JPS63315902A - Work position shift detection device - 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 Application Field] The present invention relates to industrial robots, and is particularly suitable for use in sealing robot systems that need to correct the amount of positional deviation of a car body as a workpiece. The present invention relates to a positional displacement detection device.

"Prior Art" Generally, when a robot performs a predetermined work on a workpiece, it is necessary to detect the position of the workpiece and supply the information to the robot, for example, by the following method.

■ When the workpiece is intermittently sent to the robot work position (work station) by the tact conveyor,
The three-dimensional position (one position shift) of the fed workpiece (vehicle body) is detected by multiple non-contact sensors installed at or near the working station, such as a television camera, to obtain image information of the workpiece. Determine the amount of positional deviation from the image information.

■ If the workpiece is to be continuously operated by a robot by a conveyor (passing through a device), a pulse generator is installed that generates pulses synchronized with the drive system of the conveyor.
The amount of one-dimensional displacement (positional deviation) of the vehicle body is determined on the assumption that the count value of this pulse is proportional to the moving distance of the workpiece in the flow direction.

"Problems to be Solved by the Invention" However, the above method (■) attempts to detect displacement in each direction, and a sensor is provided to detect distance data and position data along each coordinate axis. Moreover, it is necessary to perform complex arithmetic processing on a large amount of data obtained from these sensors, and since positional deviation is detected using a stopped workpiece as a surface, it is necessary to perform continuous processing on a flow conveyor. It cannot be applied to workpieces that are transported manually.

In addition, the above method (■) attempts to determine the position on the premise that the actual movement amount of the workpiece matches the conveyor pulse, or if there is surging of the conveyor, for example, , Strictly speaking, the detected pulse and the actual amount of movement of the vehicle body are not proportional.
The actual position of the vehicle body will be different. Furthermore, since the method only detects a one-dimensional movement amount in the flow direction, there is a problem that it cannot be applied to three-dimensional positional deviation detection.

The present invention has been proposed in view of the above circumstances, and an object of the present invention is to calculate the amount of positional deviation of a moving workpiece using as little data as possible and supply positional information to a robot.

"Means for Solving the Problems" In order to achieve the above object, the present invention provides a light source that is provided on one side of the robot working range in the middle of the workpiece movement path and irradiates the workpiece with a light beam throughout the working range. , a plurality of detectors provided on the other side of the robot working range, facing the light source and along the moving direction of the workpiece, and having the same height level along the transporting direction of the workpiece throughout the working range; It is composed of a one-dimensional sensor that detects the movement of each point, and a calculation device that calculates the data detected in the one-dimensional sensor, and the calculation device detects the movement of the detection point obtained from the one-dimensional sensor. The distance between the detection points is calculated from the data, and the amount of positional deviation of the workpiece in a specific direction is calculated from the calculated distance data and known data regarding the distance.

According to the above configuration, the movement of multiple detection points along the movement direction of the workpiece is detected throughout the working range of the robot, and by processing this data, the distance between the detection points at each level can be calculated. Calculated. By comparing this distance with known distance data, it is possible to know how far the position or posture of the workpiece deviates from the base color.

"Example" Hereinafter, one embodiment of the present invention will be described with reference to the drawings.

Reference symbols IA-5A and IB-5B in FIG. 1 are one-dimensional image sensors. These one-dimensional image sensors IA~
5A and IB to 5B are composed of an array having a structure in which a large number of phototransistors are arranged in a row, for example, and the second
As shown in FIG. (omitted) is placed on one side of the conveyor along the direction of movement.

In addition, the six pairs of image sensors IA-IB to 5A/5
Light sources +c to 5c are provided at positions facing B, respectively, and the rays of light emitted from these light sources nC (where n is an integer from 1 to 5) are partially blocked by the workpiece 6. , each pair of image sensors nA-n while moving in a predetermined pattern as the workpiece 6 moves.
B can be switched ON or OFF.

Further, the six pairs of image sensors nA to nB are designed to detect the light rays emitted from the light source nC from different directions, as shown in FIG.

That is, in the illustrated example, the image sensor IA is connected to the light source I.
Image sensor IB is placed at the same level as C, and image sensor IB is placed above image sensor IA, and a straight line connecting image sensor IA to light source IC and image sensor IB to
The straight line connecting the light source ICs forms an angle of α. Note that the other pairs of image sensors 2A/2B, 3A/
3B14A, 4B, 5A, and 5B are also arranged so that the straight line toward each of the light sources 20 to 5C has an angle of α, similarly to the image sensor IA-IB.

