CN209533449U - Virtual box - Google Patents

Abstract

A kind of virtual box uses in the teaching operation that training data generates system, even if simplifying the teaching operation of robot, when being loaded into glass substrate on hand and moving in box, is also easy to prevent the interference of the structure inside glass substrate and box on hand.The virtual box of the utility model uses in the teaching operation that training data generates system, it has the multiple substrate placement sections for placing glass substrate, and be formed as shape identical with the box that can store divided glass substrate, and, virtual box has detection agency, it, which is used to measure, moves in virtual box by the hand of horizontal articulated robot and the glass substrate of upside that substrate placement section is configured at before being placed in substrate placement section, and lifted and be configured at from substrate placement section by hand substrate placement section upside glass substrate in the glass substrate of either one horizontal direction position and height, and has the sensor of the height for measuring glass substrate, the sensor is the displacement sensor of mirror surface reflecting type.

Description

Virtual box

Technical field

The utility model relates to the training datas in the training data for generating horizontal articulated robot to generate system Teaching operation used in virtual box.

Background technique

Have the robot to the box moving glass substrate that can store divided glass substrate, control machine at present it is known that having The robot teaching system of the robot controller of people, the information processing terminal being connect with robot controller and teaching operation terminal It unites (for example, referring to patent document 1).In the robot teaching system documented by patent document 1, robot is for example to two sections The mode of overlapping is configured to four box moving glass substrates of the similar shape of two column.A box in four boxes is being set as benchmark Box, by remaining three boxes be set as expansion box when, in the robot teaching system, use benchmark box carry out robot teaching Operation.

In the robot teaching system recorded in patent document 1, teaching operation of the robot for benchmark box is being carried out When, operator acts robot successively to the specified position relative to benchmark box by teaching operation terminal, determines each teaching Position obtains the data of each teaching position.Specifically, in the robot teaching system, in the machine for being mounted with glass substrate Ready position, hand are inserted into benchmark box and enter and place glass substrate before storage before the hand of device people is mobile into benchmark box The storage ready position of posture on the stringer inside benchmark box, subordinate drop complete the receipts being placed on glass substrate on stringer Receive the position that retracts that down position, hand be pulled back to outside benchmark box, operator moves robot successively by teaching operation terminal Make, determines each teaching position, obtain the data of each teaching position.At this point, operator confirms robot by visual observation, and make machine Device people successively acts to specified position.

In the robot teaching system recorded in patent document 1, according to the data of each teaching position of acquirement to benchmark Training data, that is, benchmark training data of box generation robot.In addition, being based on three expansion boxes in the robot teaching system Each relative to benchmark box relative position data and benchmark training data, each that three are unfolded with boxes is automatic raw Training data is unfolded at the training data of robot.That is, in the robot teaching system, even if being carried out without using expansion box The teaching operation of robot can also automatically generate the expansion training data of each that three are unfolded with box.

Existing technical literature

Patent document

Patent document 1: Japanese Unexamined Patent Publication 2006-123157 bulletin

Utility model content

Utility model technical problem to be solved

In the robot teaching system recorded in patent document 1, for example, being loaded in order to prevent when in hand insertion box The interference of the structures such as the stringer inside glass substrate and box on hand preferably generates and is set as showing by best storage ready position Teach the benchmark training data of position.

On the other hand, in the robot teaching system recorded in patent document 1, since operator confirms machine by visual observation Device people, while making robot motion to storage ready position, therefore best storage ready position is set as teaching position in order to generate The benchmark training data set, it is necessary to confirm robot, while the position of fine tuned robot by visual observation, make robot motion to most Good storage ready position.Therefore, in the case where the robot teaching system, it is being loaded into glass substrate on hand in order to prevent When moving in box, the structure inside glass substrate and box on hand is interfered, and the teaching operation of robot may become very numerous It is trivial.

Therefore, it is the technical issues of the utility model, a kind of virtual box is provided, generates the teaching of system in training data It uses in operation, even if simplifying the teaching operation for carrying out glass substrate to the horizontal articulated robot of the carrying of box, is loading When the glass substrate on hand of horizontal articulated robot moves in box, also it is easy to prevent in glass substrate and box on hand The interference of the structure in portion.

