JPS60263682A - Selector for follow-up position - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Field of Application J) This invention relates to a follow-up position selection device in a follow-up device in which a movable base on which an automatic working machine such as a robot is mounted follows a conveyor. be.

[Conventional Technology J] In recent years, for example, in automobile manufacturing factories, automatic working machines such as industrial robots and specialized machines (automatic screw tightening machines, etc.) have been developed to perform various assembly tasks that were traditionally performed by workers. There are many attempts to make devices do this.

By the way, such automatic working machines generally move the workpiece positioned at a predetermined position within the work area.
Since the work is to be taught in advance or to perform a specific work, it is preferable to carry the work into and out of the work area using a tact conveyor (shuttle conveyor).

However, although it is desirable, most of the conveyors installed on the current lines that perform various assembly operations are continuous conveyors such as power and free conveyors and slat conveyors. However, considering the huge equipment costs required to change from a continuous conveyor to a tact conveyor, it is recommended to continuously change the working mode of automatic work equipment instead of using a tact conveyor. It is clear that it would be better to devise a way to adapt it to the assemble conveyor.

One of the ways to do this is to use the following well-known method that takes advantage of the versatility of the robot's movements.

In other words, by taking advantage of the fact that the robot can move freely within its range of motion, by synchronizing the movements of the robot and the continuous conveyor for a certain period of time, the robot itself can follow the conveyor. This is a method of getting work done.

However, although this method is suitable for very simple big-and-place work such as placing a work on a conveyor or taking a work out from a conveyor, it is hardly suitable for various types of assembly work that involve complex movements. .

This is because the work trajectory becomes very complicated in such a task, and when the work trajectory becomes complicated, the synchronization control of the robot's conhair becomes very complicated, and the time required for one task also increases. This is because, depending on the relationship between the speed of the conveyor and the movable range of the robot, the robot may not be able to complete the work within its followable range.

I did. The above method also has a problem in that it cannot be applied to special-purpose machines whose movements are not versatile.

A possible solution to this problem is to place the automatic working machine on a cart that can travel along the conveyance direction of a continuous conveyor, and to mechanically connect the cart and the conveyor for a certain distance. In this method, the automatic work machine follows the conveyor until the work is completed.

However, in order to adopt such a method, not only would it be necessary to develop an attachment/detachment device to mechanically connect and separate the two, and install multiple stages on the existing conveyor, but the conveyor There are many issues that need to be solved, such as the need to take measures to reduce the load on the vehicle and to cushion the shock at the start of tracking. □ Therefore, the applicant first detected the relative positional deviation in the conveyor conveyance direction between the follow-up position on the conveyor side and the reference position on the movable base side on which an automatic working machine device such as a robot is mounted, while sequentially A tracking method in which the movable base follows the conveyor by controlling the travel of the movable base so that the detected relative position deviation becomes zero (Japanese Patent Application No. 59-778
No. 46) and developed a tracking device applying this method to solve the above-mentioned problems.

[Problems to be Solved by the Invention] However, in the tracking device to which the above tracking method is applied, the movable base is made to follow one tracking position determined on the conveyor side.

Even when performing one operation or multiple operations on a workpiece transported by a conveyor, there is no problem as long as these operations can be performed without changing the relative positional relationship between the robot and the workpiece. However, when multiple tasks cannot be performed without changing the relative positional relationship between the robot and the workpiece, such as when assembling a front windshield glass and a rear window glass onto a car body, in other words, the robot's There is a problem in that it cannot be used when the movable range cannot cover a wide work area where multiple tasks can be performed.

This invention aims to provide the technology necessary to solve such problems.

Means for solving the L problem J Therefore, in the follow-up position selection device according to the present invention,
In the above-mentioned tracking device, a dogleg-shaped reflector formed by making at least one of the two reflecting plates as a back reflecting plate and making both reflecting plates orthogonal to each other is used to control a plurality of predetermined tracking units on the conveyor side. At the same time, the thickness of the back reflector is changed for each follow-up position, and the light is irradiated from the movable base side to the dogleg-shaped reflector. Based on the reflected light, The movable base is configured to include a selection means for selecting a follow-up position on the conveyor side to be followed by the movable base.

5. Effect] By using the following position selection device configured as described above, by changing the following position to be followed by the movable base, the movable base can be made to follow any of the plurality of following positions on the conveyor side. become.

(Embodiment J Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

FIG. 1 is a perspective view of the structure around a wind crow assembly work stage to which a follow-up position selection device according to the present invention is applied.

In the figure, 1 is a slat conveyor, and the slat 1a is driven in the direction of arrow Y at a predetermined speed (for example, 2.2 m/win) by a chain not shown in the figure.

On the slats 1a of this slat conveyor 1, jigs 2 for mounting vehicle bodies are attached at predetermined intervals.
The vehicle body 3 is placed at a predetermined position on the jig 2 in a predetermined sitting posture, and the vehicle body 3 is sequentially transported in the direction of arrow Y at a predetermined transport speed.

Each jig 2 on the slat 1a has two predetermined follow-up positions corresponding to the front side and rear side of the vehicle body 6, respectively, as shown enlarged in FIGS. 2 and 3. The dogleg-shaped reflectors 6 and 7 are attached via brackets 4 and 5.

Among these, the reflectors have a plate thickness second to that of the front reflector 6a.
As shown in the figure, eleven back reflectors 6b are formed to be perpendicular to each other, and the mounting is detailed as follows.

As shown in FIG.
The mating portion 6c of the slat conveyor 1 is perpendicular to the direction of arrow Y, which is the conveyance direction of the slat conveyor 1, and as shown in FIG. slit light Ll emitted from the slit light source 20 (FIG. 1).

