CN118204963A - Robot automatic teaching method and robot control device - Google Patents

Abstract

The center of the work table can be taught automatically to a robot for carrying a work piece between work tables more accurately. 3N (where N.gtoreq.2) directions of a position determining tool (61) having a known positional relationship with the table center C are determined (step 101), and N groups are obtained by repeating the selection of the directions for every N (th) and setting groups of three directions (step 102). For each group, the hand (14) is brought close to the position determination jig (61) using three directions of the group, and the coordinates of the position determination jig (61) are obtained (step 103). An average value of the coordinates obtained for each group is calculated (step 104), and the average value is used as the coordinates of the position determination jig (61) for automatic teaching (step 107).

Description

Robot automatic teaching method and robot control device

Technical Field

The present invention relates to an automatic teaching method for a robot and a robot control device for performing the automatic teaching method.

Background

In a manufacturing process of a semiconductor device, a transfer robot for transferring a semiconductor wafer as a workpiece between stages is used. In the following description, an object to be taken out (i.e., loaded) and unloaded (i.e., unloaded) of a workpiece by a robot is collectively referred to as a table. In the semiconductor manufacturing process, a cassette for accommodating wafers and a workpiece processing apparatus for performing a certain process on the wafers are each a table. The work table strictly defines a region where a workpiece is to be placed when being transported by a robot (referred to as a workpiece placement region), and the center of the workpiece placement position is referred to as a work table center. In order to convey a workpiece between tables using a robot, it is necessary to teach (teach) the robot coordinates of the table center in the coordinate system of the robot for each table. In the case of carrying a plate-like workpiece such as a semiconductor wafer by a horizontal multi-joint robot, the workpiece is carried while maintaining its horizontal posture, and the workpiece is slightly moved in the vertical direction in order to load and unload the workpiece in the table, so that the center of the table in the horizontal plane may be accurately taught.

Conventionally, a suspension type console is connected to a robot control device that controls a robot, and the robot is manually operated via the suspension type console, whereby the robot is taught the exact position of the table. However, there are technical problems that the manual teaching causes a deviation due to an operator who performs the teaching, and the teaching requires a long time. The general positional relationship between the robot and the table is known from design data of the robot and the table, installation data when the robot and the table are installed on site, and the like. Accordingly, patent documents 1 and 2 disclose that a cylindrical jig (also referred to as a pin) having a known positional relationship with the center of the table is placed in the table, and the center of the table is accurately obtained by detecting the jig in a noncontact manner by a sensor attached to the hand while moving the hand (also referred to as an end effector) at the front end of the robot in three different directions. According to this method, the position of the table center can be accurately determined in the coordinate system of the robot, and automatic teaching of the table center can be performed. As the sensor, a correlation sensor is used which has a light emitting portion and a light receiving portion and detects that the optical path is blocked by the jig. Patent document 3 discloses that, when detecting a jig using a correlation sensor, a regression analysis is performed by performing a plurality of detection operations or detection accuracy is improved by performing least squares approximation and numerical search.

Prior art literature

Patent literature

Patent document 1: japanese patent application laid-open No. 2010-284728

Patent document 2: japanese patent laid-open publication No. 2013-153187

Patent document 3: japanese patent application laid-open No. 2022-520052

Disclosure of Invention

Technical problem to be solved by the invention

When the coordinate value of the center of the table is obtained by detecting a columnar jig in a noncontact manner using an correlation sensor provided on the hand, there is a case where reproducibility is poor when the coordinate value is repeatedly obtained, and a position determination error is not necessarily small, whereby automatic teaching of the robot may not be performed with sufficient accuracy.

The invention aims to provide an automatic teaching method capable of automatically teaching a workbench center more accurately and a robot control device for executing the automatic teaching method.