Technical solution used by solving the problems, such as

In order to solve the above-mentioned technical problem, the utility model provides a kind of virtual box, generates showing for system in training data It is used in religion operation, the training data generates the training data that system is used to generate horizontal articulated robot, the level Articulated robot is to the similar shape that can store the divided glass substrate configured with the state of interval in the up-down direction Multiple boxes carry out the carrying of the glass substrate, wherein there are the virtual box the multiple substrates for placing the glass substrate to put Portion is set, and is formed as shape identical with the box, also, the virtual box has detection agency, the detection agency is used for Measurement is moved in the virtual box by the hand of the horizontal articulated robot and is matched before being placed in the substrate placement section It is placed in the glass substrate of the upside of the substrate placement section and is lifted and be configured at from the substrate placement section by the hand The position of the horizontal direction of the glass substrate of either one in the glass substrate of the upside of the substrate placement section It sets and height, the detection agency has the sensor of the height for measuring the glass substrate, the sensor is mirror surface The displacement sensor of reflection-type.

In the present invention, be formed as shape identical with box and in the teaching operation of horizontal articulated robot The virtual box used has detection agency, is used to measure and moves in virtual box by the hand of horizontal articulated robot, and It is configured at the glass substrate of the upside of substrate placement section before being placed in substrate placement section and is lifted from substrate placement section by hand And be configured at substrate placement section upside glass substrate at least one glass substrate horizontal direction position and height.

Therefore, in the present invention, even if not having when carrying out the teaching operation of horizontal articulated robot to virtual box There is the position of fine tuned robot, also can generate based on the detection result of detection agency and configure the hand for being mounted with glass substrate Optimum position in the position of the upside of substrate placement section is set as the benchmark training data of teaching position.That is, practical new at this In type, even if there is no the position of fine tuned robot when carrying out the teaching operation of horizontal articulated robot, can also generate by When being loaded into the glass substrate on hand of horizontal articulated robot and being moved in box, the structure inside glass substrate and box on hand Benchmark training data of the optimum position that the divine force that created the universe does not interfere as teaching position.Therefore, by based on each of multiple boxes Data, that is, the position data and benchmark training data of a relative position relative to virtual box, to each generation water of multiple boxes The training data of flat articulated robot, even if not having fine tuned robot when carrying out the teaching operation of horizontal articulated robot Position, also can be based on the structure inside the glass substrate and box that will be loaded into when glass substrate on hand moves in box on hand Benchmark training data of the optimum position that the divine force that created the universe does not interfere as teaching position generates training data.Therefore, practical at this In novel, even if simplifying the teaching operation of horizontal articulated robot, it is loaded into the glass on hand of horizontal articulated robot When substrate moves in box, also it is easy to prevent the interference of the structure inside glass substrate and box on hand.

In addition, in the present invention, detection agency has the sensor of the height for measuring glass substrate, the sensing Device is the displacement sensor of mirror surface reflecting type, therefore, detecting distance is enough long, can detect and virtual box configures along the vertical direction The corresponding distance of the spacing of adjacent bearing part；Also, the displacement sensor of mirror surface reflecting type is particularly suitable for detecting glass base Transparent object as plate；Also, even if test object tilts, the displacement sensor of mirror surface reflecting type also can be examined steadily It surveys.

In addition, in the present invention, the sensor may be configured to measure the glass substrate in a manner of range of triangle Height.

In addition, in the present invention, the sensor is for example installed on the frame of the virtual box by installing component. In such a situation it is preferred to which the installing component is the plate of stainless steel.When constituting in this way, it is avoided that installing component itself produces It is raw to vibrate and bend and range determination result is caused to change.

In the present invention, it is generally desirable to, detection agency has the glass substrate for being configured at and being formed as oblong-shaped The corresponding position in quadrangle four sensors.When constituting in this way, for example, even if glass substrate is enlarged, loads Bending or distortion are generated in glass substrate on hand or hand generates bending, as a result, the one end of glass substrate on hand Height and glass substrate another side height generate difference, also can accurately measure glass base by four sensors The height of plate.Therefore, it can generate the glass substrate that will be loaded on hand based on the detection result of detection agency and be moved in box When glass substrate and box on hand inside benchmark teaching number of the better place that does not interfere of structure as teaching position According to.

In the present invention, detection agency is for example set to upper end side and the two positions of lower end side of virtual box.? In this case, at two positions of the upper end side of virtual box and lower end side, horizontal articulated robot is carried out to virtual box Teaching operation.

(utility model effect)

As described above, in the present invention, even if simplifying the more passes of level for carrying out carrying of the glass substrate relative to box The teaching operation of section robot also can when the glass substrate on hand for being loaded into horizontal articulated robot moves in box Enough prevent the interference of the structure inside glass substrate and box on hand.

Detailed description of the invention

Fig. 1 is for illustrating that the training data comprising the utility model embodiment generates the knot of the robot system of system The block diagram of structure.

Fig. 2 is the top view for indicating horizontal articulated robot shown in FIG. 1, box and virtual box.

