JPH04250972A - Control method for robot - Google Patents

Abstract

PURPOSE:To simplify data for control of transferring working means by memorizing divided detecting points in the control part of a transferring means of the working means, and transferring the working means to the position of the detecting point of a part position to be worked detected through the detecting means. CONSTITUTION:Mark detecting robots R1,R2,R3 and grinding robots R4,R5,R6 are formed into a corresponding same sort of machine. Further, non-contact type detecting means such as mark detecting cameras are arranged, respective faces to be worked are divided into a plurality of detecting points based on the detected positions of the detecting means, and these detecting points are memorized in the grinding robots controlling part of a transferring means of a working means. Hereafter, a part position to be worked is detected by the detecting means at every detecting point, after detecting the part position to be worked the position of the detecting point of which the part position to be worked is detected based on the memory data is discriminated, and the working means is moved to the detecting point.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention detects the machining area of each machining surface of the workpiece on the upstream side for a workpiece having multiple machining surfaces such as an automobile body, and the detected machining area on the downstream side. This invention relates to a method of controlling a robot that moves and processes.

0002

2. Description of the Related Art Usually, a robot that processes a workpiece is controlled to operate according to teaching based on teaching data. Furthermore, in order to accurately process the processing area using the above-mentioned robot, it is known that a camera installed upstream of the workpiece conveyance path detects the amount of shift of the workpiece and corrects the movement of the robot.

[0003] On the other hand, in order to remove paint defects that have occurred on each painted surface of an object to be painted (workpiece) such as an automobile body that has been undercoated or intermediate coated, a worker visually inspects the transported workpiece and detects any defects. An automatic water grinding method has been proposed in which a robot on the downstream side automatically grinds water by inputting a defective part and a defective state through an indicating device (Japanese Patent Laid-Open No. 58-64517).

0004

[Problems to be Solved by the Invention] By the way, a robot equipped with a camera is used upstream of the conveyance path to detect the processing area and transfer the data to the downstream robot, and based on this transferred data, the downstream robot One possibility is to process the detected machining part, but in this case, for example, if the upstream robot and downstream robot are of different models, data processing such as coordinate data transferred from the upstream robot to the downstream robot may be difficult. It becomes very complicated. Furthermore, if the downstream robot is configured to move to the processing area based on data from the upstream robot, the workpiece and the robot arm may interfere if the downstream robot's arm posture is different from that of the upstream robot. There is. For this reason, data for controlling the posture of the robot's arm is also required, but data conversion becomes complicated.

[0005] Furthermore, in order to detect the machined part with the camera of the upstream robot and capture the position of the machined part as x, y, z coordinates, it is necessary to consider the distance between the camera and the machined part. In order to perform the operation, there are difficulties in terms of camera position accuracy, and data processing is complicated. Furthermore, in order to perform proper machining, it is necessary to obtain data while also considering the machining posture of the machining section.

[0006] Therefore, the problem is how to handle the data contents of upstream robots and downstream robots, and how to control the downstream robots.

On the other hand, in the automatic water grinding method disclosed in Japanese Patent Application Laid-Open No. 58-64517, each time an operator discovers a defective part, the operator inputs the position and condition of the defective part into an indicating device. Normally, the location of the above-mentioned defective part is indicated by coordinate data, and the robot serving as the water-sharpening device is controlled based on the above-mentioned coordinate data, so data processing takes time, reducing the time for water-sharpening. becomes a hindrance.

SUMMARY OF THE INVENTION The present invention solves the above problems, and aims to provide a robot control method that can simplify data processing for machining and shorten machining time.

[0009]

[Means for Solving the Problems] In order to achieve the above object, the present invention provides a processing method in which a processing section of a workpiece transported from upstream is processed by processing means, in which each processing surface is set at a plurality of detection points. These detection points are stored in the control unit of the moving means of the processing means, and the non-contact detection means detects the processing region for each detection point, and the position of the detection point where the processing region is detected. After determining this, the moving means is controlled to move the processing means to the detection point.

[0010] Furthermore, in claim 2, the range of the detection point is made to coincide with the detection range of the detection means.

Furthermore, in the third aspect of the present invention, the control section of the moving means stores the processing posture together with the detection point.

0012

[Operation] According to the method for controlling a robot according to claim 1,
Each machined surface is divided into multiple detection points, the machined part is detected at each detection point by a non-contact detection means such as a camera, the position of the detection point where the machined part is detected is determined, and the machined part is detected at the detection point. The moving means is controlled to move the processing means, and the processing portion is processed.

