JP2007237394A - Workpiece positioning device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To correct the position of a workpiece as a product placed on an intermediate table when a carry in-and-out robot loads a product loading table with the workpiece, whereby the products is stacked in multi-stage without destruction on the product loading table in a bending system using the robot. <P>SOLUTION: This workpiece positioning device includes: the intermediate table for placing the workpiece as the product; a workpiece imaging means fitted with the intermediate table 8; a workpiece image detecting means for detecting an image of the workpiece input from the workpiece imaging means; a workpiece reference image calculating and storing means for calculating the reference image of the workpiece, based on previously input information; a shift amount calculating means for calculating a predetermined shift amount based upon a difference between the detected image and the reference image; and a robot control means for correcting the posture of the carry in-and-out robot so that the detected image and the reference image match each other, based on the shift amount, and loading the product loading table with the workpiece. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

The present invention makes it possible to correct the position of a workpiece as a product placed on an intermediate platform when the loading / unloading robot loads it on the product loading table, and the products can be loaded in multiple stages without breaking on the product loading table. The present invention relates to a workpiece positioning device.

    Conventionally, when a bending process is automatically performed using a robot, for example, the loading / unloading robot 3 (FIG. 1) sucks a workpiece as the material T on the material table 1 and attaches it to the bending apparatus 10. The material T is delivered to the bending robot 11, and the bending robot 11 supplies and positions the material T to the bending apparatus 10 to perform bending.

After the processing, when the bending robot 11 places a workpiece as the product Q on the intermediate table 8, the loading / unloading robot 3 sucks the product Q and sequentially loads it on the product loading table 12.
None

  In such a bending system, in general, the workpiece on the material table 1 is often misaligned. When the workpiece is transferred from the loading / unloading robot 3 to the bending robot 11 and processed as it is, Since there is a deviation in the positional relationship between the workpiece before processing and the bending apparatus 10 (for example, a positional deviation in the longitudinal direction between the workpiece and the molds P and D, a positional deviation in contact between the workpiece and the abutment), a predetermined amount There is a case that the product of the size cannot be produced, the bent flange is crushed, and the workpiece does not come into contact with the abutment, so that efficient machining cannot be performed.

    Therefore, the machine has to be stopped, and therefore, the processing cannot be continued, and there are various problems such as many defective products are generated and the material is wasted.

  Further, conventionally, the work is carefully loaded on the material table 1 by hand so that the above-described positional deviation does not occur, so that the loading time is long.

  This is also true for the workpiece as the product Q after processing. In general, the workpiece as the product Q placed on the intermediate table 8 (FIG. 11) is often misaligned. Therefore, even if the work as the product Q is gripped by the loading / unloading robot 3 and placed on the product loading table 12 in the state as it is, it is separated and collapsed one by one. As a result, conventionally, workpieces as the product Q cannot be stacked in multiple stages without breaking on the product loading table 12.

An object of the present invention is to enable correction of a position when a loading / unloading robot loads a workpiece as a product placed on an intermediate platform on a product loading platform in a bending system using a robot, It can be loaded in multiple stages without breaking on the loading platform.

In order to solve the above problems, the present invention, as shown in FIG.
An intermediate platform 8 on which workpieces as product Q are placed;
A work imaging means 2 attached to the intermediate table 8;
A work image detection means 30D for detecting an image DW of a work as a product Q inputted from the work imaging means 2 (FIG. 1);
A workpiece reference image RW as a product Q is calculated based on information input in advance, or a workpiece reference image calculation for storing a workpiece image as a product Q at a normal position captured by the workpiece imaging means 2 as a reference image RW. Storage means 30E;
A deviation amount calculation means 30F for calculating a predetermined deviation amount based on the difference between the detected image DW and the reference image RW;
Based on the amount of deviation, the posture of the loading / unloading robot 3 that grips the workpiece as the product S on the intermediate platform 8 is corrected so that the detected image DW matches the reference image RW, and the product Q remains in the corrected posture. Technical means called a workpiece positioning device characterized by having robot control means 30G for loading the workpiece as a product loading table 12 was taken.

According to the configuration of the present invention described above, the workpiece imaging means 2 (FIG. 1) images a predetermined portion of the workpiece, for example, the corner portion K, and the deviation amount calculating means 30F is detected image detected by the workpiece image detecting means 30D. and detecting the corner section _{K D} in the DW, the difference between the work reference image calculating and storing means is calculated by 30E, or reference corner _{K R} of the reference image in RW stored therein _{PIC - X, PIC - Y,} PIC _- based on the theta (FIG. 6), the deviation amount ΔX of the actual gripping position H _R and the original holding position H _D of the workpiece by the robot, [Delta] Y, and calculates the [Delta] [theta], the robot control means 30G is, the deviation amount ΔX , [Delta] Y, based on [Delta] [theta], it is possible to grasp the actual gripping position H _R and the example loading and unloading position of the robot 3 as original holding position H _D coincides with the correction control to position the workpiece properly position .