A tracking position (hereinafter referred to as [first follow-up position]).

In this embodiment, it is assumed that the mating portions 6c of the two reflecting plates Fya and 6b and the slit light Ll are directed in the vertical direction.

On the other hand, the reflector 7 is formed by perpendicularly forming a front reflecting plate 7a and a back reflecting plate 7b having a thickness of 2 (t2>tI) as shown in FIG. It is installed like this.

That is, as shown in FIG.
The slit light emitted from the slit light source 20 on the movable base 11 side, which will be described later, is directed in the direction in which the joint portion 7c of b is perpendicular to the Y direction, and as shown in FIG. L] and.

A tracking position (hereinafter referred to as " It is attached at a second follow-up position.

In this embodiment, it is assumed that the mating portion 7c of the two reflecting plates 7a and 7b also faces in the vertical direction.

Returning to FIG. 1, 8 is a follow-up unit installed along the direction of arrow Y, which is the conveyance direction of the slat conveyor 1. Base 11 is indicated by arrow y,,Y2
It's supposed to make it run in the direction you want it to go.

The travel range of the movable base 11 in the follow-up unit 8 is determined in consideration of the time required for the window glass assembly work by the robot 16, which will be described later.

Further, the origin position of the movable base 11 is near the running end on the motor 10 side.

On this movable base 11, there is a wrist shaft 13 at the tip.
A multi-joint type robot 13 as an automatic working machine device with a suction type hand 12 attached is placed and fixed on a.

This robot 13 uses its suction type hand 12 to pick up rear crows 16 that are sequentially and intermittently supplied by wind crow supply devices 14 and 15 placed at the illustrated position near the origin of the movable base 11. and the front windshield 17 one by one, move the adsorbed windshield to an adhesive applicator (not shown), apply adhesive to the periphery of the windshield, and then close the window at a predetermined timing to be described later. Work is done to assemble the crow to the car body.

Next, the post 18 is attached to the robot 16 on the movable base 11.
A follow-up position detector consisting of a line image sensor 25, a drive detection circuit 26, etc., which will be described later, and a slit light source 20 can be attached to the upper end of this post 18 as follows. be.

That is, slit light I,+ (FIG. 4) directed vertically from the slit light source 20 is irradiated in a direction perpendicular to the direction of arrow Y, which is the conveying direction of the slat conveyor 1, and this slit light L1 forms a dogleg. The tracking position detector 1B and the optical axis of the slit light source 20 are spaced at a predetermined distance so that the reflected slit light L2 reflected by the shaped reflector 6 or 7 is incident on the tracking position detector 1B. They are installed parallel to each other, with both optical axes perpendicular to the Y direction.

Note that the line image sensor 25 in the tracking position detector 1 is arranged so that its longitudinal direction is parallel to the arrow Y direction, as shown in FIG. The center position of the section is set as the reference position on the movable base 11 side.

In FIG. 1, 21 is a follow-up control unit, which controls the rotation of the motor 10 of the follow-up unit 8 to perform follow-up control of the movable base 11, and performs the follow-up position selection function according to the present invention. do.

Further, 22 is a robot control device for controlling the robot 13, which includes a follow-up position detector 1, a slit light source 20. It also includes a power supply device that supplies power to the follow-up control unit 21 and the like.

Next, the configuration of the control system will be explained with reference to FIGS. 5 to 9.

First, in FIG. 5, the follow-up control unit 21 includes a central processing unit (CPU)-238, and a read-only memory (block memory IJ) [ROM] 23b.

Random access memory (data memory) [R, AM]
23c, and a data processing microcomputer 23 consisting of an input/output device <L, IO) 23d, etc.

It is configured by a follow-up controller 24.

Between the microcomputer 23 and the robot control device 22, a picking work start command SS and an assembly work start command S82 as robot operation commands from the microcomputer 23, a picking work completion signal SE from the robot control device 22, , assembly work completion signal SE2.
Signals are exchanged with the total work completion signal SEE and the total work completion signal SEE.

The microcomputer 23 also issues a follow-up control start command SS, a follow-up control stop command ST.

and controller 2 that follows the movable base return command SR.
4, clock start/stop command SS4. Detection start command SS5. and the detection stop command S' is sent to the drive detection circuit 26 that controls the line image sensor 25 in the position detector 19, and the spot light source 2 sends the light on/off command SL.
Output to 0 respectively.

The microcomputer 23 also receives a tracking start timing signal Sc and numerical data ('Wl (described later)) from the drive detection circuit 26, and an origin detection signal that is turned on when the movable base 11 in FIG. 1 returns to the origin position. The origin signal So from the limit switch 27 is input.

The follow-up controller 24 includes a microcomputer 23
Follow-up control start command SS, follow-up control stop command ST
, and in addition to the movable base return command SR, the relative position deviation datum 9 from the drive detection circuit w126;, the final bit output PE from the line image sensor 25, and the tiger fish signal So from the origin detection limit switch 27, FIG. The runaway signal EM from the runaway prevention limit switch 28 (normally, the work is completed before this switch) is turned on when the movable base 11 travels near the running end in the direction of arrow Y1, and the output of the motor 10. A speed feedback signal Sv from the tacho generator 9 attached to the shaft, and a reference speed signal S according to the conveyance speed of the slat conveyor 1 in FIG. 1 from the tacho generator 2 for line speed detection.
v are respectively input.