Technical proposal adopted for solving the technical problems

An automatic teaching method according to an aspect of the present invention is an automatic teaching method for automatically teaching a position of a table center of a table connected to a work area via a connection portion as an opening to a robot having a hand and disposed in the interior of the work area, wherein a direction from the interior of the work area toward the table center via the connection portion is defined as a Y direction, a direction perpendicular to the Y direction is defined as an X direction, an XY coordinate system is defined as a robot coordinate system, and a plane extending from the X direction and the Y direction is defined as an XY plane, the automatic teaching method comprising: a step of disposing a columnar position-determining jig at a position having a known relative positional relationship with the center of the table in the table; a first setting step of setting N to be an integer of 2 or more, and determining 3N directions in an XY plane within an angle range that can be taken by a hand toward the position determining jig when the hand is entered into the stage; a second setting step of setting different N groups by selecting directions while changing a starting point from one end side of the angle range, the groups being composed of three directions selected for each N directions from among the 3N directions determined in 1; a position calculation step of bringing the hand close to the position determination jig from three directions included in the group for each group set in the second setting step, detecting the position determination jig in a noncontact manner by a sensor provided on the hand, and calculating coordinates of the position determination jig in the robot coordinate system; and an average calculation step of obtaining an average value of the coordinates obtained for all the groups, and using the average value as the coordinates of the position determination jig used for teaching the robot.

According to the automatic teaching method of the first aspect, the position of the position-determining jig in the robot coordinate system can be determined by combining the condition of bringing the hand closer to the position-determining jig in as many directions as possible and the condition of bringing the hand closer to the position-determining jig in three directions different from each other as far as possible, so that the deviation in the determined coordinates of the position-determining jig becomes small, and the automatic teaching accuracy can be improved.

In the automatic teaching method according to one aspect, it is preferable that the position calculation step is performed by searching for a deviation value from the coordinates obtained in the position calculation step, and if the deviation value is present, the average value is calculated by excluding the deviation value. By eliminating the deviation value, the deviation of the position determining jig can be further reduced, and the automatic teaching accuracy can be further improved. In this case, in the average calculation process, the variance σ of the coordinates obtained for all the groups is also calculated, and the determination of the deviation value may be made based on the variance σ. The variance sigma is used to facilitate the determination of the deviation value.

In the automatic teaching method according to one aspect, it is preferable that in the first setting step, 3N directions are determined at equal angular intervals as viewed from the position determining jig. By determining the directions at equal intervals, various error factors are averaged, and the deviation of the positions with respect to the position determination jig is further reduced.

In one embodiment, the robot is a horizontal multi-joint robot, and the X direction and the Y direction are directions in a horizontal plane. In this case, if the horizontal multi-joint robot is a transfer robot that transfers a workpiece, the frequency of failure in placement of the workpiece can be reduced. In this case, it is preferable that the hand is moved in the horizontal plane in the position calculating step. By moving the hand in the horizontal plane, the influence of mechanical errors of the robot due to moving the hand in the vertical direction can be eliminated. In one aspect, the Y direction is, for example, a direction perpendicular to a wall surface of the work area at a position of the connecting portion. With this configuration, when the work tables are arranged on the wall surface of the work area in the X direction, the automatic teaching of the table center of each work table can be easily performed.

In one embodiment of the automatic teaching method, the sensor is, for example, a correlation sensor. In the case where the robot is a robot for transporting semiconductor wafers, such an correlation sensor is often provided for confirming the stock condition of the wafers, and in such a case, the accuracy of automatic teaching can be improved by using the conventional correlation sensor.

A robot control device of one embodiment is a robot control device for controlling a robot having a hand and disposed in an interior of a work area, automatically teaching the robot a position of a table center of a table connected to the work area via a connection portion as an opening, wherein a direction from the interior of the work area toward the table center via the connection portion is set as a Y direction, and a direction perpendicular to the Y direction is set as an X direction to determine an XY coordinate system, the XY coordinate system is a robot coordinate system, a plane developed by the X direction and the Y direction is an XY plane, when a cylindrical position determination jig is disposed in the interior of the table at a position having a known relative positional relation with the table center, N is an integer of 2 or more, when the hand is made to enter the interior of the table, the robot is oriented toward the position determination jig, 3N directions are determined in an XY plane within an angle range that the hand can take, three directions selected from among the 3N directions are set as an X direction, a group of three directions are set while a starting point is changed from one end side of the angle range, the N groups are set as an average, the directions are set as N groups are set as an average, the coordinate system is set by the robot is changed from one end side of the three directions, the three groups are set as an average value, the coordinate system is set by the robot is set as an average, the position sensor is set by the average sensor is set, and the position is determined by the position sensor is set in a position sensor is set by the average, and the position sensor is set to be determined by a position sensor is set by a position sensor, and the position sensor is set to be a position sensor is set.

According to the robot control device of the first aspect, the condition that the hand approaches the position determining jig in as many directions as possible and the condition that the hand approaches the position determining jig in three directions different from each other as far as possible can be combined to determine the position of the position determining jig in the robot coordinate system, so that the deviation in the determined coordinates of the position determining jig becomes small, and the automatic teaching accuracy is improved.