[0013] Furthermore, according to the processing method of claim 2,
The range of the detection points may match the detection range of the detection means.

Furthermore, according to the robot control method of the third aspect, the machining posture at each detection point changes depending on the shape of each machining surface of the workpiece.

[0015]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows an embodiment of the present invention in an automobile manufacturing line.

Processing line L consists of an inspection marking process L1, a mark detection process L2, a polishing process L3, and a water washing process L4 along the conveyance line 2 from upstream. The paint defects on the automobile body 1 transported by the transport line 2 are detected and corrected by water polishing.

A plurality of inspectors M1 and M2 are placed at positions corresponding to the inspection marking step L1, and visually inspect the painted surface of the automobile body 1. and,
When the inspectors M1 and M2 find a defective part of the paint such as a pinhole or foreign matter on the painted surface, they select one of several patterns of stamps, etc. that are preset according to the degree of processing (polishing).
A mark with a pattern corresponding to the degree of processing of the processed part is selected and attached to the defective part. In other words, as shown in Table 1, for example, a round mark is placed on an area that is highly defective and requires heavy polishing, a triangular mark is placed on an area that requires medium polishing, and a square mark is placed on an area that requires light polishing. I am trying to add a mark.

[0018]

[Table 1]

Note that the color of the mark may be changed, for example, to red, green, or blue, depending on the degree of processing. Further, the shape of the mark itself may be the same, but the size may be changed depending on the degree of processing.

[0020]Also, even if the processing degree is 4 or more stages, 2
It may be below the stage. In this case, the types of shapes etc. will be increased or decreased depending on the number of stages.

Vehicle type detection sensors P1, P2, and P3 are provided upstream at a position corresponding to the mark detection step L2, and
Mark detection robots R1, R2, and R3 are arranged downstream so as to face the left, right, and top surfaces of the automobile body 1, respectively. Each mark detection robot R1, R2, R3 has a mark detection camera S1, S2, S, which is a non-contact type sensor.
3 are respectively provided to detect the position, shape, etc. of the mark attached to the automobile body 1 and output it as an image signal.

Polishing robots R4, R5, and R6 are arranged at positions corresponding to the polishing process L3 so as to face the left and right surfaces and the top surface of the automobile body 1, respectively. Tools T1, T2, T3
are provided for each. Then, mark detection step L
Each polishing robot R4, R5,
R6 guides the polishing tools T1, T2, and T3 to the position of the mark and performs water polishing.

A plurality of flushing showers 3 and flushing brushes 4, which are controlled by a flushing control panel 7, are provided at a position corresponding to the flushing step L4, and are used to rinse the automobile body 1 after water polishing.

Next, the control system of the processing line L will be explained using FIG. 2.

The mark detection step L2 includes a mark detection section 51.
, 52, 53, the vehicle type detection section 54, and the detection robot control section 5.
A detection control panel 5 having 5, 56 and 57 is provided. The mark detection sections 51, 52, and 53 detect marks attached to the automobile body 1 from image signals from the mark detection cameras S1, S2, and S3, respectively. That is, a plurality of detection points whose ranges are determined based on the detection ranges of mark detection cameras S1, S2, and S3 are set on each painted surface of the automobile body 1, respectively.
The mark detection units 51, 52, and 53 are configured to store data including the number of the detection point of the detected mark and the coordinates of the mark within the detection point. Further, the mark detection units 51, 52, and 53 determine the degree of machining from the shape of the mark, etc., and store it as polishing grade data, and when the detection work is completed, these data are transferred to the polishing control panel 6, which will be described later. ing.

The vehicle type detection section 54 includes vehicle type detection sensors P1 and P.
2, the vehicle type of the automobile body 1 is detected based on the detection signal from P3, and outputted to the detection robot control sections 55, 56, and 57 as vehicle type data. The detection robot control sections 55, 56, and 57 control the mark detection robots R1, R2, and R3 based on the vehicle type data from the vehicle type detection section 54, respectively, in accordance with the size, shape, etc. of the automobile body 1.

In the polishing step L3, the polishing robot controller 61
, 62, 63 and polishing tool control sections 64, 65, 66. The polishing robot control units 61, 62, and 63 control the polishing robot R4,
This is to control R5 and R6 respectively. That is,
The polishing robot control units 61, 62, and 63 store the detection points corresponding to each painted surface of the automobile body 1, and select the corresponding detection points according to the data from the mark detection units 51, 52, and 53. I'm trying to move to. Furthermore, since the shape of each painted surface of the automobile body 1 is not flat, the polishing robot control sections 61, 62, and 63 are configured to memorize the processing posture at each detection point along with the detection points described above.