For example, as shown in FIG. 6, the tacho generator for line speed detection is a speed increaser consisting of a sprocket 30 that meshes with the drive chain CH of the slat conveyor 1, a gear train g1 to g4 connected to the sprocket 30, etc. The slat conveyor 1 is rotated at an increased speed via the slat conveyor 1, thereby outputting a reference speed signal S'v corresponding to the conveying speed of the slat conveyor 1.

Next, in FIG. 7, the follow-up controller 24 controls the gain adjustment amplifier 31. Latch circuit 32° D/A converter 33
．． Gain adjustment amplifier 34. Addition circuit 65. Analog switch '56. Amplifier 37° adder circuit 38. amp 3 days,
Set/reset type flip-flop circuit (FF) 4
0,42°OR circuit 41.43. Return circuit 44. and an analog switch 45.

Then, the gain adjustment amplifier 31 adjusts the gain of the reference speed signal Sv from the tacho generator 29 shown in FIGS. 5 and 6 and outputs it.

The latch circuit 32 latches the relative position deviation data Δτ (sometimes positive and sometimes negative) from the drive detection circuit 26 at the timing whenever the final bit output PE is output from the line image sensor 25.

The D/A converter 66 converts the relative position error datum latched in the latch circuit 32 into an analog relative position error signal ΔS, x, and the gain adjustment amplifier 34 converts the relative position error datum into analog relative position error signals ΔS and x. The obtained relative position error signal ΔSx is gain adjusted and output.

The adder circuit 35 adds the reference speed signal Sv and the relative position error signal ΔSx, which are respectively gain-adjusted and output from the gain adjustment amplifiers 31 and 34, and outputs the addition result as an addition signal SF.

Note that the relative position deviation data ΔX latched at the latch time 1'132 has positive and negative times.

When Δwork is positive, S F > S v, and when it is negative, S
F<Sv.

The analog switch 36 is turned FF by a follow-up control start command S S g from the microcomputer 23 in FIG.
40 is set, it turns on and outputs the addition signal SF to the amplifier 37, and F F 4 'Q is inputted via the OR circuit 4.1 from the microcomputer 23 to detect follow-up control stop command ST or runaway detection. When reset by the runaway signal EM from the limit switch 28, the amplifier 37 turns off and cuts off the addition signal SF. Amplify quantitatively.

The addition circuit 38 receives the amplified addition signal S from the amplifier 37.
A deviation signal ΔS based on the deviation between the F or return signal SR and the speed feedback signal Sv from the tacho generator 9 attached to the output shaft of the motor 10 is output, and the amplifier 3S amplifies the deviation signal ΔS and outputs the deviation signal ΔS to the motor 10. Output.

The FF 42 receives a movable base return command SR from the microcomputer 23 input via an OR circuit 43 or a runaway signal EM from the runaway detection limit switch 28.
is set by the origin detection limit switch 27.
It is reset by the origin signal SO from .

The return circuit 44 operates to open the FF 42 when the Q output is -1" and outputs a return signal SB that rotates the motor 10 in the direction in which the movable base 11 returns to the origin. 6 The analog switch 45 outputs the Q output of the FF 42 is on for a period of 1'', during which time the return signal SR output from the return circuit 44 is output to the amplifier 67.

The line image sensor 25 of the tracking position detector 19 is composed of, for example, a MOS type or CCD type image sensor with 2048 pixels (bits), and internally scans the video output ■S according to the light received by the light receiving section 25'a. The circuit outputs the signals sequentially.

The drive detection circuit 26 of the follow-up position detector 1st includes a clock generator 46. Counter 47. First binarization circuit 4th, set/reset type flip-prop circuit (FF) 49.
AND circuits 50, 52. Comparison circuit 51. Latch circuit 53
．． OR circuit 54. Subtraction circuit 55. Second binarization circuit 56
．． Set/reset type toggle type flip-flop circuit (FF) S7. D-type flip-flop circuit (FF
) 58, AND circuits 59, 60. and a counter 61.

The clock generator 46 operates when the clock start/stop command 5S-4 from the microcomputer 23 is a clock start (for example, 1''), and converts the clock signal Cp of a predetermined frequency (for example, 2.5 MHz) into a line image. Output to the clock input terminal of sensor 25, SS
It stops when 4 reaches a clock stop (eg 0').

The counter 47 counts the clock signal cp from the clock generator 46 and detects the scanning pixel position in the line image sensor 25.
Every time the final hit output PE is output from , the count value -Nx is reset by the output PE.

Note that this counter 47 is a binary counter with a capacity of at least 12 bits, and is r2048J (800H).
It is assumed that the number can be counted.

The first binarization circuit 48 converts the video output vS for each pixel sequentially outputted from the line image sensor 25 in synchronization with the clock signal cp to a first slice level (2
(Varization level) Vr+[Compared with FIG. 10 or FIG. 11 J, video output V S h'NV S ≧V r
((1) Outputs the binarized videotator Dv1 which becomes O'' when Vs<Vrr.

The FF 49 is set by a detection start command S S s from the microcomputer 23 and reset by a detection stop command Sr from the microcomputer 26 .

The AND circuit 50 is configured such that the Q output of the FF 49 is 1'' and the binarized Hinao output DvI from the first binarized circuit 48 is
The output is set to 1 only when it is also 1'.

The comparison circuit 51 compares the count value Nx of the counter 47 with the center pixel position of the line image sensor 25 (movable head 1
r 400 HJ (1024,
), and the output is set to l'' only when N x = 400 H.

The AND circuit 52 outputs 1'' only when the output of the comparator circuit 51 is 1'' and the binary hiteo output DvI from the first binarization circuit 48 is also 1''.

Note that the output of this AND circuit 52 is 1''.