Fig. 3 is the top view of virtual box shown in Fig. 2.

Fig. 4 is the main view of virtual box shown in Fig. 3.

Fig. 5 (A) is the figure of the structure and configuration for illustrating first sensor shown in Fig. 3, and (B) is for explanatory diagram 3 Shown in the structure of second sensor and the figure of configuration.

Fig. 6 (A) is the figure for illustrating the structure of 3rd sensor shown in Fig. 3, and Fig. 6 (B) is for shown in explanatory diagram 3 3rd sensor mounting structure figure, Fig. 6 (C) is for the figure for the light mode that illustrates to diffuse, and Fig. 6 (D) is to be used for Illustrate that the figure of specular light light mode, Fig. 6 (E) are for illustrating to measure base in the way of range of triangle by 3rd sensor The figure of the case where height of plate.

Glass substrate when Fig. 7 is position and the height for illustrating detection agency measurement horizontal direction as shown in Figure 3 State figure.

Fig. 8 (A) is the figure for illustrating another mounting structure of 3rd sensor, and Fig. 8 (B) is to be schematically illustrated in Fig. 8 (A) Mounting structure used in the first plate figure, Fig. 8 (C) is schematically illustrated in second used in the mounting structure of Fig. 8 (A) The figure of plate.

(description of symbols)

2 substrates

3 robots

4 robot control units

6 boxes

7 virtual boxes

11 hands

23 training datas generate system

24 substrate placement sections

28 detection agencies

29 first detection agencies

30 second detection agencies

31 sensors

32 sensors

33 sensors

331 semiconductor lasers

332 projection lens

333 sensitive lens

334 light receiving elements

40 installing components

X left and right directions

Y front-rear direction

Specific embodiment

Hereinafter, being described with reference to the embodiments of the present invention.

(structure of robot system)

Fig. 1 is for illustrating that the training data comprising the utility model embodiment generates the robot system 1 of system 23 Structure block diagram.Fig. 2 is the top view for indicating horizontal articulated robot 3 shown in FIG. 1, box 6 and virtual box 7.

The robot system 1 of the method has the glass substrate 2 (hereinafter referred to as " substrate 2 ") for carrying liquid crystal display Horizontal articulated robot 3 (hereinafter referred to as " robot 3 "), the robot control unit 4 for controlling robot 3 are controlled with robot Teaching operation terminal (teaching box) 5 that portion 4 is electrically connected with what the state of interval configured stores multi-disc in the up-down direction Multiple boxes 6 of the similar shape of substrate 2, the virtual box 7 used in the teaching operation of robot 3.

Substrate 2 is formed as oblong-shaped.The substrate 2 of the method is relatively large substrate, for example, the width of substrate 2 is 2900mm or so, the vertical width of substrate 2 are 3300mm or so.In addition, for example, substrate 2 with a thickness of 0.4mm~0.7mm or so.

Robot 3 carries out carrying of the substrate 2 relative to multiple boxes 6.As shown in Fig. 2, robot 3, which has, loads substrate 2 Two hands 11, front end sides link two arms 12 of two hands 11 respectively, the body part 13 for supporting two arms 12, in the horizontal direction may be used The movably matrix part 14 in support body portion 13.In the following description, the shifting by body part 13 relative to matrix part 14 Dynamic direction (X-direction of Fig. 2 etc.) is set as left and right directions, and the Y-direction of Fig. 2 orthogonal with up and down direction and left and right directions etc. is set as Front-rear direction.

Body part 13 has the base end side of supporting arm 12 and liftable arm support 15, liftably bearing arm support 15 Scaffold 16, constitute the end portion of body part 13 and relative to the horizontally movable base station 17 of matrix part 14, solid Determine the lower end of scaffold 16 and relative to the rotatable slewing frame 18 of base station 17.

Arm 12 is made of the first arm and the second arm the two arms.The base end side and body part 13 of arm 12 are rotationally Connection.Hand 11 is rotationally linked in the front end side of arm 12.Arm 12 can be flexible to horizontal direction, so that hand 11 is in direction one Substantial linear is mobile in the state of determining direction.Specifically, arm 12 can be flexible to horizontal direction, make to go smoothly 11 towards centainly The state in direction, and the linking portion substantial linear of hand 11 and arm 12 is mobile.

Scaffold 16, which has, liftably to be kept columnar first scaffold 20 of arm support 15 and liftably protects Hold columnar second scaffold 21 of the first scaffold 20.The lower end of second scaffold 21 is fixed on slewing frame 18 Front end side.The base end side of slewing frame 18 is supported on base station 17, is made it possible to and is carried out taking up and down direction as rotation Axial rotation.Robot 3 is by the lifting action of expanding-contracting action, the arm 12 of arm 12 etc., rotational action and moves horizontally movement Combination carry out handling substrate 2.