  Thus, according to the present invention, in the bending system using a robot, the loading / unloading robot 3 (FIG. 11) loads the workpiece as the product Q placed on the intermediate platform 8 on the product loading platform 12. The position of the product Q can be corrected, and the product Q can be stacked in multiple stages without breaking on the product loading table.

Hereinafter, the present invention will be described with reference to the accompanying drawings by embodiments.
FIG. 1 is an overall view showing an embodiment of the present invention.

  The bending system shown in FIG. 1 includes a bending apparatus 10, a bending robot 11 attached to the bending apparatus 10 via a base plate 15, and a loading / unloading robot 3 installed on the side of the bending apparatus 10. Have.

In the vicinity of the loading / unloading robot 3, for example, a material stand 1 of a workpiece picking device for loading a workpiece as the material T, an intermediate platform 8 on which the bending robot 11 releases and places a product Q immediately after processing, A product loading table 12 on which the outgoing robot 3 loads the product Q adsorbed from the intermediate table 8 is arranged.

  Thereby, the workpiece placing means for placing the workpiece is constituted by the material table 1, the intermediate table 8, and the product loading table 12. Among these materials, for example, the material table 1 has a workpiece imaging means 2 described later. It is attached.

  As the bending apparatus 10 in this case, for example, there is a press brake, and as is well known, the punch P mounted on the upper table 13 (FIG. 2) and the die D mounted on the lower table 14 are provided. Then, a predetermined bending process is performed by the punch P and the die D on the workpiece whose proper position is held by the grip 17 of the bending robot 11 described later.

The bending robot 11 has, for example, five axes, and has the grip 17 at the tip thereof as described above, receives a workpiece as the material T from the loading / unloading robot 3, and will be described later (FIG. 6). in the original holding state in which the position h gripping the _D of the workpiece, the bending apparatus 10 bending after supplying and positioning grips between said workpiece T to be processed into the upper middle mount 8 it if could products Q To release from the grip 17.

  The carry-in / out robot 3 has, for example, six axes, has a suction pad 3A, sucks the material T on the material table 1 with the suction pad 3A, and delivers it to the bending robot 11 as described above. The product Q is sucked from the intermediate table 8 by the suction pad 3A and loaded on the product loading table 12.

In this case, as will be described later (FIG. 4), for example, a corner portion K of the workpiece as the material T adsorbed by the loading / unloading robot 3 is imaged by the workpiece imaging means 2 attached to the material table 1. The workpiece detection image DW including the image K _D (FIG. 6) of the corner portion K is detected by the workpiece image detection means 30 </ b> D constituting the lower NC device 30 (FIG. 1).

As a result, the detected image DW is calculated by the workpiece reference image calculation / storage means 30E (FIG. 1) or input to the deviation amount calculation means 30F together with the reference image RW stored therein, and then the deviation amount. the calculation unit 30F, a predetermined shift amount, and more particularly, the difference between the detected corner portion _{K D} in the detected image DW, a reference corner _{K R} of the reference image _{RW PIC - X, PIC - Y} , PIC - θ based on the deviation amount ΔX of the actual gripping position H _R and the original holding position H _D of the workpiece by the robot, [Delta] Y, [Delta] [theta] is calculated.

Accordingly, when the robot control means 30G converts the deviation amounts ΔX, ΔY, Δθ into the correction drive signals S _a , S _b ..., The robot control means 30G matches the detected image DW with the reference image RW. Further, for example, by correcting and controlling the posture of the loading / unloading robot 3 (down arrow in step 104 in FIG. 10), the workpiece transfer position to the bending robot 11 is corrected (step 105 in FIG. 10). Since the workpiece is positioned, if the loading / unloading robot 3 delivers the workpiece to the bending robot 11 in the corrected posture (step 106 in FIG. 10), the bending robot 11 has the proper position h _{D of the} workpiece. With the gripping state (FIG. 6) (step 107 in FIG. 10), by supplying and positioning to the bending apparatus 10 (step 108 in FIG. 10), bending is performed. Is performed (step 109 of step 8 in FIG. 10).

  Thus, according to the present invention, in the bending system using the robot, the bending robot 11 that has received the workpiece from the loading / unloading robot 3 (FIG. 1) performs the bending process while holding the appropriate position of the workpiece. Since the machining can be continued without stopping the machine, the number of defective products is reduced accordingly, and the workpiece is placed on the material stand 1 which is a workpiece placing means arranged near the loading / unloading robot 3. Since the image pickup means 2 is attached and the positional deviation of the workpiece on the material table 1 is detected by image processing, the workpiece can be loaded on the material table without worrying about the positional deviation, thereby shortening the loading time. Is possible.

  As described above, the workpiece imaging means 2 is attached to the material table 1 (FIG. 4), and the workpiece imaging means 2 is composed of, for example, a CCD camera 2A and its light source 2B. A light source 2B is mounted directly below the table 1, and the light source 2B is movable in the left-right direction (X-axis direction). For example, when the light source 2B moves to the left side, it irradiates transmitted light from its lower surface. It is supposed to be.