It is output to the microcomputer 23 as a tracking start timing signal Sc.

The latch circuit 53 receives A input via the OR circuit 54.
The output of the ND circuit 50 or 52 changes from "o" to '1'
The count value Nx of the counter 47 at the time when it rises to ' is latched.

The subtraction circuit 55 subtracts r400 HJ indicating the center pixel position of the line image sensor 25 described above from the count value Nx latched in the latch circuit S3, and uses the subtraction result as relative position deviation data ΔX in the tracking controller 24. It is output to the latch circuit 32.

The second binarization circuit 56 converts the video output S for each pixel sequentially output from the line image sensor 25 in synchronization with the clock signal Cp to a second slice level (2
digitization level) VF6 (compared with J shown in FIG. 10 or 11, video output V S fJXV S';?-V
r 2 (7) Time d ” L ” te V S < V
Binarized videotator D v which becomes 0゛′ when r 2
Outputs 2.

The FF 57 is reset to 1 to 1 by the final bit output PE from the line image sensor 25 every time the final bit output PE is output, and the AND circuit 59 resets the Q output of the FF 58 to 1'.
, the Q output is inverted every time the binarization videotator D v 2 from the second binarization circuit 56 rises from 0'' to 1''.

It is assumed that the set terminal S of this FF 57 is set to 0''.

The FF 58 of the network ink is also cleared by the final bit output PE every time it is output, and when the Q output of the FF 57 falls from "i" to 0, it is input to the D terminal. Read the current l″.

Therefore, when the Q output of FF58 is cleared by the final bit output PE and becomes 1'', the A'ND circuit 5 opens, and the binary video data D v 2 input to the torque terminal T of FF57 becomes 2. When the Q output falls from 1'i' to '0', the Q output of FF513 also falls to 1' or 'o', thereby causing the AND circuit to is closed and then FF
58 is again cleared by the final bit output PE.

The AND circuit 60 generates a clock only while the Q output of the FF 57 is 1'', that is, from the first rise to the next rise of the binarized video data Dv2 from the second binarization circuit 56. clock signal c from the device 46
Let p pass.

The counter 61 is reset by the first bit output PE each time it is output, and counts the number of pulses of the clock signal cp passing through the AND circuit 60.

Therefore, the numerical value [WJ, which is the count value of this counter 61, is the period of Dv2 from the first falling edge to the second rising edge in the binarized video data Dv2, in other words, the pixel position corresponding to the first rising edge, 2
It represents the distance from the pixel position corresponding to the falling edge.

Next, before explaining the operation of this embodiment, the behavior of the slat conveyor 1 shown in FIG. 1 during operation will be specifically explained.

This slat conveyor 1 appears to be moving at a constant speed due to the two-chain drive, but in reality, the conveyance speed varies depending on various factors such as the spring constant of the chains and the friction coefficient of the movable parts, as shown in Fig. 8. The speed fluctuates wildly over time.

Therefore, if the movable base 11 in the follow-up unit 8 is made to follow the first or second follow-up position on the slat conveyor 1 side at a set speed, the robot 16 cannot accurately perform the wind crow profiting work.

Therefore, in this embodiment, the reference speed signal Sv from the tacho generator 2, which approximately corresponds to the speed of the slat conveyor 1, is successively corrected using the relative position deviation data detected by the follow-up position detector 19. It attempts to deal with speed fluctuations. I will discuss this matter later.

Hereinafter, the operation of this embodiment will be explained with reference to the drawings from FIG. 9 onwards.

First, it is assumed that the control systems such as the follow-up control unit 21 and the robot control device 22 shown in FIGS. 1 and 5 are powered on when the slat conveyor 1 is started, and that the first It is assumed that the movable base 11 and robot 13 shown in the figure have returned to their original positions.

Microcomputer 2 in follow-up control unit 21
When the power is turned on and the CPU 23a of No. 3 is started, it performs a predetermined initialization process, and then performs drive detection of the tracking position detector 1S at 5TEP l in the block diagram processing flow shown in FIG. A clock start/stop command SS4 indicating a clock start is outputted to the circuit 26, and a lighting/lighting out command SL indicating lighting is outputted to the slit light source 20.

As a result, the clock generator 46 shown in FIG. 7 starts outputting the clock signal cp, and the slit light source 20 shown in FIGS. 1 and 5 lights up to turn on the slit light L+ (
4), and CP 'U 23 a is 5TIEP in FIG. 9.
After outputting the pinking work start command SS to the robot control device 22 at step 2, the robot waits for the picking work completion signal SEL to be input from the robot control device 22 at the next step 5rEP3.

The robot control device 22 receives a picking operation start command S S l from the follow-up control unit 21 (to begin with, it picks up the front wind crow 17 from the win i crow supply device 15 and picks up the picked front wind crow 17 in the figure). After carrying the device to an adhesive applicator and applying adhesive on its periphery, the operator performs a picking operation by transporting the device to a predetermined standby position, and when this operation is completed, the follow-up control unit 21 performs the picking operation. A completion signal SE is output.

In addition, FIG. 1 shows a state in which after the work of assembling the front window glass 17 to one vehicle body 3 has been completed, the rear window glass 16 is about to be assembled to the vehicle body B.

Pinking work completion signal SE from robot control device 22
, is input, the CPU 23a is 5T in FIG.
Proceed to process from EP 3 to 5TEP 4, and ROM23
b or the numerical data (W x J) as described below stored in the RAM 23C are read out as follows.