Robot control unit 4 is constituted such as the information processing unit by robot controller, personal computer.Robot control Device processed has the servo control portion of the various motor of control robot 1.In addition, robot control unit 4 has the operations list such as CPU The input-output unit of the input and output of the storage units such as member, ROM and RAM, progress data.As be described hereinafter, robot control unit 4 is raw At the training data of robot 3.The robot control unit 4 of the method is that the training data for the training data for generating robot 3 is raw At portion.In addition, in this mode, the training data for generating robot 3 is made of robot control unit 4 and virtual box 7 etc. Training data generate system 23.

Box 6 is formed as the cuboid box-like of one end side opening in the front-back direction.Multiple boxes 6 be configured in the lateral direction that This is adjacent.In addition, multiple boxes 6 are configured at the side of robot 3 in the longitudinal direction.3 side of opening portion object manipulator of box 6. As described above, substrate 2 is formed as oblong-shaped.The longitudinal direction for being accommodated in the substrate 2 of box 6 is consistent with front-rear direction, is accommodated in The short side direction of the substrate 2 of box 6 is consistent with left and right directions.

When moving in substrate 2 to box 6 and when moving out the substrate 2 for being accommodated in box 6, the linear movement in the longitudinal direction of hand 11, It is inserted into box 6.In the following description, the side (the Y1 direction side of Fig. 2 etc.) that robot 3 is configured in front-rear direction is set as Side (the Y2 direction side of Fig. 2 etc.) configured with box 6 is set as " rear " side by " preceding " side.

As described above, can store in box 6 with the multi-piece substrate 2 of the state configuration of interval in the up-down direction.Box 6 have the multiple substrate placement sections 24 for placing substrate 2.Multiple substrate placement sections 24 are arranged in the up-down direction with constant pitch Column.Substrate placement section 24 has from multiple bearing parts 25 in the left and right ends portion of downside supporting substrates 2 and in the lateral direction Specified position from multiple bearing parts 26 of downside supporting substrates 2.

Bearing part 25 is formed as using left and right directions as the rodlike of length direction.Bearing part 25 is from a left side for the framework of box 6 Right both ends are stretched out towards the inside of left and right directions.In addition, multiple bearing parts 25 to separate constant interval in the longitudinal direction State arrangement.Bearing part 26 is formed as using front-rear direction as the rodlike of length direction.Bearing part 26 is from the framework of box 6 Rear end is stretched out towards front side.In addition, multiple bearing parts 25 are arranged with the state for separating constant interval in the lateral direction.This Outside, box 6 does not have the detection agency 28 that will be described later.

Virtual box 7 is formed as shape identical with box 6.Hereinafter, illustrating the structure of virtual box 7.In addition, as described above, empty Quasi- box 7 uses in the teaching operation of robot 3.When carrying out the teaching operation of robot 3, by a box 6 in multiple boxes 6 It is replaced into virtual box 7.Virtual box 7 is arranged at position identical with the position of a box 6 to be replaced is provided with.Scheme in Fig. 2 State when the teaching operation for carrying out robot 3 is shown.

(structure of virtual box)

Fig. 3 is the top view of virtual box 7 shown in Fig. 2.Fig. 4 is the main view of virtual box 7 shown in Fig. 3.Fig. 5 (A) is For illustrating that the structure of sensor 31 shown in Fig. 3 and the figure of configuration, Fig. 5 (B) are for illustrating sensor 32 shown in Fig. 3 Structure and the figure of configuration.Fig. 6 (A) is the figure for illustrating the structure of sensor 33 shown in Fig. 3, and Fig. 6 (B) is for explanatory diagram The figure of the mounting structure of 3rd sensor shown in 3, Fig. 6 (C) are the figures for the light mode that illustrates to diffuse, and Fig. 6 (D) is For illustrating that the figure of specular light light mode, Fig. 6 (E) are for illustrating to survey in the way of range of triangle by 3rd sensor The figure for the case where determining the height of substrate.Fig. 7 be for illustrate detection agency 28 as shown in Figure 3 measure horizontal direction position and The figure of the state of substrate 2 when height.

As described above, virtual box 7 is formed as shape identical with box 6, can store with interval in the up-down direction The multi-piece substrate 2 of state configuration.In addition, virtual box 7 has the multiple substrate placement sections 24 for placing substrate 2.Substrate placement section 24 Has bearing part 25,26.In addition, as described above, virtual box 7 is arranged at and sets when carrying out the teaching operation of robot 3 It is equipped with the identical position in position of a box 6 being replaced.