Therefore, as the workpiece 6 passes in front of the light source IC, as shown in FIG. Patterns corresponding to the detection points 6B are respectively detected by the image sensor IB.

And the image sensor IA-II3-5A/5B
The analog outputs of converters 7A and 713 to II are respectively
After being converted into a dinotal signal by A.IIB, it is supplied to the arithmetic unit 12. Further, a storage device 13 is connected to the arithmetic device 12 .

Hereinafter, the operation of the workpiece position deviation detection device and the calculation processing in the calculation device 12 will be described in (6).

As shown by the arrow in FIG. 1, the workpiece 6 is transported and the light source tC
~5C and the image sensor IA-IB~5A/5B, and when the workpiece 6 (strictly speaking, a pillar etc. that forms part of the workpiece 6) blocks the light beam, the image sensor (strictly speaking In this case, the output of the large number of elements (such as phototranostas) that make up this becomes "dark" below the threshold value, and the output of the unobstructed portion becomes "bright." At this time, the arithmetic unit 12 processes all the image sensors IA and IB.
~The output of 5A and 5B is constantly monitored,
First, it is recognized that the pillar detection points 6A and 6B have passed in front of the image sensors (5A and 5B in FIG. 1) that have been switched from "bright" to "dark" for the first time.

In the case of this embodiment, as shown in FIG. 3, the measurement ranges of the image sensors are arranged so as to partially overlap, so that the movement of the workpiece 6 can be detected seamlessly.

The storage device 13 also includes image sensors IA-IB.
The relationship between the elements of ~5A and 5B and the corresponding movement amount is stored in advance, and based on this data, the brightness/darkness switching in each bit is determined by the absolute value of the movement amount of the workpiece 6. It is converted to . Furthermore, the storage device stores the vehicle body length QAO% at detection line A (the line corresponding to the height detected by image sensors IA to 5A) and detection line B (the line corresponding to the height detected by image sensors IA to 5A).
5B) is stored as a known value, and as shown in FIG. 5, after the detection point 6A of the front pillar is detected, . -Δc ...... The position of the specific bit of the image sensor IA-5A that is switched at a position more than the distance corresponding to (1) is recognized as the position of the detection point 6A' of the rear pillar. .

In addition, the above Δm is the width of the rear pillar (dimension in the X direction)
It represents the amount of change (error) in the length of the vehicle body that occurs on the detection line A in accordance with the positional deviation of the vehicle body due to the tilt and tilt, and taking into account this error ΔQA, as shown in FIG.
) After moving the image sensor 5 by a length corresponding to the formula
The rear pillar passage detection data obtained from A is imported.

On the other hand, before the switching operation is performed, it is assumed that only a portion of the vehicle body 6 exists within the detection range of the image sensors IA to 5A. Furthermore, image sensor IB~5B
Similar processing is also performed at detection point 6 of the rear pillar.
B' is recognized, and as a result, it is recognized that the vehicle body 6 has entered the robot's working range.

As described above, detection points 6A, 6B, 6A',
When the passage of 6B' is recognized, as shown in Fig. 4, depending on which bit of which image sensor is switched, the detection point 6A or 6B of the front pillar is changed from the detection start point to the front pillar detection point 6A or 6B.
The distance X l + X 2 is calculated, and the distance X 3.
is calculated. Then, based on the above x1 to x4, the vehicle body lengths C and QB at the detection lines A and B are calculated as follows.

Bar Xl-X4 ・・・・・・(2) QB””X
2X3 ... (3) Furthermore, as shown in Fig. 4, the car body 6 is arranged in a south reference system consisting of three sleeves X-y and Z, that is, the flow direction of the conveyor is the X-axis, the left-right direction is the y-axis, Fitting into a coordinate system with the vertical direction as two axes, and assuming that the amount of positional deviation parallel to the y-axis direction is small, the X-axis direction of the vehicle body 6,
When trying to find the positional displacement parallel to the Z-axis direction and the rotation around the y-axis, the following results are obtained.

That is, the above C and σ8 are known constant values, QAO and QB, when the vehicle body 6 is kept in a constant direction in the horizontal plane with respect to the moving direction and there is no vertical vibration.
Since it should match Q, QA−QAo
...= (,4)QB=Ce.・
...-If (5) holds true, the posture is maintained in which the rotational displacement around the y-axis is Δθy−O and the parallel displacement in the z-axis direction is Δ2=0, that is, a horizontal posture with no deviation. If it is, the detection start position of the vehicle body 6 (
It is recognized that the moving distance of the rear pillar from the detection points 6A' and 613' (a position beyond the detection start point) is equal to x4.