First, regarding the numerical data [W x J, for example, as shown in Fig. 10, a slit light source 20 (Figs.
When the slit light L1 is irradiated from the front reflector 6a, the slit light L1 is reflected by the front reflector 6a and then reflected by the back reflector 6 with a thickness tl.
A part of it is reflected on the surface of b as reflected slit light L2+, and most of the slit light that is refracted and incident on the inside of the reflector 1 is back-reflected on the back surface. Slit light L2 reflected from the surface of 6b
The reflected slit light L is emitted as 2, and the reflected slit light L is gradually decreased based on the remaining slit light.3. ... White is also emitted from the surface of the back reflector 6b.

Therefore, these reflected slit lights L2++L22
+ L 23 + . It is possible to obtain a video output vs in which a small peak appears in the distribution shown in the figure.
When it is binarized by the second binarization circuit 56 (FIG. 7) which binarizes with r2.

After all, a binarized videotator Dv2 as shown in the figure can be obtained.

Similarly, as shown in FIG. 11, when the slit light L1 is irradiated onto the dogleg-shaped reflector 7 attached to the second follow-up position, the reflection reflected or emitted from the back reflector 7b having a thickness tl. Slit light L21 rL 22 + L 2
3 +... is input to the light receiving section 25a of the line image sensor 25, the video output vS as shown in the figure
It is possible to obtain binarized video data D v 2 corresponding to .

Now, when we compare the binarized video data Dv2 shown in Fig. 10 and Fig. 11, we find that the cycle from the first rise to the next rise of both data is LW + J
In other words, the distance between the incident pixel position of the reflected slit light L2+ and the incident pixel position of the reflected slit light L22 depends on the plate thicknesses i1+i2 of the back reflectors 6b and 7b of the doglegged reflectors 6 and 7. By using numerical data indicating this period or distance (WI J (W2 J), it is possible to identify the doglegged reflectors 6 and 7, respectively.
In the first place. A second follow-up position can be identified.

Therefore, the above numerical data [WI J LW2 J to R
Store it in OM23b or RAM23c.

In 5TEP 4 of FIG. 9, its ROM23b or RA
From M23C to [WIJ
WI J [W2 J (WI J (W2 J...
The data are read out alternately in the order of...

That is, the first execution special number of this 5TEP 4; read numerical data [WxJ is 9, which corresponds to, for example, the first follow-up position (W+J).

Note that 5 TEP 4 may be replaced by a process in which required numerical data (W x J) is sequentially received from the robot control device 22 or the sequence controller (PC) that manages the line, for example.

Next, the CPU 2a at 5TEP5 operates the drive detection circuit 26.
It waits for the tracking start timing signal Sc to be input.

The slit light L1 from the slit light source 20 reaches the front part of the vehicle body B sent in the direction of arrow Y by the slat conveyor 1 in FIG. It becomes incident on the surface reflection plate 6a of the dogleg-shaped reflector 6 attached to the following position, and
Reflected slit light L 21 J L 22 r L 23+ from the back reflector 6b based on the slit light L1.
... becomes incident on the light receiving section 25a of the line image sensor 25 of the tracking position detector 1st, sequential scanning is performed based on the clock signal cp output from the clock generator 46 in FIG. line image sensor 2
From 5, reflected slit light L 2+ + L during scanning
22 + L 23 + ... Video output where the signal level is low in pixels where the reflected slit light is not incident, and the signal level is high in pixels where the reflected slit light is incident, depending on the incident pixel position. (2) S is output in time series for each scan.

Then, the dogleg-shaped reflector 6 moves in the direction of the arrow Y, and L22 of the reflected slit light is directed to the reference position (400H) at the center of the light receiving part 2Sa of the line image sensor 25, as shown in FIG. When the line image sensor 25 outputs the video output ■S as shown in the figure, the slice level V r 1 (Vr
t > Vr2) to binarize the video output vs.
The digitizing circuit 48 outputs the binarized video data DvH for one scan shown in the figure.

Then, when such binary video data Dv1 is output, the clock signal cp from the clock generator 46 is scanned by one scan (2048 pulses worth of clock signal c
The count value Nx of the counter 47 in FIG.
When the output of the comparison circuit 51 becomes 1'', A
The output of the ND circuit 52 changes from O' to "1", and this
i'' is output as the tracking start timing signal Sc to the microcomputer 26 in FIG.
The x is latched.

Returning to FIG. 9, the CPU 23a is connected to A of the drive detection circuit 26.
When the tracking start timing signal Sc is received from the ND circuit 52, the process proceeds from 5TEP 5 to 5TEP 6, and numerical data (WJ) is taken in from the counter 61 in the drive detection circuit 26 in FIG.

Then, in the first 5TIEP 7, [WJ imported in 5TIEP 6 and L W x read out in 5TEP 4]
J and check whether LWJ = LWxJ or not, [WJ≠LWxJ is this the follow-up position of the conveyor side that is currently passing in front of the reference position on the movable base 11 side? Assuming that it is not the tracking position that the movable base 11 should follow, return to 5TEP 5 and wait for the tracking start timing signal Sc to be input from the AND circuit 52 of the drive detection circuit 26 again, and if [WJ=[WxJ] For example, it is assumed that the follow-up position on the conveyor side that is currently passing in front of the reference position on the movable base 11 side is the follow-up position to be followed by the movable base 11, and the process proceeds to 5TEP 8.

At first, [WxJ read out with 5TEP 4 is [W+J
So, [Wl taken in at 5TEP 6 is also (W1), so the following position on the conveyor side during one pass is the movable base 1
Assuming that 1 is the first tracking position to be followed first, 5
Proceed to TEP8.