When carrying out the teaching operation of robot 3, hand 11 is inserted into virtual box 7, carries out the moving in virtual box 7 of substrate 2 Movement and substrate 2 from virtual box 7 move out movement at least one movement.Movement is acted and moved out carrying out moving in for substrate 2 When, hand 11 is moved linearly by along the longitudinal direction.That is, front-rear direction (Y-direction) is carrying-in/carrying-out of the substrate 2 relative to virtual box 7 Direction.

Virtual box 7 has detection agency 28, is used to measure and is moving in virtual box 7 by hand 11 and be placed in substrate placement It is configured at the substrate 2 of the upside of substrate placement section 24 before portion 24 and is lifted from substrate placement section 24 by hand 11 and is configured at base The position of the horizontal direction of at least one substrate 2 (referring to Fig. 7) of the substrate 2 of the upside of plate placement section 24 and height.Detection The first detection agency 29 and the height for measuring substrate 2 that mechanism 28 has the position of the horizontal direction for measuring substrate 2 The second detection agency 30.

In addition, detection agency 28 is set at the upper end side and lower end side two of virtual box 7.That is, virtual box 7 has two spies Survey mechanism 28.The detection agency 28 for being configured at the upper end side of virtual box 7 measures the upper of the substrate placement section 24 for being configured at uppermost The position of the horizontal direction of the substrate 2 of side and height.The measurement of detection agency 28 for being configured at the lower end side of virtual box 7 is configured at most The position of the horizontal direction of the substrate 2 of the upside of the substrate placement section 24 of lower section and height.

Before first detection agency 29 has the sensor 31 and measurement substrate 2 of the position of the left and right directions of measurement substrate 2 The sensor 32 of the position of rear direction.Sensor 31,32 is that have the transmissive optical formula of light-emitting component and light receiving element sensing Device.The light-emitting component and light receiving element of sensor 31,32 are arranged opposite in the up-down direction with the state for separating specified interval.Separately Outside, sensor 31,32 is the linear transducer for having the multiple light receiving elements being arranged in a straight line.First detection agency of the method 29 have two sensors 31 and a sensor 32.Sensor 31,32 is electrically connected with robot control unit 4.The biography of the method Sensor 31 is first sensor, and sensor 32 is second sensor.

Two sensors 31 are configured at the one end of the left and right directions of virtual box 7.In addition, two sensors 31 are in front and back The state configuration of interval on direction.Specifically, two sensors 31 are configured at the both ends in the front-back direction of virtual box 7 Portion.As shown in Fig. 5 (A), the length direction (arrangement of light receiving element of sensor 31 is configured to for the sensor 31 of linear transducer Direction) it is consistent with left and right directions.Sensor 31 detects the position of an end face of the left and right directions of substrate 2 by light receiving element to survey Determine the position of the left and right directions of substrate 2.

Sensor 32 is configured at the rear end of virtual box 7.In addition, sensor 32 is configured at the virtual box 7 on left and right directions Approximate centre position.As shown in Fig. 5 (B), for linear transducer sensor 32 be configured to sensor 32 length direction and Front-rear direction is consistent.Sensor 32 measures the front and back of substrate 2 by the position of the rear end face by light receiving element detection substrate 2 To position.

Second detection agency 30 has four sensors 33.Four sensors 33 are electrically connected with robot control unit 4.Four Sensor 33 is configured in position corresponding with the quadrangle of substrate 2 of oblong-shaped is formed as respectively.That is, four sensors 33 are respectively arranged on four angles of virtual box 7.The sensor 33 of the method is 3rd sensor.

Herein, as shown in Fig. 6 (B), sensor 33 is installed on installing component 40 (such as mounting plate) using screw etc., installation The installing component 40 of sensor 33 is installed on the frame of virtual box 7 by screw etc..In this mode, installing component 40 is stainless The metal plate of steel.That is, the rigidity of installing component 40 is larger, therefore, be avoided that 40 Self generating oscillations of installing component and It bends and range determination result is caused to change.Also, in this mode, installing component 40 is in relatively thin shape, therefore, energy Installing component 40 and the substrate 2 for the bearing part 25 for being loaded into virtual box 7 is avoided to generate interference.

Sensor 33 is the optical sensor with the reflection-type of light-emitting component and light receiving element.In addition, sensor 33 It is so-called range sensor (displacement sensor).The light-emitting component of sensor 33 is penetrated towards the lower surface of substrate 2 or upper surface Light out.For example, the light-emitting component of sensor 33 projects laser.Sensor 33 is for example according to by the lower surface or upper surface of substrate 2 The light of reflection is incident on the position of the light receiving element of sensor 33 to measure the height of substrate 2.

