot path created in the preliminary examination process.

In addition, when correcting to the collision avoidance path, the teaching point as the start point and the teaching point as the end point of the movement path in which the interference is generated are respectively set as the start point and the end point of the movement path to be partially corrected. An interference avoidance route connecting selected teaching points is searched. Therefore, it is possible to properly teach a robot path in which no interference occurs in the real robot system.
As described above, the robot path taught by the robot system model constructed using the three-dimensional model of the component in the virtual space for the preliminary examination is a three-dimensional model of the component measured after construction of the real robot system. Depending on the shape, it is possible to automatically correct only the part that needs correction so as to make the robot pass the shortest path or let the robot pass a path avoiding interference. Therefore, even when there is a difference in shape between the virtual robot system components and the real robot system components, it is possible to create appropriate off-line teaching data and save time and time for re-teaching. Can.

(Modification)
In the above embodiment, although the case where the shape of the peripheral component of the robot 10 is measured by the sensor unit 11 has been described, the shape of the end effector 13 of the robot 10 may be measured by the sensor unit 11. In this case, the height H or width of the end effector 13 used to determine whether the difference in shape between the existing three-dimensional model and the actual three-dimensional model affects the robot route created in the preliminary examination process The measured value can be used as the value of W. Therefore, more accurate judgment can be made.

  In the above embodiment, the shape of the component on the real machine robot system is measured by the sensor unit 11, and the model shape is obtained by comparing the actual three-dimensional model and the existing three-dimensional model created based on the measurement result. Although the case of confirming the difference of is described, it is not limited thereto. For example, information on the shape error of the existing three-dimensional model (difference in shape from the corresponding component of the actual robot system) may be directly input from the outside by the user or the like. In this case, the processes of steps S11 and S12 of FIG. 7 become unnecessary.

(Other embodiments)
The present invention supplies a program that implements one or more functions of the above-described embodiments to a system or apparatus via a network or storage medium, and one or more processors in a computer of the system or apparatus read and execute the program. Can also be realized. It can also be implemented by a circuit (eg, an ASIC) that implements one or more functions.

  DESCRIPTION OF SYMBOLS 10 ... Robot, 11 ... Sensor part, 12 ... Position and attitude change mechanism, 13 ... Hand, 20 ... Offline teaching device, 201 ... System construction part, 202 ... Robot path | route preparation part, 203 ... Operation check part, 204 ... Measurement and model Creation unit 205 Shape difference determination unit 206 Path correction unit 207 External IF unit 208 Input / output control unit

Claims (15)

Acquisition means for acquiring a measured three-dimensional model indicating the shape of a peripheral structure of a real robot;
In the acquired actual three-dimensional model, if the movement path of the robot can be made shorter according to the teaching data indicating the movement path of the robot, or if the movement of the robot due to the teaching data causes interference with the surrounding structure And a correction unit configured to correct the teaching data.   The teaching data creation device according to claim 1, wherein the acquisition unit is a visual sensor or a distance sensor.   Furthermore, the virtual robot system has a virtual robot system in which virtual three-dimensional models of the robot and a peripheral structure of the robot are arranged in a virtual space, and the teaching data is created. The teaching data creation device described in. The virtual three-dimensional model is a three-dimensional model to which attribute information indicating the reliability of a model shape is added,
The teaching data creation device according to any one of claims 1 to 3, wherein an actual measurement three-dimensional model is acquired by the acquisition unit based on attribute information added to the virtual three-dimensional model.   The teaching data creation device according to any one of claims 1 to 4, wherein the correction means partially corrects a movement path of the robot based on the teaching data.   The teaching data creation device according to any one of claims 1 to 5, further comprising transfer means for transferring the corrected teaching data to the robot system. An acquisition step of acquiring a measured three-dimensional model indicating a shape of a peripheral structure of a real robot;
In the acquired actual three-dimensional model, if the movement path of the robot can be made shorter according to the teaching data indicating the movement path of the robot, or if the movement of the robot due to the teaching data causes interference with the surrounding structure And a correction step of correcting the teaching data.   The control method of the teaching data creation device according to claim 7, wherein in the acquisition step, a measured three-dimensional model is acquired using a visual sensor or a distance sensor.   The virtual robot system may further comprise a virtual robot system in which virtual three-dimensional models of the robot and a peripheral structure of the robot are arranged on a virtual space, and the creating step of creating the teaching data. A control method of the teaching data creation device according to the above. The virtual three-dimensional model is a three-dimensional model to which attribute information indicating the reliability of a model shape is added,
The control of the teaching data creation device according to any one of claims 7 to 9, wherein an actual measurement three-dimensional model is acquired by the acquisition step based on attribute information added to the virtual three-dimensional model. Method.   The control method of the teaching data creation device according to any one of claims 7 to 10, wherein, in the correction step, a movement path of the robot based on the teaching data is partially corrected.   The control method of the teaching data creation device according to any one of claims 7 to 11, further comprising a consonant including a transferring step of transferring the corrected teaching data to a robot system. A robot, and the teaching data creation device according to any one of claims 1 to 6.
A control system for operating the robot based on the corrected teaching data.   A program for executing the control method according to any one of claims 7 to 12 by being read and executed by a computer.   A storage medium storing a program for executing the control method according to any one of claims 7 to 12 by being read and executed by a computer.

Images