In this way, the slit light is irradiated from the movable base 11 side to the dogleg-shaped reflector, and the numerical data based on the reflected slit light [WJ] determines the tracking position on the conveyor side that the movable base 11 should follow. Selected.

In addition, in 5TEP 7, instead of checking whether [WJ = [WxJ], it is more practical to check whether I [WJ - (Wxi l≦α (α is the tolerance). Highly desirable.

Next, at 5TEP8, the CP LT 23a outputs a follow-up control start command S S HH to the follow-up controller 24, and also outputs a detection start command Is S to the drive detection circuit 2nd day.
Outputs 5.

After that, the process advances to 5TEP9, and after performing the tailing process of waiting for a certain period of time, the next 5TEPIO
An assembly work start command S82 is output to the robot control device 22.

Then, at the next 5 TEPII, the CPU 23a waits for the set (=1 work completion signal SE2) to be input from the robot control device 22.

In addition, for Tayley processing in 5TEP9, moveable base 11
This is done in order to wait for the relative positional relationship between the robot 13 and the car body 3 on the slat conveyor 1 to stabilize, and if it is stabilized while the robot 13 moves to the starting position for assembling the wind crow, This process can be omitted.

In the process at 5TEP 8, the follow-up control start command S S3 is output to the follow-up controller 24, and the detection start command S
85 is output to the drive detection circuit 26 of the follow-up position detector 19, both control the movement of the movable base 11 as follows, and move the robot 13 to the follow-up position on the slat conveyor 1 side, in this case the cylinder 1. Follow the tracking position.

That is, first, when the follow-up control start command SS4 is input to the follow-up controller 24, the FF 40 in FIG. 7 is set and the analog switch 36 is turned on.

At this point, the analog switch 45 is turned off because the FF 42 remains reset by the origin signal SO from the origin detection limit switch 27 when the movable base 11 returned to the origin last time.

Then, when the analog switch 36 is turned on.

Initially, the reflected slit light L22 is incident on the reference position on the movable base 11 side, which is the center position of the line image sensor 25 in the tracking position detector 19, and the relative position deviation signal ΔS,7': is zero (this In particular, as will be described later), the reference speed signal Sv whose gain has been adjusted by the gain adjustment amplifier 31 from the tachogenerator 2nd is outputted from the addition circuit 35 to the amplifier 37 as the addition signal SF.

Therefore, on the third day of the amplifier, the addition signal 3F (=S v',) amplified by the amplifier 37 from the addition circuit 38 and the speed feedback signal S from the tacho generator 9 are combined.
Deviation signal ΔS based on the deviation from υ (however, initially zero)
(ΔS − S v ) is amplified and output to the motor 10, so the motor 10 starts rotating so that the movable base 11 travels in the direction of arrow Y at approximately the same speed as the conveyance speed of the slat conveyor 1. .

On the other hand, when the detection start command S85 is input to the drive detection circuit 26 of the follow-up position detector IS, the FF 49 in FIG. The binarized videotator Dv1 that is sequentially output from the gate is inputted to the latch circuit 53 via the AND circuit 50 and the OR circuit 54, so that the binarized video data Dvt (see FIG. 10) becomes 0''. From” 1
The count value Nx of the counter 47 is latched by the latch circuit 53 every time the counter 47 rises.

Therefore, the movable base 1 whose first follow-up position on the slat conveyor 1 side is the center position of the line image sensor 25
At the time when the reference position on the 1 side is reached and the reflected slit light L22 is incident on the reference position, as described above, r
The count value Nx of 400 HJ is latched in the latch circuit 5'5, but the movable base 11 begins to run as described above and reaches the center position 1F (4oon) of the line image sensor 25, which is the reference position on the movable base 11 side. and conveyor 1
When a slight relative positional deviation in the direction of the arrow Y in FIG. 1 based on the speed fluctuation shown in FIG. 8 begins to occur between the line image sensor 25 and the first tracking position on the Since the position of incidence on the line image sensor 25 shifts, the latch circuit 53 latches a count value Nx indicating the position on the line image sensor 25 corresponding to the first tracking position every time the line image sensor 25 scans.

Therefore, the relative position deviation data ΔX calculated by the subtraction circuit 55 is zero at the beginning, and thereafter is Nx-400H (positive or negative This relative position deviation data ΔX is latched into the latch circuit 32 of the tracking controller 24 at the timing of the final bit output PE output from the line image sensor 25 every time one scan is completed. be done.

Then, relative position deviation data ΔX is input to this latch circuit 32.
If the data ΔX is latched every time one pedal stroke is completed, this data ΔX becomes velocity dimension data, and this velocity dimension-based data Δ
After D/A conversion of It is possible to obtain a speed dimension-based relative position error signal Δs which can be corrected to the signal shown in FIG.

Therefore, this relative position deviation signal Δs2:! is input into the addition circuit 35 to form an addition signal 5F=Sv+Δf3yc.The motor 10 rotates based on this addition signal SF, and the movable base 11 is moved between the reference position and the follow-up position (first follow-up position). ) so that the relative positional deviation in the direction of arrow Y becomes zero, that is, the relative positional relationship between the robot 13 on the movable base 11 and the vehicle body 6 on the slat conveyor 1 is always constant. It will now follow the slat conveyor 1 to the first follow-up position.

Note that the line image sensor 25 has the function 2.
Since scanning is performed using the 5 MHz tarokk signal CP (1 scanning time: 819.2 μsec), even if the conveyance speed of the slat conveyor 1 fluctuates as shown in FIG. 8, it can be adequately coped with.