Displacement sensor is there are two types of light mode, i.e., from the body part BT of displacement sensor towards object as shown in Fig. 6 (C) Light mode that what DX irradiation light and receiving reflect from object DX diffused diffuse and as Fig. 6 (D) is shown from displacement The specular light light of the body part BT of sensor specular light from object DX towards object DX irradiation light and reception Mode.

In this mode, sensor 33 uses the displacement sensor of mirror surface reflecting type, and sensor 33 is configured to when looking down It is Chong Die with substrate 2, to measure at a distance from the lower surface of substrate 2 (or upper surface).Also, as shown in Fig. 6 (B), sensor 33 can detect distance corresponding with the spacing P of adjacent bearing part 25 of virtual box 7 configured along the vertical direction, that is to say, that Substrate 2 is loaded into the hand 11 of robot 3 with above-mentioned spacing P and is carried.

In addition, sensor 33 for example measures the height of substrate 2 in the way of range of triangle.Specifically, such as Fig. 6 (E) institute Show, sensor 33 is used with flowering structure: by irradiating substrate 2 using laser beam (irradiation light in the utility model) L, can export It can determine the signal S of the distance between the surface 2a of substrate 2 (the measure object face in the utility model) D.More specifically, As shown in Fig. 6 (E), sensor 33 includes semiconductor laser 331, projection lens 332, sensitive lens 333 and light receiving element 334.Semiconductor laser 331 issues laser, and the laser that semiconductor laser 331 issues is converted into having by projection lens 332 The laser beam of predetermined spot diameter, and irradiate substrate 2.Sensitive lens 333 collects the reflected light R of the laser beam reflected by substrate 2. Light receiving element 334 is made of such as PSD (position sensitive detector) or CCD (charge-coupled device), is received by 333 meeting of sensitive lens Poly- reflected light R, and export detection signal S.

(the teaching operation of robot and the generation method of training data)

When carrying out the teaching operation of robot 3 and generating the training data of robot 3, virtual box 7 is used.That is, into The teaching operation of row robot 3 and when generating the training data of robot 3 a, as described above, box 6 is replaced into virtual box 7. Virtual box 7 is arranged at position identical with the position of a box 6 to be replaced is provided with.

When carrying out the teaching operation of robot 3, the operator of robot 3 while confirming robot 3 by visual observation, Robot 3 is operated by teaching operation terminal 5.In addition, when carrying out the teaching operation of robot 3, operator is in order to determine phase For the regulation teaching position of virtual box 7, act robot 3 successively to specified position relative to virtual box 7.In the method In, the teaching operation of robot 3 is carried out at two positions of upper end side and lower end side of virtual box 7.

In addition, in this mode, virtual box 7 will be moved to by hand 11 and be placed in the base before substrate placement section 24 Plate 2 is configured in the position of the upside of substrate placement section 24 and is configured at base from substrate placement section 24 by the substrate 2 that hand 11 lifts The substrate 2 for being loaded into hand 11 (is configured at substrate to place by at least one party of the position (referring to Fig. 7) of the upside of plate placement section 24 The position of the upside in portion 24) also as the teaching position of robot 3.

Specifically, the substrate 2 being loaded on hand 11 is configured at the position of the upside of the substrate placement section 24 of uppermost, And it is all showing for robot 3 that the substrate 2 being loaded on hand 11 is configured to the position of the upside of the substrate placement section 24 of lowermost Teach position.In order to determine the teaching position, operator passes through teaching operation terminal 5 while confirming robot 3 by visual observation Act robot 3, until the position.The position is arrived in the movement of robot 3, and the substrate 2 being loaded on hand 11 is configured When the upside of substrate placement section 24, position and the height of the horizontal direction of substrate 2 are measured by detection agency 28.

The measurement result of detection agency 28 is entered robot control unit 4.Robot control unit 4 is based on detection agency 28 Measurement result determines teaching position.In addition, to the position of 4 input operator of robot control unit substrate 2 and robot 3 by visual observation Set and determine other teaching positions data.Robot control unit 4 is determined according to the measurement result based on detection agency 28 The data of teaching position and the other teaching positions for being entered robot control unit 4 generate teaching of the robot 3 to virtual box 7 Data, that is, benchmark training data.That is, robot control unit 4 generates benchmark training data based on the measurement result of detection agency 28.