Effects of the invention

According to the invention, the center of the workbench can be automatically taught to the robot more accurately.

Drawings

Fig. 1 (a) is a plan view showing a robot, (B) is a schematic cross-sectional view taken along line B-B of fig. 1 (a), and (c) is an enlarged plan view showing a hand.

Fig. 2 (a) to (d) are diagrams illustrating automatic teaching of the table center.

Fig. 3 is a diagram illustrating a method of determining coordinates of the jig center O.

Fig. 4 is a flowchart illustrating an automatic teaching method.

Description of the reference numerals

1 … Robots; 5 … working areas; 10 … base stations; 11 … first arms; a second arm 12; 13 … third arm; 14 … hands; 15 … lifting cylinders; 20 … prongs; 21-24 … motor; 25 … correlation sensor; 26 … light-emitting parts; 27 … light receiving sections; 28 … optical paths; 30 … robot controls; 50 … pieces; 51 … table; 52 … connections; 53 … workpiece configuration positions; 61. 62 … a position-determining jig; c … workbench center; and O … is arranged at the center of the clamp.

Detailed Description

Next, an embodiment of the present invention will be described with reference to the drawings. Fig. 1 shows a robot to which an automatic teaching method according to an embodiment of the present invention is applied. The illustrated robot 1 is a horizontal multi-joint robot, and is provided in a work area 5, which is a rectangular space, for example, and is configured to convey a plate-like workpiece 50 between work tables 51 provided on a wall surface surrounding the work area 5. The work area 5 is a space in which the robot 1 can move the arms 11 to 13 and the hand 14 without interfering with a wall surface or the like when the work 50 is conveyed between the tables 51. Each table 51 is a place where the robot 1 loads and unloads the workpiece 50, and is connected to the work area 5 via the connection portion 52. The connection portion 52 is configured as an opening provided in the wall surface of the work area 5 so that the hand 14 of the robot 1 on which the workpiece 50 is mounted can enter the interior of the table 51. Therefore, when the robot 1 loads and unloads the work 50 on and from the table 51, the movement direction of the hand 14 when passing through the connection portion 52 is generally a direction perpendicular to the wall surface of the work area 5. A workpiece placement position 53 is defined on the table 51, and the workpiece placement position 53 is a position at which the workpiece 50 is placed on the table 51 when the robot 1 performs loading and unloading of the workpiece 50. The center of the work placement position 53 is set as a table center C.

Next, a detailed configuration of the robot 1 will be described. The robot 1 includes: a base 10 disposed and fixed on the floor of the work area 5; first arm 11, second arm 12, and third arm 13, which are three arms connected in series to base 10; and a hand 14 mounted on the third arm 13. The base 10 includes a lift cylinder 15 that is driven by a lift motor (not shown) to move up and down. Each of the arms 11 to 13 and the hand 14 has a base end portion and a tip end portion, and the base end portion of the first arm 11 is rotatably coupled to the lift cylinder 15, whereby the first arm 11 is held by the base 10. The first arm 11 can be lifted and lowered relative to the base 10 in accordance with the lifting and lowering of the lifting and lowering cylinder 15. The lifting of the lifting cylinder 15 causes the arms 11 to 13 and the hand 14 to be lifted integrally, but the present embodiment relates to teaching of the horizontal articulated robot 1 in the horizontal plane, and since the movement in the height direction of the lifting cylinder 15 is smaller than the movement of the arms 11 to 13 and the hand 14 in the horizontal plane, the movement of the robot 1 in the height direction by the lifting cylinder 15 will not be described in detail below.

The first arm 11 is driven by a motor 21 incorporated in the lift cylinder 15 to rotate about a rotation axis J0 in a horizontal plane. The base end portion of the second arm 12 is rotatably coupled to the tip end portion of the first arm 11, and the second arm 12 is held by the first arm 11 and is driven by a motor 22 incorporated in the first arm 11 to rotate in a horizontal plane about the rotation axis J1. Similarly, the base end portion of the third arm 13 is rotatably held at the tip end portion of the second arm 12, and the third arm 13 is driven by a motor 23 incorporated in the second arm 12 to rotate about the rotation axis J2 in the horizontal plane. The base end portion of the hand 14 is rotatably held at the tip end portion of the third arm 13, and the hand 14 is driven by a motor 24 incorporated in the third arm 13 to rotate about the rotation axis J3 in the horizontal plane.