However, if the sensor is continuously scanned with a 2.5 MHz clock signal cp, there is a risk that sufficient charge accumulation will not be achieved when using a CCD type image sensor. It is necessary to repeat the next scan after a pause of 8μ5ec). In this case, the update time of data ΔX is, for example, 2m5ec.
However, this can also adequately cope with the above-mentioned fluctuations.

In this way, the tracking controller 24 and the tracking position detector 1
If the movable base 11 is made to accurately follow the flow jL of the car body 6 from the time when the first follow-up position on the slat conveyor 1 side reaches the above-mentioned reference position, the ninth
51'lEP in the figure] The robot control device 22, which received the assembly work start command SS2 outputted at O, moves the robot 13 on the movable base 111-, which was waiting while holding the front windshield crow 17 by suction, to a predetermined position. By braking at the right timing, it is possible to correctly assemble the front winch 1 and the sword lath 17 to the vehicle body 6 flowing in the direction of arrow Y, as shown in FIG. 1. Next, returning to FIG. 9, the robot 16 When the front windshield crow 17 is completed and the robot control device 22 outputs an assembly work completion signal SE2 to the microcomputer 26 of the follow-up control unit 21, the CPU 2
3a advances the processing from 5TEP l 1 to 5TEP l 2 and sends the follow-up control stop command ST to the follow-up controller 24.
Along with outputting to.

A detection stop command Sr is output to the drive detection circuit 26 of the follow-up position detector 19.

As a result, FF4Q in FIG. 7 is reset via the OR circuit 41 and the analog switch 36 is turned off, so the addition signal SF is cut off, the motor 10 stops, and the movable base 11 also stops.

Further, since the FF 49 is also reset and the AND circuit 50 is closed, the new count value Nx is no longer latched in the latch circuit 53.

However, if the follow-up control stop command ST is not output for some reason, the movable base 11 continues to run, but if it overruns to a certain extent, the runaway detection limit switch 28 (Fig. 5) turns on and the runaway signal EM is activated. Output.

Then, FF4Q is reset by this runaway signal EM, and the movement of the movable base 11 is stopped, and FF4Q is reset.
2 is set. Note that when the FF 42 is set, the movable base 11 returns toward the origin as described later.

Further, when the runaway signal EM is output, an abnormality indicator (not shown) may be activated to notify maintenance personnel of the abnormality, and further work may be temporarily halted.

Returning to FIG. S, after the CPO 23a completes the processing of 5TEP 1 2, it proceeds to 5TEP 1 :( and outputs the movable base return command SR to the tracking controller 24, and then proceeds to 5TEP 14 to output the origin signal So Wait for input.

When the movable heath return command SR is input to the follow-up controller 24, the FF 42 in FIG. 7 is set via the OR circuit 43, the return circuit 44 is activated, and the analog switch 45 is turned on.

As a result, the return signal SB output from the return circuit 44 is input to the amplifier 37.
The motor 10 begins to rotate to return the movable base 11 to its origin.

Note that while the movable base 11 is returning, the robot 16 returns to the work origin.

Then, when the movable base 11 returns to the origin and the origin detection limit switch 27 shown in FIG.

When the origin signal So is input to the tracking controller 24,
The FF 42 in FIG. 7 is reset and the return circuit 44 stops operating, and the analog switch 45 is turned off, so the motor 10 stops and the movable base 11 also stops.

Further, when the origin signal So is input to the microcomputer 23, the CPU 23a operates from 5TEP14 to 5TEP14 in FIG.
The process advances to TEP15, and it is checked whether or not the all work completion signal SEE is input from the robot control device 22. If the all work completion signal SEE is not input, the process returns to TEP15.
Returning to TEP 2, if the all-work completion signal SEE has been input, the process proceeds to 5 TEP 16, where a glock start/stop command S84 representing a clock stop is outputted to the drive detection circuit of the following position detector on the 1st day, and the slit After outputting a lighting/lighting-off command SL indicating turning off to the light source 20, all processing ends.

It is assumed that the robot control device 22 outputs an all-work completion signal SEE in response to a command from a sequence controller (PC) that sequentially controls various devices installed on the line, for example.

When it is confirmed in the processing of S'rEP15 that the all work completion signal SEE has not been input and the process returns to 5TEP2, C
The PU 23a again outputs the picking work start command 3s to the robot control device 22.
When receiving the picking work start command SS, the robot 13 performs play pack control so that the robot 13 picks up the rear window crow 16 from the wind crow supply device 14 and carries it to a predetermined standby position.

“Then, when the CPU 23a receives the picking work completion signal SE from the robot control device 22 at 5TEP 3 in FIG. Corresponding numerical data (W
x)(=(W2 J).

Then, in the primary 5TEPs 5 and 6, the binarized video data Dv1. After performing processing based on the tracking start timing signal Sc and numerical data [W] (= [W2 )) obtained from Dv2 in the same manner as in the above case, check whether [WJ'' (W x J) is performed in 5TEP7. Then, only when (WJ = [Wx), the processing related to the above-mentioned follow-up and assembly work after 5TEP 8 is performed.

As a result, while the movable base 11 follows the second follow-up position on the slat conveyor 1 side, the robot 13 on the movable base 11 assembles the rear window crow 16 to the vehicle body A flowing in the direction of the arrow Y. After the assembly is completed, the movable base 11 can be returned to its original position.

Then, as described above, both the front and rear Indian crow I are attached to one car body 3 flowing in front of the following unit 8.
After completing the work of assembling E+, 17, 5 in Figure 9
When it is checked in TEP15 that the complete work signal SEE has not been inputted from the robot control device 22, the process returns to sn:p2 and the front windshield crow 17 is then assembled on the next vehicle body B flowing in front of the follow-up unit 8. 4=J
Similarly, the six processes for performing the work to be performed are sequentially performed.