With this configuration, the unloading example corner K of the work as gripped material T by the suction pads 3A of the robot 3, are captured by the CCD camera 2A, the image K _D of one-dimensional electrical signal of the corner portion K by being converted into two-dimensional electric signal by being converted into work image detecting means 30D, in deviation amount calculating means 30F, the reference image RW including the detected image DW comprising detecting corners K _D, the reference corner K _R Are compared (FIG. 6). In this case, the object to be imaged by the CCD camera 2A is an arbitrary part of the workpiece for determining the position of the workpiece in the X-axis direction, the Y-axis direction, and the angular direction, and is not limited to the corner portion K described above. It may be a hole or a notch formed in the workpiece.

  As a control device of the bending system having the above-described configuration, there are a high-order NC device 29 (FIG. 1) and a low-order NC device 30.

  Among them, the host NC device 29 incorporates CAD information, which includes the thickness of the workpiece, the material, the length of the bending line, the position of the corner portion K that is the object to be imaged by the workpiece imaging means 2, and the like. Product information such as workpiece information and product bending angle is included, and these are configured as a three-dimensional solid view and a development view.

  CAD information composed of these pieces of information is input to the lower NC device 30 and can also be used for detecting a positional deviation of a workpiece by image processing of the present invention.

  The subordinate NC device 30 (FIG. 1) includes a CPU 30A, an information calculation means 30B, an imaging control means 30C, a work image detection means 30D, a work reference image calculation / storage means 30E, a deviation amount calculation means 30F, a robot The control unit 30G, the bending control unit 30H, the conveyance control unit 30J, and the input / output unit 30K are included.

  The CPU 30A performs overall control of the entire apparatus shown in FIG. 1, such as the work image detection means 30D and the work reference image calculation / storage means 30E, according to the operation procedure of the present invention (for example, corresponding to FIG. 10).

  The information calculation means 30B is information necessary for detecting the positional deviation of the workpiece such as the bending order, correcting the posture of the robot, and bending work based on the CAD information input from the host NC device 29 via the input / output means 30K described later. Is determined by calculating.

  The information calculated and determined by the information calculating means 30B includes, in addition to the bending order, the mold to be used (punch P and die D) and which mold is disposed at which position of the upper table 13 and the lower table 14. A mold layout (mold station), an operation machining program for the bending robot 11 and the loading / unloading robot 3 are also included.

  Thereby, for example, based on the difference between the detected image DW detected by the workpiece image detecting unit 30D and the reference image RW calculated by the workpiece reference image calculating / storing unit 30E, the deviation amount calculating unit 30F has a predetermined deviation amount ΔX. , ΔY, Δθ are calculated (step 103 in FIG. 10), and it is determined based on the deviation amounts ΔX, ΔY, Δθ whether the carry-in / out robot 3 or the bending robot 11 performs the posture correction (FIG. 10). 10 step 104).

Then, as will be described later, correction of the workpiece delivery position by the loading / unloading robot 3 → workpiece delivery (down arrow in step 104 in FIG. 10, steps 105 to 106), or correction of the workpiece reception position by the bending robot 11 → receiving (RIGHT step 104 in FIG. 10, step 110-111) after either operation is performed, in a state where the bending robot 11 grips the proper position _{h D} of the workpiece (Fig. 10
Step 107), the workpiece is supplied to a predetermined mold station and positioned (Step 108 in FIG. 10), and bending is performed (Step 109 in FIG. 10).

  The imaging control means 30C controls the movement of the light source 2B (FIG. 4) constituting the workpiece imaging means 2 described above based on the bending order determined by the information calculation means 30B, the position of the corner portion K of the workpiece, and the like. In addition, the imaging operation such as the control of the imaging range S of the CCD camera 2A is controlled.

Workpiece image detecting means 30D is (Fig. 1), as described above, the image K _D of the corner part K of the workpiece consisting of the workpiece pickup means one-dimensional electrical signal sent from 2, the two-dimensional electrophoresis Convert to signal.

Thus, as will be described later (FIG. 5, FIG. 6), the image K _D of the corner portion _K, i.e. obtained detected image DW comprising detecting corners K _D, the detecting corners _the K _D, the reference corner It is compared to the K _R.

Work reference image calculating and storing unit 30E (FIG. 1), the information calculating means 30B is determined by the bending sequence, based on such position of the corner part K of the workpiece, calculates a reference image RW including the reference corner K _R (FIGS. 5 and 6). Alternatively, the workpiece reference image calculation / storage means 30E captures a workpiece placed at a normal position (a workpiece in which no positional deviation has occurred) with the CCD camera 2A, and uses the captured workpiece image as a reference image RW. For example, it is stored in a memory.

  The deviation amount calculation means 30F (FIG. 1) inputs the detected image DW and the reference image RW, and calculates predetermined deviation amounts ΔX, ΔY, Δθ from the difference therebetween.

The principle of the image processing in this case, for example, as shown in FIG. 5 (A), to set the window frame M _R of a predetermined magnitude and direction includes the reference corner K _R, the window frame M Capture data in _R.