The polishing tool control units 64, 65, and 66 control the polishing tools T1, T2, and T3, respectively, based on the polishing grade data from the detection control panel 5. That is, each of the polishing tools T1, T2, and T3 wet-polishes the paint defective portion of the automobile body 1 with polishing time and polishing intensity depending on the degree of polishing.

The production information network terminal 8 outputs data from a production management computer (not shown) to the detection control panel 5 and polishing control panel 6.

[0030] Furthermore, the mark detection robots R1 and R2
, R3 and the polishing robots R4, R5, and R6 are of the same type as shown in FIG. 3, for example, in order to simplify data processing. That is, if the models are of the same type and have the same control data, they will perform the same operation. Therefore,
By teaching the mark detection robots R1, R2, and R3 according to the undulations of the painted surface of the automobile body 1, and using this teaching data as is as operation control data for the polishing robot control units 61, 62, and 63, The polishing robot R moves to the detection point of the processing area detected by the mark detection cameras S1, S2, and S3.
4, R5, and R6 can be moved.

Here, mark detection robots R1, R2,
An example of data processing operations between R3 and polishing robots R4, R5, and R6 will be described.

That is, for example, mark detection robot R1
During the teaching work, as shown in Table 2, coordinate data (x, y, z) for each detection point, posture data (α, β, γ) of mark detection camera S1, and each arm axis R
11, R12, R13 (Fig. 3) angle data (θ1, θ
2, . After that, the floppy is transferred to the polishing robot controller 61.
Load it into After starting the detection work, only the number data of the detection point to be processed is transferred from the detection robot control section 55 to the polishing robot control section 61, and the polishing robot R4 moves to the detection point.

[0033]

[Table 2]

On the other hand, mark detection robots R1, R2, R
3 and the polishing robots R4, R5, and R6, coordinate data (x
, y, z) and the posture data of mark detection camera S1 (α,
β, γ) and the angle data (θ1, θ2, . . . , θ4) of each arm axis R11, R12, R13 may be transferred. In this case, although the amount of transferred data increases, there is no need to register detection point data in advance.

[0035] Furthermore, the above-mentioned mark detection robot R1,
The teaching data of R2 and R3 is data of two-dimensional detection points, and the polishing robots R4, R5, and R6 are within the operating range of the mark detection robots R1, R2, and R3.
It is controlled only by where the detection point is located on a plane. Therefore, the polishing robots R4, R5, and R6 can operate very easily and accurately.

Note that the mark detection process L2 and the polishing process L3
are respectively provided with absolute reference positions for processing positioning of the automobile body 1, and the mark detection robot R1
, R2, R3 and the polishing robots R4, R5, R6 have their coordinates set based on the above absolute reference position. Therefore, if the coordinates are the same, the mark detection robots R1, R2, R3 and the polishing robots R4, R5, R6 are designed to move to the same location.

Next, the operations of the mark detection step L2 and polishing step L3 will be explained using the flowcharts of FIGS. 4 and 5. Note that the flowchart in FIG. 4 shows the operation of the mark detection step L2, and the flowchart in FIG. 5 shows the operation of the polishing step L3.

After a paint defect is found and marked by the inspectors M1 and M2 in the marking process L1, the automobile body 1 is conveyed along the conveyance line 2 to the mark detection process L2 and is detected by vehicle type detection sensors P1, P2, The vehicle type is detected by P3 (step S1), and a control program for mark detection robots R1, R2, and R3 is set according to the vehicle type (step S2).

Next, it is determined whether the car body 1 is set at the detection work position (step S3), and if the car body 1 is at the detection work position (YES at step S3), the mark detection robots R1, R2 , R3 are activated, and mark detection is started by mark detection cameras S1, S2, and S3 (step S4). That is, in the case of the mark detection robot R3, the mark detection camera S3 detects marks in the order of the hood surface, the roof surface, and the trunk surface. Further, for example, on the bonnet surface 11,
As shown in , a plurality of detection points 1a, 1b, ..., 1c, ... are set so as to partition the bonnet surface 11, and the mark detection robot R3 uses a mark detection camera S3.
Accordingly, mark detection is performed along the bonnet surface 11 in the order of the detection points 1a, 1b, . . . , 1c, . Note that detection points are also set on the roof surface and the trunk surface, respectively, and the mark detection robot R3 detects marks along the roof surface and the trunk surface.