Fig. 1 (c) is an enlarged plan view showing the structure of the hand 14. The distal end portion of the hand 14 is bifurcated into two branches to form a fork portion 20. The workpiece 50 is horizontally placed on the surface of the fork 20 in the hand 14 during conveyance. A light emitting portion 26 for emitting laser light is provided at the tip of one branch of the fork portion 20, a light receiving portion 27 into which the laser light from the light emitting portion 26 is incident is provided at the tip of the other branch, and the correlation sensor 25 is constituted by the light emitting portion 26 and the light receiving portion 27. In the figure, an arrow from the light emitting portion 26 to the light receiving portion 27 indicates an optical path 28 of light from the light emitting portion 26 to the light receiving portion 27. According to the correlation sensor 25, whether or not an object blocking the optical path 28 between the light emitting portion 26 and the light receiving portion 27 is present can be detected in a noncontact manner, depending on whether or not the light emitting portion 26 can be detected in the light receiving portion 27. Such an correlation sensor 25 is generally provided in the robot 1 for carrying semiconductor wafers, so that, when the table 51 is a cassette for storing the workpieces 50 such as semiconductor wafers for each cassette, mapping of the stock condition of each cassette is examined.

In order to explain the operation of the robot 1 as a horizontal multi-joint robot, XY coordinates are set in the horizontal plane. Here, as shown in the drawing, when the work area 5 is a rectangular space and a plurality of tables 51 are arranged along the long side thereof, the extending direction of the long side is defined as the X direction, and the direction perpendicular to the X direction is defined as the Y direction. When the robot 1 enters the table 51 using the hand 14 thereof, the hand 14 moves in the Y direction and passes through the connection portion 52 to enter the inside of the table 51. The coordinate system of the robot 1 in the horizontal plane (hereinafter referred to as a robot coordinate system) is represented by the XY coordinate system described above with the position of the rotation axis J0 as the origin. Hereinafter, the horizontal plane in which the XY coordinate system is set in this way is referred to as an XY plane. In addition, when loading or unloading the workpiece 50 on or from the table 51, the robot 1 needs to be operated using the coordinates of the table center C in the robot coordinate system, and therefore, when teaching the robot 1, it is necessary to teach the exact position of the table center C in the XY plane. As shown in fig. 1b, a robot controller 30 for controlling the robot 1 is connected to the robot 1, and the robot controller 30 can drive and control the motors 21 to 24 and a lifting motor (not shown) based on an externally input command. The robot control device 30 can control the robot 1 to perform automatic teaching described below.

Next, automatic teaching for accurately determining the position of the table center C in the robot coordinate system will be described. In the automatic teaching, the coordinates of the table center C in the robot coordinate system in the horizontal plane, that is, the coordinates of the table center C in the XY plane are obtained. Therefore, in the table 51, the columnar position determination jigs 61, 62 are disposed at two positions which are away from the table center C and have a known relative positional relationship with the table center C, respectively. In the figure, the position specifying jigs 61 and 62 are disposed on the outer periphery of the workpiece disposition position 53, but the positions where the position specifying jigs 61 and 62 are disposed are not limited to this. Since the position determination jigs 61, 62 are provided to stand upright with respect to the XY plane, that is, the horizontal plane, the axes of the position determination jigs 61, 62 as cylinders extend in the vertical direction, and the position of the axes in the XY plane is set as the jig center O. If the coordinates of the jig centers O of the two position determination jigs 61, 62 in the robot coordinate system are known, the positional relationship between each jig center O and the table center C is also known, and therefore the position of the table center C in the robot coordinate system, that is, the XY coordinate value thereof can be calculated.

Fig. 2 is a diagram illustrating automatic teaching. In the present embodiment, XY coordinate values of the jig center O of each of the position determination jigs 61, 62 in the robot coordinate system are determined in a noncontact manner using the correlation sensor 25 provided on the hand of the robot 1. The position determining jigs 61, 62 are cylindrical, and the cross-sectional shape of the horizontal plane thereof is a perfect circle, and the center of the perfect circle is the jig center O. Therefore, the robot 1 is operated, and the hand 14 is brought closer to the position specification jigs 61 and 62 in three different directions from the work area 5 side for each position specification jig 61 and 62. In this case, since the hand 14 is inserted into the table 51 through the opening of the connecting portion 52, the angular range that the hand 14 can take for each of the position determining jigs 61, 62 is restricted by the size of the opening of the connecting portion 52. From the design data of the work area 5 and the table 51 and the installation data of where the robot 1 is installed in the work area 5, the approximate positions of the table center C and the position determination jigs 61 and 62 indicated by the robot coordinate system are known, and therefore the hand 14 can be made to approach the position determination jigs 61 and 62 without causing the robot 1 to collide with the wall surface of the work area 5 and the wall surface of the table 51.