Further, on the one hand, it sets the same window frame _{M D} including the detecting corner _{K D} (FIG. 5 (B)), takes in the data in the window frame _{M D,} Ryomadowaku _M D, the _{M R} The deviation amounts ΔX, ΔY, Δθ when the two data coincide with each other are obtained.

Specifically, in the present invention, the bending robot 11 (FIG. 1), the proper position h _D of the workpiece from doing bending (Figure 6) gripped state, during the actual image processing, the actual and holding position H _R by loading and unloading robot 3, shift amount ΔX of the original holding position H _D, [Delta] Y, thereby obtaining the [Delta] [theta]. For example, although unloading robot 3 the workpiece as the material T was the actual in the gripping position H _R (6) gripped by the suction pads 3A (FIG. 1), the workpiece (detected since that time misalignment occurs rather than the center of the image DW), as shown, it means that grips the end of the workpiece, held by the original holding position H _D to grip the center of the workpiece (detected image DW), both Deviation amounts ΔX, ΔY, Δθ are obtained. Note that in FIG. 6, H _R and H _D is carried out robot 3 shows the actual gripping position and the original grip position of the suction pads 3A as (Figure 1) has, h _R and h _D is the bending robot 11 shows the actual gripping position and the original gripping position of the grip 17 as shown in FIG.

That is, in FIG. 6, to set the code of each component as the upper right, and detecting the corner section K _D detected by the workpiece image detection means 30D and imaged with a CCD camera 2A, calculated by the work reference image calculating means 30E X-axis direction component of the difference between the reference corner K _R, Y-axis direction component, the angle direction component respectively shall be expressed as follows.

PIC _- X, PIC _- Y, PIC _- θ (1)

Further, on the detected image DW, to X-axis direction component of the difference between the original holding position H _D of the detection corners K _D and the workpiece, the X-axis direction component shall respectively expressed as follows.

TCP _- X, TCP _- Y (2)

Further, the detected image DW, if caused to pivot by θ in the clockwise direction about the original holding position H _D, the detected image DW obtained by correcting the deviation of the angular direction 'is obtained, the detection image DW' is the reference image RW Becomes parallel.

Then, a corner K _D of the original detected image DW, the X-axis direction component X of the difference between the corner portion e of the detected image DW after turning '(theta), Y-axis direction component Y (theta), the ( It can be expressed as follows using the components 1) and (2).

X (θ) = {TCP ₋ X * COS (−PIC ₋ θ) −TCP ₋ Y * SIN (−PIC ₋ θ)} − TCP ₋ X (3)
Y (θ) = {TCP ₋ X * SIN (−PIC ₋ θ) + TCP ₋ Y * COS (−PIC ₋ θ)} − TCP ₋ Y (4)

In this case, in FIG. 7 which is an enlarged view of FIG. 6, the original detected image DW and this is represented by θ (−θ (−PIC ₋ θ if expressed in the amount of deviation in the angular direction detected via the CCD camera 2A) in the clockwise direction. )), The X (θ) in (3) above and Y (θ in (4) above are taken into consideration when supplementing the right-angled triangles K _D ac and aO _D b taking into account the relationship with the detected image DW ′ swirled only by )
X (θ) = (ac−ab) −de (5)
_{Y (θ) = (cK D} + bO D) -ef ·· (6)
It can be expressed.

A (5) is a point of intersection of the original detected image DW and the line segment E constitute a right triangle K _D ac, b is the intersection of the sides A and line B of the right-angled triangle K _D ac , c is the intersection between the sides a and C at right angles of the right-angled triangle K _D ac, d is detected intersection of the image DW 'and the line B after turning, e is, after turning as described above This is a corner portion of the detected image DW ′.

As is apparent from the figure, since ac = TCP ₋ X * COS (−PIC ₋ θ), ab = TCP ₋ Y * SIN (−PIC ₋ θ), and de = TCP ₋ X, these are expressed as (5 (3) described above can be obtained.

Meanwhile, the _{O D} (6), the center of the original holding position _{H D,} f is the intersection between the detected image DW 'after turning the line segment F, also as already described, _{K D} is the original corners of the detected image DW, b, together with an intersection of the sides a and line B of the right-angled triangle _K D ac, constitute a right triangle _aO-D b, c is a right of a right triangle _K D ac An intersection point e between sides A and C, which is formed, is a corner portion of the detection image DW ′ after turning.

Moreover, as is clear from _{FIG, cK D = TCP - X *} SIN (-PIC - θ), bO D = TCP - Y * COS (-PIC - θ), ef = TCP - since it is Y, these Substituting into (6), the above-described (4) is obtained.

Next, in FIG. 6, the detected image DW 'after turning, the difference between the reference image RW, each a corner portion e, equal to the difference between K _R, X-axis direction component of the difference, Y-axis direction component, as is clear from _{FIG, PIC - X + X (θ} ), PIC - a Y + Y (θ), these components are the actual gripping position _{H R,} the shift amount of the original holding position _{H D} Since it is also an X-axis direction component ΔX ′ and a Y-axis direction component ΔY ′,
ΔX ′ = PIC ₋ X + X (θ) (7)
ΔY ′ = PIC ₋ Y + Y (θ) (8)
It becomes.