[0040] When a mark indicating a polished area is detected (YES in step S5), the data of the detection point number, coordinates, and polishing grade of the mark are stored (step S6). The mark detection described above continues until all detection points are completed, and when mark detection is completed at all detection points (YES in step S7), mark detection robots R1, R2, and R3 return to their original positions (step S8).
As shown in FIG.
(step S9).

After this, the automobile body 1 undergoes a polishing process L.
3 (YES in step S11 of FIG. 5) and set at the polishing work position (YES in step S12).
, polishing robots R4, R5, and R6 start up (
Step S13). Each of the polishing robots R4, R5, and R6 moves to the center of the detection point with the smallest number among the detection points of the mark detected in the mark detection step L2 (step S14), and further moves to the coordinates of the mark ( Step S15). That is, as shown in FIG. 6, the mark M is located at the detection point 1c on the bonnet surface 11.
When the polishing robot R6 is attached to the detection point 1
Move to the center O of c, and further move to the coordinates of mark M shown in Fig. 8 (
x1, y1).

Next, a polishing control program for controlling the degree of polishing such as polishing time and polishing intensity is set based on the polishing grade data (step S16), and the polishing tool T1
, T2, and T3 according to the processing degree (step S17).

[0043] When the water polishing process at all detection points is completed (YES in step S18), the polishing robot R
4, R5, and R6 return to their original positions (step S19).

Next, the polishing robots R4, R5, R6
The polishing operation will be explained using FIGS. 9 and 10.

The polishing robots R4 and R5 are arranged on the left and right sides of the automobile body 1, respectively, and perform hydro-polishing in the order of the front fender surface, front door surface, rear door surface, and rear fender surface as shown in FIG. . On the other hand, the polishing robot R6 performs water polishing on the bonnet surface 11, the roof surface 12, and the trunk surface 13 in this order, as described above.

The front fender surface, front door surface, rear door surface, and rear fender surface have small undulations between each surface, so they can be treated as substantially the same two-dimensional surface. That is, even if the polishing robots R4 and R5 move in parallel from one surface to the other, contact with the automobile body 1 does not occur.

On the other hand, the bonnet surface 11 and the roof surface 1
2. As shown in FIG. 9, the trunk surface 13 has a fairly large difference in height between the surfaces, so the polishing robot R6 detects the mark detected on the roof surface 12, for example, after finishing the water polishing process on the bonnet surface 11. If you try to move directly to the point, there is a risk of contact with the car body 1. On the other hand, it is conceivable to move from the bonnet surface 11 to the roof surface 12 along the undulations of the pillars of the automobile body 1 after finishing the hydro-polishing process of the bonnet surface 11, but this would take a long time to move.

For this reason, as shown in FIG. 10, a plurality of interference avoidance points 14 and 15 are provided above the automobile body 1, and when the polishing robot R6 moves from the bonnet surface 11 to the roof surface 12, the bonnet surface 11 to the interference avoidance point 14 and then to the roof surface 12, and when moving from the roof surface 12 to the trunk surface 13, the vehicle moves from the roof surface 12 to the interference avoidance point 15 and then moves to the trunk surface 13. do it like this.

An example of the operation of moving the polishing robot R6 to the interference avoidance point will be described below with reference to the flowcharts of FIGS. 11 and 12. Note that each detection point on the bonnet surface 11 is numbered 1 to 49 on the roof surface 1.
It is assumed that each detection point on the trunk surface 13 is numbered from 50 to 99, and each detection point on the trunk surface 13 is numbered from 100 to 149.

The automobile body 1 is transported from the mark detection process L2 to the polishing process L3 (YES in step S21),
When set to the polishing work position (YES in step S22)
S), the polishing robot R6 starts (step S23),
The interference avoidance flags FF1 and FF2 are each reset (step S24). Subsequently, the detection points at which marks are detected are read out in numerical order from the data from the mark detection step L2, and the mark coordinate data and polishing grade data related to the detection points are read out (step S25).