In this mode, each opposite position relative to virtual box 7 of multiple boxes 6 is stored in robot control unit 4 The data set i.e. position data.That is, in robot control unit 4, be stored with remaining multiple boxes 6 each relative in machine The data i.e. position data of the relative position for the box 6 replaced when the teaching operation of people 3.Robot control unit 4 is generating base After quasi- training data, it is based on benchmark training data and position data, generates robot 3 to the teaching number of each of multiple boxes 6 According to.

That is, in this mode, robot 3 is automatically generated relative to the training data of each of multiple boxes 6, without The teaching operation of robot 3 is carried out using box 6.In addition, benchmark training data is directly as robot 3 to showing in robot 3 Teach the training data for the box 6 replaced when operation.

(main effect of the method)

As described above, in this mode, virtual box 7 has detection agency 28, is used for measurement and is loaded on hand 11 simultaneously It is configured at position and the height of the horizontal direction of the substrate 2 of the upside of substrate placement section 24.In addition, in this mode, robot Control unit 4 generates training data, that is, benchmark teaching number of the robot 3 relative to virtual box 7 based on the measurement result of detection agency 28 According to.

Therefore, in this mode, when carrying out the teaching operation of robot 3 to virtual box 7, even if operator does not finely tune The position of robot 3 also can be generated the hand 11 for being mounted with substrate 2 being configured at base based on the detection result of detection agency 28 Benchmark training data of the optimum position as teaching position in the position of the upside of plate placement section 24.That is, in this mode, i.e., Make there is no the position of fine tuned robot 3 when carrying out the teaching operation of robot 3, can also generate will be loaded into the substrate of hand 11 The optimum position that the structure of the inside of substrate 2 and box 6 on 2 hands 11 when moving in box 6 does not interfere is as teaching position The benchmark training data set.

In addition, in this mode, each based on multiple boxes 6 is ascended the throne relative to the data of the relative position of virtual box 7 Data and benchmark training data are set, to the training data of each generation robot 3 of multiple boxes 6.Therefore, in this mode, It, also can be based on will be loaded on hand 11 even if there is no the position of fine tuned robot 3 when carrying out the teaching operation of robot 3 The optimum position that the structure of the inside of substrate 2 and box 6 when substrate 2 moves in box 6 on hand 11 does not interfere, which is used as, to be shown The benchmark training data of position is taught, training data is generated.Therefore, in this mode, even if simplifying the teaching operation of robot 3, When the substrate 2 being loaded on hand 11 moves in box 6, the structure of the inside of the substrate 2 and box 6 on hand 11 can be also prevented Interference.

In addition, in this mode, detection agency 28 has the sensor 33 of the height for measuring substrate 2, the sensor 33 be the displacement sensor of mirror surface reflecting type, and therefore, detecting distance is enough long, can be detected and the configuration along the vertical direction of virtual box 7 Adjacent bearing part 25 the corresponding distance of spacing；Also, the displacement sensor of mirror surface reflecting type is particularly suitable for detecting glass Transparent object as glass substrate；Also, even if test object tilt, the displacement sensor of mirror surface reflecting type also can steadily into Row detection.

In this mode, sensor 33 is installed on the frame of virtual box 7, also, installing component 40 by installing component 40 It is the plate of stainless steel, therefore, is avoided that 40 Self generating oscillations of installing component and flexure and range determination result is caused to be sent out Changing.

In this mode, the first detection agency 29 has with two biographies of the state configuration of interval in the longitudinal direction Sensor 31.Therefore, in this mode, it is able to use two sensors 31 and measures substrate 2 relative to gradient in the front-back direction.Cause This can be generated based on the detection result of detection agency 28 in this mode and be moved the substrate 2 being loaded on hand 11 in box 6 The better place that the structure of the inside of substrate 2 and box 6 when dynamic on hand 11 does not interfere is shown as the benchmark of teaching position Teach data.

In this mode, four sensors 33 be respectively arranged at and be formed as the substrate 2 of oblong-shaped quadrangle it is right respectively The position answered.Therefore, in this mode, even if 2 enlargement of substrate, the substrate 2 being loaded on hand 11 generates bending or distortion, or Person's hand 11 generates bending, as a result, the height of the height of the front end side of the substrate 2 on hand 11 and rear end side generates difference, or The height of the one end of the left and right of substrate 2 on hand 11 and the height of another side generate difference, can also utilize four sensings Device 33 accurately measures the height of substrate 2.Therefore, in this mode, can be based on the detection result of detection agency 28, generating will What the structure of the inside of substrate 2 and box 6 when the substrate 2 being loaded on hand 11 moves in box 6 on hand 11 did not interfered Benchmark training data of the better place as teaching position.