Fig. 2 (a) shows a position determining jig 61 for bringing the hand 14 closer to one of the three directions. When the hand 14 is brought close to the position determination jig 61, the optical path 28 of the correlation sensor 25 is blocked by the position determination jig 61. At the moment when the optical path 28 is blocked during the movement of the hand 14, the optical path 28 coincides with a tangent line of a circle representing the outer periphery of the position determining jig 61 in the XY plane. The angles of the axes of the robot 1 at the moment the optical path 28 is blocked can be obtained from the outputs of the encoders connected to the motors 21 to 24, and the lengths of the arms 11 to 13 and the hand of the robot 1 are known, so that an equation of a tangent line of a circle represented by the position determination jig 61 in the XY plane can be obtained based on the case where the optical path 28 is blocked by the position determination jig 61. When the hand 14 is moved closer to the position determining jig 61 from three different directions, as shown in fig. 2 (b), three tangential lines L1 to L3 in the XY plane are obtained, and equations of the tangential lines L1 to L3 can be obtained. If the intersection of bisectors between these tangents, for example, the bisectors of tangents L1, L2, is taken as a straight line M1 and the bisectors of tangents L2, L3 are taken as a straight line M2, the intersection of the straight lines M1, M2 becomes the jig center O. Therefore, the accurate XY coordinate value of the jig center O of the position determination jig 61 can be obtained from the equation of the tangential lines L1 to L3. By applying the same procedure to the other position determining jig 62, the accurate XY coordinate values of the jig center O of the position determining jig 62 can be obtained, and the accurate positions of the position determining jigs 61, 62 in the robot coordinate system can be obtained. Since the accurate relative positional relationship between the table center C and the position determining jigs 61 and 62 is known, the accurate position of the table center C in the robot coordinate system can be determined using the accurately determined position determining jigs 61 and 62 as shown in fig. 2 (C). That is, the position of the table center C can be automatically taught.

The above description is of the case where the two position determining jigs 61, 62 are disposed away from the table center C, but if the position determining jigs 61 can be disposed at the table center C as shown in fig. 2 (d), the coordinates of the table center C can be obtained by determining the coordinates of the position determining jigs 61 in the robot coordinate system in the same manner as described above. In other words, in this case, the automatic teaching of the table center C can be performed using only one position determining jig 61 disposed in the table center C.

However, in the case of obtaining the coordinates of the jig center O, which is the position of the position determination jigs 61 and 62 in the robot coordinate system, as described above, due to various mechanical errors in the robot 1, a limited beam diameter of the optical path 28 of the correlation sensor 25, an influence of the ambient light, and the like, reproducibility may be reduced and the obtained coordinate values may deviate. This reduction in reproducibility means a large error in the coordinates of the jig center O, resulting in a teaching error in the teaching of the table center C. In order to reduce the error in the coordinate determination of the jig center O, it is conceivable to perform the process of approaching the hand 14 from three different directions to determine the coordinate of the jig center O a plurality of times while changing the moving direction of the hand 14, and to average the coordinate values obtained in each position determination. Therefore, in the present embodiment, the process of approaching the hand 14 from three different directions to determine the coordinates of the jig center O is performed a plurality of times, and the coordinates of the jig center O are determined on the basis of the average. Here, if the number of times of coordinate determination is set to N (where N is equal to or greater than 2), the hand 14 is brought close to the position determining tool 61 a total of 3N times, but in the present embodiment, the moving direction of the hand 14 in each time of position determination is selected so that the determination error of the coordinates of the tool center O becomes small. Fig. 3 is a schematic plan view illustrating a method of determining coordinates of the jig center O in the present embodiment. Here, a description will be given of a method of determining the coordinates of the jig center O of the position determining jig 61, and the coordinates of the jig center of the position determining jig 62 can be determined in the same manner.