Next, it is determined whether the number of the detection point is the number of the detection point on the bonnet surface 11, that is, a number between 1 and 49 (step S26). (Step S26
(YES), the polishing robot R6 moves to the center O (FIG. 8) of the detection point (step S27), further moves to the coordinates of the mark M (FIG. 8) (step S28),
Water polishing is performed using the polishing tool T3 according to the degree of polishing (step S29). The water polishing process continues until the water polishing process is completed at all detection points on the bonnet surface 11 (NO in step S30), and if there is no mark on any detection point other than the bonnet surface 11 (YES in step S30), the polishing robot R6 It returns to its original position (step S31). On the other hand, after the water grinding process is completed on the bonnet surface 11, a mark is added on the roof surface 12 or the trunk surface 1.
3, the number of these detection points is 50 or more, so the answer in step S26 is NO and the process moves to step S32 in FIG. Then, it is determined whether the number of the detection point is the number of the detection point on the roof surface 12, that is, a number between 50 and 99 (step S32), and if the number is between 50 and 99 (YES at step S32). , it is determined whether the interference avoidance flag FF1 is set (step S33). At this time, since the interference avoidance flag FF1 is not set (NO in step S33), the interference avoidance point 14 is set in the polishing robot control unit 63 (step S34), and after the polishing robot R6 moves to the interference avoidance point 14. (Step S35), and the interference avoidance flag FF1 is set (Step S36). Next, the process returns to step S27, where the polishing robot R6 moves to the corresponding mark on the roof surface 12, and water polishing is performed (
Steps S27 to S29).

[0052] After the water polishing of the roof surface 12,
Furthermore, if there is a mark on the roof surface 12, the interference avoidance flag FF1 has already been set in step S36 (
If YES in step S33), the polishing robot R6 moves directly to the corresponding mark portion on the roof surface 12, and water polishing is performed (steps S27 to S29).

On the other hand, if there is a mark only on the trunk surface 13, or if there is a mark on the trunk surface 13 after the water polishing of the roof surface 12 is completed, the detection point of the mark will be a number between 100 and 149. (Step S
37: YES), it is determined whether the interference avoidance flag FF1 is set (step S38). That is, for example, if there is no mark on the roof surface 12 (NO in step S32), the interference avoidance flag FF1 is not set in step S36 (NO in step S38). The interference avoidance point 14 is the polishing robot control unit 6
3 (step S39), and the polishing robot R6 moves from the bonnet surface 11 to the interference avoidance point 14 (
Step S40), and the interference avoidance flag FF1 is set (Step S41).

Next, it is determined whether the interference avoidance flag FF2 is set (step S42). Since the interference avoidance flag FF2 remains reset in step S24, the result is NO in step S42, the interference avoidance point 15 is set in the polishing robot control unit 63 (step S43), and the polishing robot R6 moves to the interference avoidance point 15 (step S42). S44), after the interference avoidance flag FF2 is set (step S45), the process returns to step S27 to move to the corresponding marked part on the trunk surface 13, and water polishing is performed (steps S27 to S29).

If there is a mark on the trunk surface 13 after the water polishing of the trunk surface 13, the interference avoidance flags FF1 and FF2 are set at this time, so YES is determined in steps S38 and S42. Without performing the processes of steps S39 to S41 and steps S43 to S45, the process returns to step S27 and moves to the mark site, where water polishing is performed (steps S27 to S29).

In this way, as shown by the two-dot chain line in FIG. 10, the polishing robot R6 immediately moves to the interference avoidance point 14 after the water grinding process on the bonnet surface 11, and then moves to the roof surface 12. After the surface 12 is polished, it immediately moves to the interference avoidance point 15 and then moves to the trunk surface 13. That is, by passing through the interference avoidance points 14 and 15, the polishing robot R6 can polish the bonnet surface 1 without contacting the automobile body 1.
1 or roof surface 12 directly to roof surface 12 or trunk surface 13. Therefore, mark detection robots R1, R2, R3
Similarly, compared to the case where the polishing robot R6 is configured to move to the next processing surface after passing through the entire area of each processing surface, the moving time of the polishing robot R6 can be shortened.

As mentioned above, the polishing robots R4 and R5 have small undulations between the front fender surface, front door surface, rear door surface, and rear fender surface, so there is no need for interference points, and each surface can be directly connected between the surfaces. Can be moved.

Next, the mark detection operations of the mark detection cameras S1, S2, and S3 in the mark detection step L2 will be explained.

In order to prevent marks from being omitted from being detected near the boundaries of the detection points, the mark detection cameras S1, S2, and S3 are configured to detect adjacent detection points on the upper, lower, left, and right sides, respectively, in an overlapping manner. For this reason, the marks in the overlapping portion are detected in duplicate.