Since the above (7) and (8) are representations in the absolute coordinate system that determines the position on the reference image RW and the detected image DW ′ after turning, these are the materials that determine the position on the original detected image DW. If converted to a representation in the coordinate system,
ΔX = ΔX ′ * COS (PIC ₋ θ) −ΔY ′ * SIN (PIC ₋ θ) (9)
ΔY = ΔX ′ * SIN (PIC ₋ θ) + ΔY ′ * COS (PIC ₋ θ) (10)
Δθ = PIC ₋ θ (11)
It becomes.

In this case, in FIG. 7, which is an enlarged view of FIG. 6, if the detected image DW ′ parallel to the reference image RW is represented by θ (+ θ (angle deviation detected through the CCD camera 2A) in the counterclockwise direction. , PIC _- theta)) by turning the original detected image DW right triangle supplemented by considering the relationship between _mO R h, when focusing on mO _D k, and ΔX above (9), [Delta] Y (10) Is
ΔX = hO _R −hg (12)
ΔY = gk + kO _D (13)
It can be expressed.

G of (12), the side G and the intersection of the line segment E of a right triangle mO _R h, h is the intersection of the side G and J at right angles of the right-angled triangle mO _R h, _{O R} is above as was a center of the actual gripping position H _R.

Further, as is apparent from the figure, hO _R = ΔX ′ * COS (PIC ₋ θ) and hg = mk = ΔY ′ * SIN (PIC ₋ θ), so if these are substituted into (12), (9) described above is obtained.

On the other hand, a point of intersection of k, and the perpendicular L drawn to the line segment E from the an intersection of the hypotenuse H and the segment B of the right-angled triangle mO R _h m, a line segment E (13) at right angles constitute a triangle mO _D k, g, _{O D,} as described above, edges G and the intersection of the line segment E of each right-angled triangle mO _R h, is the center of the original holding position _{H D.}

Moreover, as is clear from FIG, gk = hm = ΔX '* SIN (PIC - θ), kO D = ΔY' * COS - because it is (PIC theta), by substituting them into (13), already The aforementioned (10) is obtained.

Further, in FIG. 6, when the coordinate conversion from the absolute coordinate system to the material coordinate system is performed, as described above, the detection image DW ′ parallel to the reference image RW is turned counterclockwise by θ, and thus Δθ is set. , It can be expressed by the amount of deviation in the angular direction detected through the CCD camera 2A, so that PIC ₋ θ is obtained, and the above-described (11) is obtained.

Such a specific example of FIG. 6, when the contrasted with the principle of FIG. 5, as shown in FIG. 8, the reference corner K actual gripping position comprises _R H _R center O window frame through _R M _R of When the window frame M _D passing through the center O _D of the original holding position H _D comprises detection corners K _D respectively set, the window frame M _R, the deviation amount ΔX as data in M _D match, This is obtained by converting ΔY and Δθ into the material coordinate system.

  The robot control means 30G corrects the posture of the robot so that the detected image DW and the reference image RW coincide with each other based on the amount of deviation represented by (9) to (11) (FIG. 1), and positions the workpiece appropriately. Hold the position.

  In this case, in the bending system having the carry-in / out robot 3 and the bending robot 11 as shown in FIG. 1, the robot control means 30 </ b> G controls both the robots 3 and 11.

  For example, the robot control means 30G attracts the work on the material table 1 to the carry-in / out robot 3 (step 101 in FIG. 10), and images the corner portion K of the work with the CCD camera 2A as the detection image DW. After taking in and comparing with the reference image RW and calculating the predetermined shift amounts ΔX, ΔY, Δθ (steps 102, 103 in FIG. 10), the posture is corrected based on the shift amounts ΔX, ΔY, Δθ. It is determined whether the robot 3 is a loading / unloading robot 3 or a bending robot 11 (step 104 in FIG. 10).

  When the robot controller 30G determines that the loading / unloading robot 3 corrects the posture (down arrow in step 104 in FIG. 10), the loading / unloading robot 3 corrects the workpiece transfer position to the bending robot 11. (Step 105 in FIG. 10).

That is, when the robot control unit 30G transmits the correction drive signals S _a , S _b ... To the loading / unloading robot 3 based on the deviation amounts ΔX, ΔY, Δθ of (9) to (11), the loading / unloading is performed. robot 3, by correcting the posture to grasp at the original holding position H _D the adsorbed workpiece from the actual gripping position H _R, the correction of the workpiece handover position (step 105 in FIG. 10).

In other words, the loading and unloading robot 3, as the actual gripping position H _R of the workpiece itself is holding consistent with (FIG. 6) the original holding position H _D, steps of correcting the posture (FIG. 10 105) The workpiece is delivered to the bending robot 11 with the posture (step 106 in FIG. 10).