As described above, there is a limit to the angular range that can be taken by the hand 14 when approaching the position determining jig 61. This angle range is an angle range toward the work area 5 with the jig center O as the center, and is denoted by θ in fig. 3 (a). The angle range θ is usually about several tens of degrees, regardless of the magnitude, smaller than 180 degrees. On the other hand, when the hand 14 is moved closer to the position determining jig 61 from three different directions to determine the position of the jig center O in the robot coordinate system, the accuracy of the coordinate determination improves when the three directions in which the hand 14 is moved closer are moved away from each other as much as possible. In addition, the angle transmission error of the speed reducer in the robot 1 is included in the factors of the error in determining the coordinates of the jig center O, and in order to reduce the influence of the angle transmission error of the speed reducer, it is preferable to cancel the influence of the error by approaching the hand 14 to the position determination jig 61 in as many directions as possible in the angle range θ as possible. In order to reduce the influence of other errors in determining the coordinates of the jig center O, it is also preferable to bring the hand 14 as much as possible toward the position determination jig 61 as possible.

In the present embodiment, 3N directions are set in the above-described angle range θ, considering a condition that the hand 14 is made to approach the position determining jig 61 from three directions different from each other as far as possible each time the position is determined, and a condition that the hand 14 is made to approach the position determining jig 61 as many directions as possible as a whole. Each direction is used for position determination only once. Preferably, these directions are arranged at equal angular intervals as viewed from the clamp center O. In fig. 3 (a), each set direction is indicated by an arrow, and each small circle indicates a position where the optical path 28 of the correlation sensor 25 of the hand 14 will come into contact with the outer periphery of the position determination jig 61 when the hand 14 is brought close to the position determination jig 61 in a direction corresponding to the small circle. In the figure, n=5 is shown, and the hand 14 approaches the position determining jig 61 from 15 (=3×5) directions. By making N larger than 5, the mutual angle in the direction in which the hand 14 approaches is made smaller, and the accuracy of determining the coordinates of the jig center O can be further improved.

In the example shown in fig. 3 (a), since three directions are used in one position determination, five position determinations can be made by using fifteen directions. Since it is preferable that three directions used for each position determination are as far as possible, the first, sixth, and eleventh directions are set as a first group, the second, seventh, and twelfth directions are set as a second group, and the fifth, tenth, and fifteenth directions are set as a fifth group from one end side of the angle range θ. Fig. 3 (b) shows how the 15 directions shown in fig. 3 (a) are combined into a group by hatching within a small circle. Then, the position of the jig center O is determined as coordinates for each group using three directions belonging to the group. Fig. 3 (c) is a diagram showing the position of the jig center O calculated for each group, and the position calculated for each group is indicated by a cross-shaped mark. The small circles indicate positions at which the outer periphery of the position determination jig 61 is detected by the optical paths 28 of the correlation sensors 25, as shown in fig. 3 (b). Therefore, although the small circle shown in fig. 3 (c) is a small circle of the robot coordinate system, the small circle represents a position determined by including the errors described above, and is deviated from the position of the outer periphery of the original position determination jig 61 represented by a thick line. In each position determination, the position of the jig center O is calculated from the position including the error, and thus the calculated position of the jig center O also deviates.

In the present embodiment, the final coordinates of the jig center O are determined by averaging the coordinates of the jig center O obtained by each position determination, and the coordinates of the table center C are calculated based on the final coordinates, thereby performing automatic teaching of the table center C. In this case, when there is a deviation value in the coordinates of the jig center O measured based on the positions of the respective times, it is preferable to calculate the average value of the coordinates of the jig center O by excluding the coordinates of the deviation value. As a method for detecting the offset value, for example, when an average value of coordinates of the jig center O measured based on each position is obtained, a variance σ of the coordinates is also obtained, and coordinate data having a distance of 3 σ or more from the calculated average value may be used as the offset value. In the drawing, a circle of a broken line indicates a range of 3σ, and in the example shown here, a position of the jig center O obtained by bringing the hand 14 close to the position determination jig 61 using the first group of three directions is outside the range of 3σ, and is handled as a deviation value. In this way, the final coordinates of the jig center O are determined by excluding the offset value, and the coordinates of the jig center O can be determined by excluding the group having a large error, so that the reproducibility of the coordinates of the jig center O can be improved, and the teaching error of the position of the table center C can be reduced.