For this overlap detection process, as shown in FIG. 13, the overlap portion is divided into a plurality of detection windows and numbers W1 to W8 are assigned to these windows. If a mark exists in these surrounding detection windows W1 to W8, the detection process is performed assuming that a mark exists only in either one of the detection point or the adjacent detection point.

The operation of the above mark detection will be explained below with reference to FIG.
This will be explained using the flowchart of No. 4.

When the automobile body 1 transported from the marking step L1 is set at the detection work position, mark detection is started by the mark detection cameras S1, S2, and S3 (step S51). Then, when the mark is detected (
YES in step S52), the data of the detection point number, coordinates, and polishing grade of the mark are stored, and if the mark is within the peripheral detection windows W1 to W8, the number of the detection window is stored ( Step S53).

On the other hand, if there is no mark (step S52
(NO in step S54), the process moves to step S54 without performing the process in step S53, and mark detection is continued until mark detection at all detection points is completed (NO in step S54). Then, when mark detection is completed at all detection points (
(YES in step S54), each of the stored data is read out (step S55), and the output flag corresponding to each stored data is set to "1" (step S56). Next, marks in the surrounding detection windows W1 to W8 are searched from the above stored data in the order of detection point numbers (step S57), and if a corresponding mark is found (YES in step S58), the corresponding mark is searched based on Table 3. An adjacent detection point that overlaps the detection point is extracted, and as shown in Table 4, the output flag of the stored data corresponding to the adjacent detection point is reset to "0" (step S59).

[0064]

[Table 3]

[0065]

[Table 4]

When the search for all stored data is completed (YES in both steps S60 and 61), only the data for which the output flag is set from among the stored data is transferred to the polishing control panel 6 (step S62). . After this,
Based on the above transferred data, polishing robot R4,R
5, R6 is moved and water polishing is performed.

Next, step S5 of the above flowchart
The operations from step S7 to step S59 will be explained using FIG. 15 and Tables 3 and 4 using marks A to E as examples. Note that the detection points are only "1" to "9". Also, the coordinates of mark A (80, 50) and the coordinates of mark B (250, 120)
), the coordinates of mark C (200, 235), and the coordinates of mark D (250, 235) each indicate the value at detection point "1", and the coordinates of mark E (200, 235) indicate the value at detection point "2". It shows.

Further, the reference window number indicates the number of the peripheral detection window of the adjacent detection point that overlaps with the corresponding peripheral detection window. That is, for example, if it is the surrounding detection window W2 of detection point "1",
The adjacent detection point is "2", and the number of the overlapping peripheral detection window is W4. Furthermore, the detection window W0 indicates an area that does not overlap with any adjacent detection points.

As shown in FIG. 15, since mark A is located in the detection window W0 of detection point "1", it is detected only at detection point "1". On the other hand, mark B is located in the peripheral detection window W2 of detection point "1" and is detected redundantly at detection point "1" and detection point "2". The mark C is located in the peripheral detection window W3 of the detection point "1", and is detected redundantly at the detection point "1" and the detection point "4". In addition, mark D is located in the surrounding detection window W7 of detection point "1", and detection point "1", detection point "2", and detection point "
4" and detection point "5" are detected redundantly. Furthermore, mark E is located in the surrounding detection window W3 of detection point "2", and detection point "2" and detection point "
5'' is detected redundantly (steps S51 to S54).

[0070] Then, each data of the marks A to E is read out in relation to the number of the detection point, etc. (
Step S55), all output flags of these data are set (Step S56).

As mentioned above, mark A is the detection point "
1" is detected and there is no overlap, so
If NO is determined in step S58, the stored data A of mark A is
The output flag of 1 remains set.

On the other hand, for mark B, the adjacent detection point number "2" of the peripheral detection window W2 of the detection point "1" and the reference window number W4 are extracted from Table 2. That is, since mark B overlaps in the surrounding window number W4 of detection point "2" and the surrounding detection window W2 of detection point "1", the output flag of the stored data B2 of mark B of detection point "2" is The mark B is reset and processed so that it exists only at the detection point "1" (step S59).

Similarly, for mark C, the adjacent detection point number "4" and reference window number W1 are extracted from Table 2, and the stored data C of mark C at detection point "4" is extracted.
The output flag of No. 4 is reset, and the mark C is processed so that it exists only at the detection point "1" (step S
59). Furthermore, for mark E, the adjacent detection point number "5" and reference window number W1 are extracted from Table 2, the output flag of the memory data E5 of mark E at detection point "5" is reset, and mark E is detected. Point "2"
” (step S59)
.