  When the robot control unit 30G determines that the bending robot 11 corrects the posture (right arrow in step 104 in FIG. 10), the bending robot 11 corrects the workpiece receiving position from the loading / unloading robot 3. Control is performed (step 110 in FIG. 10).

That is, when the robot control means 30G transmits the correction drive signals S _a , S _b ... To the bending robot 11 based on the deviation amounts ΔX, ΔY, Δθ of (9) to (11), the bending robot 11 but by correcting the posture to receive work loading and unloading robot 3 is gripped at the original holding position h _D, corrects the workpiece receiving position (step 110 in FIG. 10).

In other words, the bending robot 11, if it is assumed that grips the workpiece, so that the actual gripping position h _R of the workpiece at that time coincides with (FIG. 6) the original grip position h _D, corrects the posture (Step 110 in FIG. 10), and the workpiece is received from the carry-in / out robot 3 in the posture (Step 111 in FIG. 10).

  Thereafter, in either case (steps 105 to 106 in FIG. 10 or steps 110 to 111), the bending robot 11 holds the appropriate position of the workpiece (step 107 in FIG. 10), so that the robot control means 30G re-bends. If the robot 11 is driven and controlled, the bending robot 11 supplies and positions the workpiece holding the appropriate position to the bending apparatus 10 (step 108 in FIG. 10), and is controlled under the control of the bending control means 30H (FIG. 1). Next, bending is performed (step 109 in FIG. 10).

  The bending control means 30H (FIG. 1) controls the press brake, which is the bending apparatus 10, based on the bending order determined by the information calculation means 30B, and the workpiece whose proper position is gripped by the bending robot 11. Then, bending is performed by the punch P and the die D (step 109 in FIG. 10).

  The conveyance control means 30J (FIG. 1) controls the conveyance of the material table 1, the intermediate table 8, and the product loading table 12, for example, drives and controls the belt conveyor constituting the product loading table 12.

  The input / output means 30K (operation box) is provided, for example, in the vicinity of the upper table 13 constituting a press brake (FIG. 2), and is composed of a screen such as a keyboard and a liquid crystal.

  This input / output means 30K has an interface function to the above-described upper NC apparatus 29 (FIG. 1), and by connecting the lower NC apparatus 30 to the upper NC apparatus 29 by wire or wirelessly, the CAD information Can be entered.

  The flow of data of the control device of the bending system having the above-described configuration is shown in FIG. 9, which is an example in the case where the loading / unloading robot 3 (loading robot) described above corrects the posture (FIG. 9). 10 step 104 down arrow, steps 105-106).

That is, from the image sensor control (2) in FIG. 9 (corresponding to the imaging control means 30C, work image detection means 30D, work reference image calculation means 30E, and deviation amount calculation means 30F in FIG. 1) to the image sensor (FIG. When the image data request is issued to the workpiece image pickup means 2 (corresponding to the workpiece imaging means 2 of 1), the image sensor is configured to select the corner portion of the workpiece on the workpiece picking device (corresponding to the material stand 1 of FIG. 1). sends K (FIG. 1) captured by it the image data (corresponding to the detected corner portion K _D of FIG. 6) to the image sensor control, already described displaced amount [Delta] X, [Delta] Y, the [Delta] [theta] (FIG. 6) is calculated The

The image data transmitted at this time, as shown in FIG. 9, is transferred to the monitor (built in the deviation amount calculating means 30F in FIG. 1), the target value (corresponding to the reference corner K _R of FIG. 6) A match / mismatch is always monitored.

In this state, the posture correction data is transferred from the loading robot controller (4) in FIG. 9 (built in the robot controller 30G in FIG. 1) to the sequencer (3) in the robot controller 30G in FIG. When a request is issued, if the sequencer issues a displacement amount data request to the image sensor control to transfer the displacement amount data, converts it into posture correction data, and transmits it to the loading robot controller, the loading robot The control device transmits the operation command output (corresponding to the correction drive signals S _a , S _b ... In FIG. 1) whose posture has been corrected to the loading robot (corresponding to the loading / unloading robot 3 in FIG. 1).

  The operation of the first embodiment of the present invention having the above-described configuration (bending system having the bending robot 11 and the carry-in / out robot 3 shown in FIG. 1) will be described below with reference to FIG. In FIGS. 10 and 11 below, the object to be imaged by the CCD camera 2A is the corner K of the workpiece, and the reference image RW is an image of the workpiece at the normal position captured by the CCD camera 2A. The reference image RW stored in the workpiece reference image calculation / storage means 30E is used.

(1) Operations until calculating the predetermined deviation amounts ΔX, ΔY, Δθ.

In step 101 of FIG. 10, the loading / unloading robot 3 sucks the workpiece on the material table 1, and in step 102, the corner portion K of the sucked workpiece is imaged by the CCD camera 2A, and this is captured as a detection image DW. in step 103, it calculates a detection corners K _D in the detected image DW, based on the difference between the reference corner K _R of the reference image RW, the predetermined deviation amount [Delta] X, [Delta] Y, the [Delta] [theta].