Then, if the movable base 11 on which the robot 13 is mounted can be made to follow either the front side or the rear side of the six vehicle bodies 3 as shown in FIG. 1-The robot 16 cannot obtain the necessary range of motion for attaching the crow 16°17 to the car body 3.
In addition, in the embodiment described in 2, the front me wind crow 17 can be made to perform such work.
After the movable base 11 is assembled to the vehicle body 6, the movable base 11 is returned to its original position.
Depending on the arrangement position of the rear window glass 14, the rear window glass 16 may be returned to an intermediate position where it is easy to pick, instead of returning to the original position.

In addition, in the above embodiment, the work of assembling the front and rear window glasses 16 and 17 to the vehicle body B is explained as an example, but the work is not limited to this, and for example, the sealing work performed on each part of the vehicle body 3 is explained. The present invention can be similarly applied to the case where the method is carried out.

When performing such sealing work, if two follow-up positions on the slat conveyor 1 side are insufficient, a new third follow-up position will be added. It is sufficient to set a fourth follow-up position and attach a dogleg-shaped reflector having a back reflector having a different thickness for each position.

Furthermore, in the above embodiment, the reflected slit light L21°12
Numerical data corresponding to the interval of 2 ([Wt J .

Although the example in which the position to be followed by the movable base 11 is selected by LW2)) is described above, the present invention is not limited to this. , a selection is made based on the reflected slit light L 2+ * r-22
The numerical data corresponding to each interval of r L 21 +... may be detected and the selection based on the average value thereof may be made.Furthermore, in the above embodiment, the In the above example, only one side of the shaped reflector is used as a back reflector.

Of course, both may be used as back reflectors.

In the above embodiment, the work performed by the robot is described in which a special machine such as an automatic screw tightening machine or an automatic welding machine is placed on a movable base and the special machine follows the workpiece. It is also possible to have the robot perform the dedicated work while the robot is being trained.

[Effect of the Invention J As described above, in the tracking position selection device according to the present invention, in the tracking device in which the movable base on which the automatic working machine device is mounted follows the conveyor, two reflecting plates are used. A dogleg-shaped reflector formed by making at least one of them a back reflector and making both reflectors orthogonal to each other,
The back reflectors are attached to a plurality of predetermined follow-up positions on the conveyor side with different thicknesses for each follow-up position, and light is irradiated from the movable base side to the dogleg-shaped reflector. Since the following position on the conveyor side to be followed by the movable base is selected based on the reflected light, it is possible to expand the work area of an automatic working machine such as a robot for the workpiece conveyed by the conveyor.

[Brief explanation of the drawing]

FIG. 1 is a perspective view of a wind crow assembly work stage and its surroundings to which a follow-up position selection device according to the present invention is applied. FIGS. 2 and 3 show the dogleg-shaped reflector 6 and FIG. 1, respectively.
FIG. FIG. 4 shows doglegged reflectors 6, 7. FIG. 3 is a schematic diagram showing the posture and positional relationship of the slit light LL+reflected slit light L2+ and the line image sensor 25; 5 is a block diagram of the control system shown in FIG. 1, FIG. 6 is a sectional view showing how the tachogenerator shown in FIG. A block diagram showing a specific configuration of the detection circuit 26. FIG. 8 is a diagram showing fluctuations in the conveying speed of the slat conhair 1 shown in FIG. 1. FIG. 9 is a processing flow diagram showing a program executed by the CPU 23a of the microcomputer 23 in the follow-up control unit 21 of FIG. FIGS. 10 and 11 are explanatory views for explaining the function of the dogleg-shaped reflector 6.7, respectively. 1...Slat con hair 2...Jig 3...Car body 4, 5...Bracket 6.7...Dovetail-shaped reflector Eia, 7a...Back reflector 8...Following Unit 9.29...Tachogenerator 10...Motor 11 - Movable head 12°" Hunt 13...Robot 14.15...Wind crow supply device 16...Rear window crow 17...Front window glass 1st...Following position detector 20...Slit light source 2
1...Following control unit (selection means) 22...Robot control device Fig. 2 Fig. 3 Fig. 4 Fig. 10 DV2 -Top-4

Claims (1)

[Scope of Claims] 1. While sequentially detecting the relative position deviation in the conveyance direction of the conveyor between the follow-up position on the conveyor side and the reference position on the movable base side on which an automatic working machine device such as a robot is mounted, In the following device, the traveling of the movable base is controlled so that the movable base follows the conveyor so that the detected relative position deviation becomes zero, and at least one of the two reflectors is configured to reflect the back surface of the movable base. A dogleg-shaped reflector formed by making both reflecting plates orthogonal to each other as a plate is placed at a plurality of predetermined follow-up positions on the conveyor side, and the thickness of each back reflector is made different for each follow-up position. At the same time, a selection means is provided for selecting a tracking position on the conveyor side to be followed by the movable base 1 based on reflected light that is irradiated from the movable base side to the dogleg-shaped reflector and reflected. A follow-up position selection device characterized in that it is configured by: 2. A slit irradiating each of the doglegged reflectors from the movable base side in a direction in which the joint portion of each two reflectors is perpendicular to the conveying direction of the conveyor and perpendicular to the conveying direction. 2. The follow-up position selection device according to claim 1, wherein the follow-up position selection device is attached to each follow-up position on the conveyor side so that both optical paths of the light and the reflected slit light reflected by the two reflecting plates are parallel to each other.