On the other hand, for mark D, the adjacent detection point numbers "2", "4", and "5" of the surrounding detection window W7 of the detection point "1" are extracted from Table 2, and these Reference window numbers W8, W6 and W5
are extracted respectively. Then, based on this extraction result, the detection points "2", "4", and "5" are marked D.
The output flags of the stored data D2, D4, and D5 are reset, respectively, and the mark D is processed so that it exists only at the detection point "1" (step S59).

After this, among the stored data in Table 3, the data A1, B1, C1, D1, whose output flags are set are
E2 is transferred to the polishing control panel 6 (step S62).

In this way, even if duplicate marks are detected, only one piece of data for each mark is transferred to the polishing control panel 6, so that it is possible to prevent the same mark from being wet-polished twice. It can be prevented.

Next, a first embodiment of data processing in the mark detection sections 55, 56, and 57 to enable polishing of a plurality of marks at once will be described with reference to flowcharts shown in FIGS. 16 and 17.

In this embodiment, a plurality of marks included in one polishing range of polishing tools T1, T2, and T3 are extracted by image processing. Note that the shape of the polishing range varies depending on the shape of the polishing tools T1, T2, T3, etc.
For example, the shape is set to be round, square, or oval, and the size of the polishing range changes depending on the degree of processing. Furthermore, the memory area of the image memory is divided according to the degree of processing. Furthermore, mark detection sections 55 and 56
, 57 extracts, for example, a white-painted area on the image memory. Further, the filled-in area indicates a processed portion such as a pinhole.

When the automobile body 1 is set at the detection work position, the mark detection robots R1, R2, and R3 are activated, and mark detection is started by the mark detection cameras S1, S2, and S3 (step S71). When a mark is detected (YES in step S72), the degree of processing is determined from the shape of the mark, etc., and an image memory corresponding to the degree of processing is selected. Subsequently, in the selected image memory, an area having a shape and size corresponding to the polishing range and degree of processing (hereinafter referred to as the mark section) centered around the coordinates of the mark is filled with white (step S73).
.

On the other hand, if there is no mark (step S72
(NO in step S74), the process moves to step S74 without performing the process in step S73, and it is determined whether mark detection has been completed at all detection points (step S74), and if it has not been completed (NO in step S74), Step S7
1 and mark detection continues. When mark detection at all detection points is completed (YES in step S74), each mark portion in each image memory area is labeled (numbered) (step S75). Here, if multiple mark parts are connected to each other, these mark parts are treated as one mark part with a single label (number).
is now attached.

[0081] Subsequently, if the labeled mark part is present (YES in step S76), it is extracted in numerical order (step S77), and the widths of the extracted mark part in the x direction and the y direction are respectively extracted. (step S78, S
79). Next, it is determined whether the widths of the mark portions in the x and y directions are within twice the widths of the polishing range in the x and y directions (step S80). That is,
As mentioned above, the filled-in area is set around machining parts such as pinholes, so if this machining part falls within the polishing range, even if multiple machining parts are processed, water polishing can be performed at once. Can be done. Therefore, 2 of the above polishing range
If it is within the same time (YES in step S80), by aligning the center of gravity of the mark to the center of the polishing range, it is possible to water-grind multiple parts within the mark at once. After the barycenter coordinates of the part are determined on the image memory (step S81), the barycenter coordinates are stored (step S82). Then, if there is another mark part (YES in step S83), step S
Return to 76.

On the other hand, in step S80, the mark part x
If the width in the direction and y direction is twice or more the polishing range (NO in step S80), the mark portion is divided into areas within twice the polishing range on the image memory (step S84 in FIG. 17). . Subsequently, the barycenter coordinates of each of the divided mark portions are extracted and stored (step S85).
- Loop of YES in step S87 and step S88). Then, when the processing of the divided mark parts described above is completed (
(NO in step S88), the process returns to step S76.

[0083] When the processing for all mark portions is completed (NO in step S76), the data of the center of gravity coordinates is transferred to the polishing control panel 6, and the polishing robots R4, R5, and R6 are driven based on the transferred data. Then, water polishing is performed.

Next, a second embodiment of data processing in the mark detection sections 55, 56, and 57 to enable polishing of a plurality of marks at once will be described using the flowchart of FIG.

In this embodiment, the center of gravity of a plurality of marks included in one polishing range of polishing tools T1, T2, T3 is calculated from the coordinates of each mark.