That is, the CPU 30A (FIG. 1) causes the loading / unloading robot 3 (FIG. 4) to pick up the workpiece as the material T on the material table 1 via the robot control means 30G, and the workpiece is placed above the CCD camera 2A. The workpiece corner portion K is imaged by the CCD camera 2A via the imaging control means 30C, the imaged corner portion K is detected as an image by the workpiece image detection means 30D, and the deviation amount calculation means through 30F, the detection corners _{K D,} the difference between the reference corner _{K R} stored in the work reference image calculating and storing unit _{30E PIC - X, PIC - Y} , PIC - θ ( previously described (1) ), Finally, the shift amounts ΔX, ΔY, Δθ that can be expressed by the above (9) to (11) are calculated.

(2) Robot posture correction operation.

  In step 104 of FIG. 10, it is determined whether the posture correction is performed by the loading / unloading robot 3 or the bending robot 11. When the loading / unloading robot 3 performs posture correction (down arrow in step 104), step 105 is performed. Then, the loading / unloading robot 3 corrects the workpiece delivery position to the bending robot 11, and in step 106, the loading / unloading robot 3 delivers the workpiece to the bending robot 11 with the corrected posture, and the bending robot 11 corrects the posture. When performing (right arrow of step 104), in step 110, the bending robot 11 corrects the work receiving position from the loading / unloading robot 3, and in step 111, the bending robot 11 carries in / out with the corrected posture. A workpiece is received from the robot 3.

  That is, the CPU 30A (FIG. 1), when the deviation amounts ΔX, ΔY, and Δθ are calculated in FIG. 10 (FIG. 6) (FIG. 6), instructs the robot control means 30G to correct the posture. It is determined whether the loading / unloading robot 3 or the bending robot 11 is used.

When the robot controller 30G determines that the carry-in / out robot 3 corrects the posture, the correction drive signal S _a , based on the deviation amounts ΔX, ΔY, Δθ of (9) to (11). When sent to the robot 3 are carried into and out S _b · · ·,該搬and out robot 3, as described above, the actual gripping position H _R of the workpiece itself is holding (FIG. 6) the original grip position to match the H _D, because to correct the posture and passes the workpiece to remain bending robot 11 of this posture.

In addition, when the robot control unit 30G determines that the bending robot 11 corrects the posture, the correction drive signals S _a and S based on the deviation amounts ΔX, ΔY, and Δθ of (9) to (11). sending _b · · · bending robot 11, the bending robot 11, as already described, if it is assumed that grips the workpiece, the actual gripping position h _R of the workpiece at that time (Fig. 6) to match the original grip position h _D, because to correct the attitude, receiving the workpiece from the robot 3 loading and unloading remain in that position.

(3) Bending operation.

  As described above, in any case (steps 105 to 106 in FIG. 10 or steps 110 to 111 in FIG. 10), the workpiece positioning is performed regardless of whether the posture is corrected by the loading / unloading robot 3 or the bending robot 11. Since the bending robot 11 grips the appropriate position of the workpiece (step 107 in FIG. 10), if the robot control unit 30G drives and controls the bending robot 11 again, as described above, the bending robot 11 Then, the workpiece holding the proper position is supplied to and positioned in the bending apparatus 10 (step 108 in FIG. 10), and bending is performed under the control of the bending control means 30H (FIG. 1) (step 109 in FIG. 10). ).

  FIG. 11 is a diagram showing a second embodiment of the present invention.

The bending system shown in FIG. 11 has a workpiece imaging means 2 attached to an intermediate table 8.

  That is, after bending, the corner portion K of the workpiece as the product Q placed on the intermediate table 8 (FIG. 11) via the bending robot 11 (FIG. 1) is imaged by the CCD camera 2A, and the same is taken. The detected image DW (corresponding to FIG. 6) is compared with the reference image RW, and predetermined deviation amounts ΔX, ΔY, Δθ are calculated based on the difference between the two, and based on the deviation amounts ΔX, ΔY, Δθ, intermediate The position of the product Q on the table 8 when the loading / unloading robot 3 sucks and loads it on the product loading table 12 is corrected. In particular, the product Q is loaded in multiple stages without breaking on the product loading table 12. It is effective when you do.

    The operation of the second embodiment shown in FIG. 11 will be described with reference to FIG.

(1) Operations until calculating the predetermined deviation amounts ΔX, ΔY, Δθ.

The loading / unloading robot 3 sucks the workpiece as the product Q on the intermediate platform 8 (corresponding to step 101 in FIG. 10), and images the corner portion K of the sucked workpiece with the CCD camera 2A, which is used as the detection image DW. uptake (equivalent to step 102), calculates a detection corners K _D in the detected image DW, based on the difference between the reference corner K _R of the reference image RW, the predetermined deviation amount [Delta] X, [Delta] Y, the [Delta] [theta] (step 103).