Fig. 4 is a flowchart showing the above-described process of obtaining the coordinate values of the jig center O of the position determination jig 61 in the robot coordinate system. This process is performed by the robot control device 30. Firstly, in step 101, in the range of the angle range θ that the hand 14 can take when approaching the hand 14 to the position determining tool 61, for example, 3N directions are set at equal angular intervals (wherein N.gtoreq.2). Next, in step 102, in 3N directions, the coordinates of the jig center O in the robot coordinate system are calculated in the order described using FIG. 2, after the calculation of the jig center O for all of the N groups in the first direction and the (n+1) th direction, the second and the (n+2) th directions, and the (N) th directions and the (2) th directions are combined, and in each of the (N) th selective directions, a group consisting of three directions is set, for example, in the step 103, in such a manner that the hand 14 approaches the position determining jig 61 at equal angular intervals, in the order described using FIG. 2, the step 104, the average value of the jig center O is automatically calculated for all of the N groups, and in the step 104, in the order of description using FIG. 2, the average value of the jig center O is calculated, and then in the step 107, in the case that the average value of the average value is not calculated in the step 107, the step 107 is determined in the case that the average value of the average value is not calculated in the step 107, and the average value of the average value is calculated in the step 107 is calculated in the order of the step of the difference value is determined in the step 107, and in the case of the average value is not calculated in the step 107, and the average value is determined in the step of the difference value of the difference of the step 107, coordinates of the jig center O for automatic teaching are determined. The coordinates of the jig center O for automatic teaching can be determined by the same process for the other position determination jig 62, and the automatic teaching of the table center C can be performed using these coordinates.

According to the present embodiment described above, the coordinates of the jig center O can be obtained with good reproducibility, and the accuracy can be improved, whereby the teaching error with respect to the table center C can also be reduced. By improving the reproducibility and accuracy of the automatic teaching, the frequency of failure in placement of the workpiece 50 as the conveyance object or the frequency of re-performing the automatic teaching can be reduced, and the productivity of the entire system including the robot 1 can be improved.

In addition, the present technology may employ the following structure.

(1) An automatic teaching method for automatically teaching a position of a table center of a table connected to a work area via a connection portion as an opening to a robot having a hand and disposed in the work area,

Defining an XY coordinate system by setting a Y direction from the inside of the work area toward the center of the table via the connection portion, and an X direction perpendicular to the Y direction, and a robot coordinate system by setting the XY coordinate system, and a plane extending from the X direction and the Y direction as an XY plane,

The automatic teaching method comprises the following steps:

A step of disposing a columnar position determining jig at a position having a known relative positional relationship with the center of the table in the table;

A first setting step of setting N to be an integer of 2 or more, and determining 3N directions in the XY plane within an angle range that the hand can take, toward the position determining jig when the hand is moved into the table;

A second setting step of setting different N groups by selecting a direction while changing a starting point from one end side of the angle range, the group being composed of three directions selected for each N directions from among the 3N directions determined in the first setting step;

A position calculation step of bringing the hand close to the position determination jig from three directions included in the group for each of the groups set in the second setting step, detecting the position determination jig in a noncontact manner using a sensor provided on the hand, and calculating coordinates of the position determination jig in the robot coordinate system; and

An average calculation step of obtaining an average value of the coordinates obtained for all the groups,

The average value is set as coordinates of the position determining jig used for teaching of the robot.

(2) The automatic teaching method according to (1), wherein the average value is calculated by searching for a deviation value from the coordinates obtained in the position calculating step, and when the deviation value exists, excluding the deviation value in the average calculating step.

(3) The automatic teaching method according to (2), wherein in the average calculation step, variance of the coordinates obtained for all the groups is also calculated, and the determination of the deviation value is performed based on the variance.

(4) The automatic teaching method according to any one of (1) to (3), wherein in the first setting step, the 3N directions are determined at equal angular intervals as viewed from the position determining jig.

(5) The automatic teaching method according to any one of (1) to (4), the robot is a horizontal multi-joint robot, and the X direction and the Y direction are directions in a horizontal plane.

(6) The automatic teaching method according to (5), wherein in the position calculating step, the hand is moved in the horizontal plane.

(7) The automatic teaching method according to any one of (1) to (6), wherein the Y direction is a direction perpendicular to a wall surface of the work area at a position of the connecting portion.

(8) The automatic teaching method according to any one of (1) to (7), wherein the sensor is a correlation sensor.