(other embodiment)

Above-mentioned mode is an example of the best mode of the utility model, but not limited to this, it is practical not changing Various modifications implementation can be carried out in the range of novel objective.

In above-mentioned mode, the quantity for the sensor 31 that the first detection agency 29 has is also possible to one.This In the case of, for example, a sensor 31 is configured at the approximate centre position of the virtual box 7 on front-rear direction.In addition, being visited first In the case that the quantity for the sensor 31 that survey mechanism 29 has is one, the first detection agency 29 can also have in right and left Two sensors 32 of the state configuration at spaced up interval.In which case it is possible to use two sensors 32 measure base Gradient of the plate 2 relative to left and right directions.That is, in this case, being also able to use two sensors 32 and measuring substrate 2 relatively In gradient in the front-back direction.

In above-mentioned mode, the quantity for the sensor 33 that the second detection agency 30 has is also possible to three or less.Separately Outside, in above-mentioned mode, detection agency 28 can only be configured at the upper end side of virtual box 7, can also only be configured at virtual box 7 Lower end side.In addition, the position or more other than the upper end side and lower end side of virtual box 7 also can be set in detection agency 28 A position.

In above-mentioned mode, when carrying out the teaching operation of robot 3, virtual box 7 also can be set to be wanted in setting The different position in the position of one box 6 of displacement.In addition, can also be configured more than two boxes 6 in above-mentioned mode It is overlapped in the up-down direction.In addition, robot 3 is also possible to carrying the use other than liquid crystal display in above-mentioned mode The robot of glass substrate 2 used in way.

In above-mentioned mode, virtual box 7 and box 6 are had the same shape and dimensions, the machine carried to substrate 2 The hand 11 of people 3 is of similar shape, therefore, be able to confirm that on hand 11 there are the detection position of the substrate 2 when substrate 2 and The flexure and gradient of substrate 2；Based on this, by implementing sunykatuib analysis in advance, and the inclination of substrate 2 is followed according to its result Sensor 33 is set in an inclined manner, can steadily be detected.As specific structure, for example, as Fig. 8 (A), Fig. 8 (B) and Shown in Fig. 8 (C), the first plate 51 and the second plate 52 of L-shaped can be used, wherein the first plate 51 has cross one another First plate portion 511 and the second plate portion 512, and the first plate portion 511 has a screw hole and a longer long hole of short transverse, the Two plates 52 have cross one another first plate portion 521 and the second plate portion 522, and there are two screw holes for the first plate portion 521 tool, and And first plate 51 sensor 33 is fixed on by screw (not shown), installation is fixed in the second plate portion 522 of the second plate 52 Component 40, screw (not shown) pass through the screw hole of the second plate 52 and the long hole of the first plate 51, exist as a result, by adjusting screw The gradient of sensor 33 can be adjusted in position in above-mentioned long hole.

Claims (7)

1. a kind of virtual box uses in the teaching operation that training data generates system, the training data generates system and is used for Generate the training data of horizontal articulated robot, the horizontal articulated robot is to can store to separate in the up-down direction Multiple boxes of the similar shape of the divided glass substrate of the state configuration at interval carry out the carrying of the glass substrate, and feature exists In,

The virtual box has the multiple substrate placement sections for placing the glass substrate, and is formed as shape identical with the box Shape, also, the virtual box has detection agency, and the detection agency is for measuring by the hand of the horizontal articulated robot It moves in the virtual box and is configured at before being placed in the substrate placement section described in the upside of the substrate placement section Glass substrate and lifted from the substrate placement section by the hand and be configured at the substrate placement section upside the glass The position of the horizontal direction of the glass substrate of either one in substrate and height,

The detection agency has the sensor of the height for measuring the glass substrate,

The sensor is the displacement sensor of mirror surface reflecting type.

2. virtual box according to claim 1, which is characterized in that

The sensor is configured to measure the height of the glass substrate in a manner of range of triangle.

3. virtual box according to claim 1, which is characterized in that

The sensor is installed on the frame of the virtual box by installing component.

4. virtual box according to claim 3, which is characterized in that

The installing component is the plate of stainless steel.

5. virtual box according to any one of claim 1 to 4, which is characterized in that

The detection agency, which has, is configured at position corresponding with the quadrangle of the glass substrate of oblong-shaped is formed as Four sensors.

6. virtual box according to any one of claim 1 to 4, which is characterized in that

The detection agency is set to upper end side and the two positions of lower end side of the virtual box.

7. virtual box according to claim 5, which is characterized in that

The detection agency is set to upper end side and the two positions of lower end side of the virtual box.