Also, a＾≠mu.・・・・・・(6) i! a#
(2no --... (7), it is recognized that the vehicle body 6 is not in a horizontal position, and in this case, the deviation amounts Δθy and ΔZ and the detected IaQA and QB are approximately as follows. A relationship is established.

In other words, the conversion from the coordinate system represented by the variables IL and ρ8 to the coordinate system represented by the variables Δθy and Δ2 is performed by multiplying the matrix ■, which is generally defined as By substituting % as the conversion formula from the C-ρ8 system to the ΔθyΔZ system in the above example, and further transforming the above equation (9), the following can be obtained. Therefore, the detected Δha and ΔQ8
The deviation amount ΔZ and Δθy can be determined from the known T and the deviation amount ΔZ.

Note that (2) above differs depending on the type of vehicle, that is, the shape of the workpiece.For example, the vehicle body 6 is displaced by a known Δ2 from the reference posture to measure Δc・ΔeB, and then further displaced by a known Δθy. It can be determined by measuring Δc·ΔI2B and performing partial differentiation according to the above equation (8).

Therefore, Δ2 and Δθy can be determined by determining T in advance for each vehicle type and storing it in the storage device 13, and applying this T to the equation (9).

Furthermore, the amount of movement X after the vehicle body 6 enters the robot's working range will match X4 in FIG. There is an error between the amount of movement and the amount of movement X. The correction amount ΔX for correcting this error can be approximated by the following formula.

As for t1 and t, the correction amount ΔX can be easily determined by calculating them in advance for each vehicle type and storing them in the storage device 13 in the same way as in the case (2) above.

Then, from ΔX and X4 obtained by the above formula (11),
...Movement fttx can be obtained as shown in (!2).

Then, the arithmetic unit 12 calculates [x, Δ2. Δθy] is supplied to the robot control device as position correction data of the vehicle body 6, and this correction data is supplied to the robot control device,
Control is performed to follow the movement and attitude change of the vehicle body 6. Note that the deviation ff1x is essential data in order for the robot to perform work on the work being transported, and ΔZ is the deviation that occurs due to the inevitable deflection of the conveyor that is intermittently supported. Furthermore, Δθy is a deviation (so-called pitching) that inevitably occurs in a workpiece whose shape matches the conveyance direction and the longitudinal direction, such as an automobile body.On the other hand, deviation in other directions, for example Δy (Horizontal deviation along the width direction of the vehicle body), Δθx (so-called rolling centered on the x sleeve, so-called yawing centered on the Cz axis)
are very small compared to x1ΔZ and Δθy, so by appropriately detecting these shifts, it is possible to perform a practically sufficient shift correction.

"Modified Embodiment of the Invention" (1) In the above embodiment, two systems of one-dimensional image sensors were provided to perform measurements at two levels, upper and lower.
If it is clear that the deviation in a particular direction is small due to factors such as the shape and weight of the workpiece or hanger, measure only at one level, that is, install only one system of image sensor nA. , only H may be determined.

For example, rolling ΔθX1 pitching ΔO of workpiece 6
If the y1 yawing ΔθZ and the deviation Δy in the y-axis direction are negligibly small, [x, Δ2], that is, the amount of deviation between the amount of movement accompanying the conveyance of the conveyor and the weight of the vertical parallelism. It is sufficient to obtain the data of , and in this case, the above-mentioned deviation amounts can be calculated based on the measured value C as follows.

In other words, in a typical passenger car, the front pillar and rear pillar are tilted at a constant angle in opposite directions, so if the vehicle body position changes by ΔZ,
The length of the vehicle body detected on an inspection line at a specific height (for example, inspection line A). and the vehicle body length b at a height deviated by Δ2 from this inspection line, ΔQA20A-QA.

There will be a difference. (In general passenger cars, both pillars are so-called stationary, so above the inspection line, +2A<+2AO, and below the inspection line, (!A>(!AO
becomes. ) Then, the vehicle body lengths Δc and Δ with such a relative relationship
The ratio of 2, that is, 1-ΔQA/ΔZ, is a constant value for each vehicle type in cars with general straight pillars (ΔQA and ΔZ are linear functions), so this 1 in advance and memorize it, and by performing the calculation Δz - (1/)・Δ1 (11) corresponding to the measured car body length C, the vertical direction can be calculated. The amount of positional deviation ΔZ can be determined.