That is, the CPU 30A (FIG. 1) causes the work as the product Q on the intermediate platform 8 to be attracted to the loading / unloading robot 3 shown in FIG. 11 via the robot control means 30G, and the work is placed above the CCD camera 2A. The workpiece corner portion K is imaged by the CCD camera 2A via the imaging control means 30C, the imaged corner portion K is detected as an image by the workpiece image detection means 30D, and the deviation amount calculation means through 30F, the detection corners _{K D,} the difference between the reference corner _{K R} stored in the work reference image calculating and storing unit _{30E PIC - X, PIC - Y} , PIC - θ ( previously described (1) ), Finally, the shift amounts ΔX, ΔY, Δθ that can be expressed by the above (9) to (11) are calculated.

(2) Robot posture correction operation.

  Thereafter, in the case of FIG. 11, only the loading / unloading robot 3 corrects the posture, and the loading / unloading robot 3 corrects the workpiece loading position on the product loading table 12 (FIG. 10). Step 104⇒Step 105 or 110).

  That is, when the deviation amounts ΔX, ΔY, Δθ are calculated (FIG. 6) (FIG. 1), the CPU 30A issues an instruction to the robot control means 30G to correct the posture of the loading / unloading robot 3.

More specifically, the robot control means 30G sends the correction drive signals S _a , S _b ... On the basis of the deviation amounts ΔX, ΔY, Δθ of (9) to (11) (FIG. 1). sending to 3, the該搬and out robot 3, as the actual gripping position H _R of the workpiece itself is holding consistent with (FIG. 6) the original holding position H _D, corrects the posture.

  (3) Product Q loading operation on the product loading table 12.

As described above, loading and unloading robot 3, as the actual gripping position H _R of the workpiece itself is holding consistent with (FIG. 6) the original holding position H _D, by correcting the posture, the workpiece Is carried out, and the loading / unloading robot 3 grips the appropriate position of the workpiece as the product Q (corresponding to step 106 or 111 → step 107 in FIG. 10), so that the positional deviation of all products Q is corrected. And gripped in the state of facing the same position.

Therefore, if the robot control means 30G drives and controls the loading / unloading robot 3 again (FIG. 11), the loading / unloading robot 3 (see FIG. 11) keeps the workpiece as the product Q holding the appropriate position in the corrected posture in the product loading table If the product Q is loaded on the product loading table 12, the product Q can be loaded in multiple stages without breaking on the product loading table 12 (step 107 → step 108 in FIG. 10).

  As described above, according to the present invention, when a corner portion of a workpiece as a product placed on the intermediate platform is imaged by the CCD camera, the loading / unloading robot 3 loads the product on the product loading platform. The position can be corrected, and there is an effect that the product can be stacked in multiple stages without breaking on the product loading table.

The present invention is used when correcting the position of the product Q on the intermediate table 8 when the carry-in / out robot 3 sucks and loads the product Q on the product loading table 12, and the product Q is broken on the product loading table 12. This is useful when loading in multiple stages.

1 is an overall view showing a first embodiment of the present invention. It is a perspective view of the bending process system in FIG. It is a side view of the bending process system in FIG. FIG. 2 is a detailed view of the workpiece imaging means 2 in FIG. 1. It is a figure which shows the principle of the image processing in FIG. It is a figure which shows the specific example of deviation | shift amount calculation in FIG. FIG. 7 is an enlarged view of FIG. 6. FIG. 6 is a correspondence diagram between the specific example of FIG. 6 and the principle of FIG. 5. It is a figure which shows the flow of the data in FIG. It is a flowchart for demonstrating the operation | movement in FIG. It is a figure which shows 2nd Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Material stand of workpiece picking device 2 Work imaging means 3 Loading / unloading robot 3A Suction pad 8 Intermediate stand 10 Bending device 11 Bending robot 12 Product loading stand 13 Upper table 14 Lower table 15 Base plate 17 Grip 18 Grab change stand 19, 21, 23 Other bending robot 20 Other material base 22 Belt conveyor 29 Upper NC device 30 Lower NC device 30A CPU
30B Information calculation means 30C Imaging control means 30D Work image detection means 30E Work reference image calculation / storage means 30F Deviation amount calculation means 30G Robot control means 30H Bending control means 30J Input / output means T Material Q Product S Imaging range DW Work detection image original actual gripping position _{h D} grip 17 of the original grip position _{h R} grip 17 of the actual gripping position _{H D} suction pads 3A of the reference image _{K D} detecting corner _{K R} reference corner _{H R} suction pad 3A of RW workpiece Gripping position K Work corner

Claims (1)

An intermediate platform for placing workpieces as products,
A workpiece imaging means attached to the intermediate platform;
Workpiece image detection means for detecting an image of a workpiece as a product input from the workpiece imaging means;
A workpiece reference image calculation / storage unit that calculates a reference image of a workpiece as a product based on information input in advance, or stores a workpiece image as a product at a normal position captured by the workpiece imaging unit as a reference image;
A deviation amount calculating means for calculating a predetermined deviation amount based on a difference between the detected image and the reference image;
Based on the amount of deviation, correct the posture of the loading / unloading robot that grips the workpiece as a product on the intermediate platform so that the detected image matches the reference image, and load the workpiece as a product in the corrected posture. A workpiece positioning device comprising a robot control means for loading on a table.

Images