(9) A robot control device for controlling a robot having a hand and disposed in a work area, the robot being automatically taught a position of a table center of a table connected to the work area via a connection portion as an opening,

When a direction from the inside of the work area toward the center of the table through the connection portion is defined as a Y direction and a direction perpendicular to the Y direction is defined as an X direction, and an XY coordinate system, which is a robot coordinate system, is defined as an XY plane, a plane extending from the X direction and the Y direction is defined as an XY plane, and a cylindrical position-determining jig is disposed in the table at a position having a known relative positional relationship with the center of the table,

When the hand is moved into the table, 3N directions are determined in the XY plane within an angle range which the hand can take, a group of three directions selected from the 3N directions for each N directions is set, a starting point is changed from one end side of the angle range, and different N groups are set by selecting directions,

For each set, the hand is brought close to the position-determining jig from three directions included in the set, the position-determining jig is detected in a noncontact manner by a sensor provided to the hand, and coordinates of the position-determining jig in the robot coordinate system are calculated,

The coordinates obtained for all the groups are averaged,

And teaching the robot by taking the average value as the coordinates of the position determining jig in the robot coordinate system.

Claims (9)

1. An automatic teaching method for automatically teaching a position of a table center of a table connected to a work area via a connection portion as an opening to a robot having a hand and disposed in the work area,

Defining an XY coordinate system by setting a Y direction from the inside of the work area toward the center of the table via the connection portion, and an X direction perpendicular to the Y direction, and a robot coordinate system by setting the XY coordinate system, and a plane extending from the X direction and the Y direction as an XY plane,

The automatic teaching method comprises the following steps:

A step of disposing a columnar position determining jig at a position having a known relative positional relationship with the center of the table in the table;

A first setting step of setting N to be an integer of 2 or more, when the hand is moved into the table, 3N directions are determined in the XY plane within an angle range that the hand can take, toward the position determining jig;

A second setting step of setting different N groups by selecting a direction while changing a starting point from one end side of the angle range, the second setting step being provided to form groups from three directions selected for each N of the 3N directions determined in the first setting step;

A position calculation step of bringing the hand into proximity with the position determination jig from three directions included in the group for each of the groups set in the second setting step, detecting the position determination jig in a noncontact manner by a sensor provided to the hand, and calculating coordinates of the position determination jig in the robot coordinate system; and

An average calculation step of obtaining an average value of the coordinates obtained for all the groups,

The average value is set as coordinates of the position determining jig used for teaching of the robot.

2. The automatic teaching method according to claim 1, characterized in that,

And searching for a deviation value from the coordinates obtained in the position calculating step, and if the deviation value exists, excluding the deviation value in the average calculating step to calculate the average value.

3. The automatic teaching method according to claim 2, characterized in that,

In the average calculation step, variance of the coordinates obtained for all the groups is also calculated, and determination of the deviation value is made based on the variance.

4. The automatic teaching method according to any of claims 1 to 3,

In the first setting step, the 3N directions are determined at equal angular intervals as viewed from the position determining jig.

5. The automatic teaching method according to any of claims 1 to 3,

The robot is a horizontal multi-joint robot, and the X-direction and the Y-direction are both directions in a horizontal plane.

6. The automatic teaching method according to claim 5, characterized in that,

In the position calculating step, the hand is moved in the horizontal plane.

7. The automatic teaching method according to any of claims 1 to 3,

The Y direction is a direction perpendicular to a wall surface of the work area at a position of the connection portion.

8. The automatic teaching method according to any of claims 1 to 3,

The sensor is a correlation sensor.

9. A robot control device for controlling a robot having a hand and disposed in a work area, the robot being automatically taught a position of a table center of a table connected to the work area via a connection portion as an opening,

When a direction from the inside of the work area toward the center of the table through the connection portion is defined as a Y direction, and a direction perpendicular to the Y direction is defined as an X direction, and an XY coordinate system, which is a robot coordinate system, is defined, a plane extending from the X direction and the Y direction is defined as an XY plane, a cylindrical position-determining jig is disposed at a position having a known positional relationship with respect to the center of the table in the inside of the table,

When the hand is moved into the table, 3N directions are determined in the XY plane within an angle range which the hand can take, a group of three directions selected from the 3N directions for each N direction is set, a starting point is changed from one end side of the angle range, and different N groups are set by selecting directions,

For each set, the hand is brought close to the position-determining jig from three directions included in the set, the position-determining jig is detected in a noncontact manner by a sensor provided to the hand, and coordinates of the position-determining jig in the robot coordinate system are calculated,

The coordinates obtained for all of the groups are averaged,

And teaching the robot by taking the average value as the coordinates of the position determining jig in the robot coordinate system.