In addition, as described in the illustration, the positional deviation ΔZ (in the case shown, Δ
2) occurs downward, the starting point 6 of the measurement of the car body length C, A z, becomes the original starting point (
The starting point in the case of Δz=0) is located above 6Az', and the point 6Az is mistakenly recognized as the original starting point 6Az', resulting in an error of 19×ΔX in the transport direction of the workpiece.

However, regarding the above ΔX, if the g1 degree of the front pillar is a known constant value, the following relationship holds true: -ΔX/ΔZ Therefore, calculate 2 as a constant in advance and store it in the storage device 13. Set aside, add 2 to this, and add 1
1) From Δ2 obtained from equation ΔX−ζ,・ΔZ
...By calculating ΔX based on the formula (12) and adding (or subtracting) this ΔX to the data of the detected front pillar position, the actual The position of the workpiece can be known.

Note that when the downward movement of the workpiece 6 is small and the pitching is large, X and ΔOy can be determined in the same manner as in the above case.

(11) The number of one-dimensional image sensors is not limited to the above-mentioned example, but as long as the detection points can be determined continuously without any break along the conveyance direction of the workpiece, all the detection points can be detected. Either a method that covers the range with one device or a method that uses multiple devices may be used.

(1ii) The structure of the one-dimensional sensor does not necessarily have to be an image sensor as in one embodiment, but it may be of another type that can one-dimensionally detect the movement of the detection point, or it may be a two-dimensional image sensor. Of course, it is also possible to use a system that simultaneously detects the movement of a detection point at a plurality of levels for the first time.

"Effects of the Invention" As is clear from the above explanation, the following effects can be obtained according to the present invention.

■ Since the one-dimensional sensor detects the movement of the detection point throughout the robot's working range, deviations from the reference posture of the workpiece being transported are continuously detected. By moving the robot in parallel along the transport direction of the workpiece, the robot can be made to work on the moving workpiece in a self-evident manner.

■ Since the amount of movement of the workpiece is detected based on data from a one-dimensional sensor, the conventional pulse transmitter of the conveyor drive system or the start phototube that detects when the workpiece enters the robot work range is used. becomes unnecessary.

■ Since the amount of deviation is calculated based on a relatively small amount of data obtained from a one-dimensional sensor, the amount of deviation can be determined more quickly than, for example, a method that obtains data on all six degrees of freedom. Therefore, it is possible to follow changes in the posture of the workpiece almost in real time.

- Since the one-dimensional sensor and its essential light source can be installed outside the conveyor line, they can be easily additionally installed without affecting the existing conveyor line.

[Brief explanation of drawings]

1 to 5 show one embodiment of the present invention,
Figure 1 is a plan view, Figure 2 is a side view, Figure 3 is a front view, Figure 4 is an explanatory diagram of the measurement principle, Figure 5 is a diagram showing the relationship between the shape of the workpiece and the waveform of the detection pulse. FIG. 6 is an explanatory diagram of the measurement principle in another embodiment of the present invention. IA~5A・・・One-dimensional image sensor, IB~
5B...One-dimensional image sensor, tC~5C...
...Light source, 6... Work (car body), 6A, 6B,
...To detection, 6. ~° ・6B' ・・・・・・
Detection point, 7A-11A...Converter, 7B-11B...
...Converter, 12... Arithmetic device, 13...
···Storage device. Applicant Tokiko Co., Ltd. Figure 1-2-3

Claims (2)

[Claims] (1) A light source that is provided on one side of the robot working range in the middle of the workpiece movement path and irradiates the workpiece with a light beam throughout the working range, and a light source that is provided on the other side of the robot working range and that faces the workpiece. a one-dimensional sensor that is provided along the movement direction of the workpiece and detects the movement of a plurality of detection points that are arranged along the movement direction of the workpiece and have the same height level throughout the work range; and a calculation device that calculates the distance between the detection points from the movement data of the detection points obtained from the one-dimensional sensor, and the calculation device calculates the distance between the detection points from the movement data of the detection points obtained from the one-dimensional sensor. A workpiece positional deviation detection device, characterized in that the workpiece positional deviation detection device calculates a positional deviation of the workpiece in a specific direction from data. (2) The one-dimensional sensor is provided with a plurality of systems to detect the movement of the detection point at different height levels, and the calculation device calculates the distance between the detection points at different height levels, and The workpiece positional deviation detection device according to claim 1, wherein positional deviations of the workpiece in a plurality of directions are calculated from known data at a height level and the distance data.