When the automobile body 1 is set at the detection work position, the mark detection robots R1, R2, and R3 are activated, and mark detection is started by the mark detection cameras S1, S2, and S3 (step S91). Then, when a mark is detected (YES in step S92), the x and y coordinates of the mark are extracted (step S93),
The x coordinates are stored in ascending order (step S94).

On the other hand, if no mark is detected (NO in step S92), it is determined whether or not mark detection has been completed at all detection points (step S95), without performing the processing in steps S93 and S94, and the process ends. If not (NO in step S95), the process returns to step S91 and mark detection is continued. Then, mark detection at all detection points is completed (YES in step S95),
If there is no mark at this time (NO in step S96),
The processing of this flowchart ends.

On the other hand, if there is a mark (YES in step S96), the detected marks are read out in order from the one with the smallest x coordinate, and the polishing range is determined with respect to the coordinates of the one read mark. Other marks within the width in the x direction and within the width in the y direction are searched for (step S97). Then, if there are no other marks within the polishing range (step S
98: NO), the coordinates of the mark 1 are stored as they are (step S99). On the other hand, if there are other marks within the polishing range (YES in step S98), these marks can be polished at once, so the x and y coordinates of the centers of gravity of these marks are 2 (step S100), and the calculated coordinate data of the center of gravity is stored (step S101).

[0089]

[Math 1]

[0090]

[Math 2]

[0091] Here, n indicates the number of marks searched in step S97.

[0092] The processes of steps S97 to S101 are repeated until the reading and searching of the x-coordinates for all marks is completed, and when the above processes are completed (YES in step S102), the stored data is polished. The data is transferred to the control panel 6, and polishing robots R4, R5, and R6 are driven based on the transferred data to perform water polishing.

[0093] In the above description, the water training process of the automobile body 1 after painting has been described, but the workpiece may be other than the automobile body 1, or may be a workpiece other than the object to be painted. In addition, it is not limited to water grinding,
For example, the marked positions may be spray painted.

In this way, the mark detection robots R1, R
2. Since the same model robots were used for R3 and the polishing robots R4, R5, and R6, the mark detection robots R1, R
2. Data can be shared between R3 and polishing robots R4, R5, and R6, making data processing easier.

As shown in FIGS. 5 and 8, each painted surface of the automobile body 1 is divided by detection points whose ranges are set based on the detection ranges of the mark detection cameras S1, S2, and S3 in the mark detection step L2. be done. Then, as shown in the flowchart of FIG. 3 above, mark detection step L2
The mark part is detected, and the detection point of the mark is detected by the polishing robot controllers 61, 62, 6 in the polishing process L3.
3, as shown in the flowchart of FIG. 4, the polishing robots R4, R5, and R6 move to the detection point where the mark area is detected, and then move to the polishing area and remove the polishing tools T1, T2. , T3, the data can be simplified compared to the case where the polishing robots R4, R5, and R6 are controlled by specifying the position on the painted surface of the automobile body 1 using coordinate data.

Note that if the polishing ranges of the polishing tools T1, T2, and T3 are configured to almost match the detection ranges of the mark detection cameras S1, S2, and S3, the polishing tools will be placed at the center O of the detection point where the machined part is detected. By simply aligning the centers of T1, T2, and T3, the part to be processed within the detection point can be processed. In this case, step S in FIG.
There is no need to perform the process of moving to the mark coordinates shown in 15.

[0097] Furthermore, the polishing robot controllers 61, 62, 6
3 stores the machining posture at each detection point along with the above detection points, for example, the angle data (θ1, θ2, ..., θ4) of the arm axes R11, R12, R13, so that the polishing tool T1, T2 and T3 are in reliable contact and proper processing can be performed.

[0098]

According to the present invention, a plurality of divided detection points are stored in the control section of the moving means of the processing means, and the processing means is moved to the position of the detection point of the processing part detected by the detection means. Data for controlling movement of the processing means can be simplified. Therefore, data processing can be performed in a short time, and processing time can be shortened. Furthermore, teaching of the robot as a means of transportation can be easily performed. Furthermore, since the processing means corresponds to the detection means, data can be shared between the detection means and the processing means, and data processing can be simplified.

Furthermore, since the range of the detection point is set based on the detection range of the detection means, the detection ability of the detection means can be used efficiently.

Furthermore, since the machining posture is also memorized along with the detection points, proper machining can be performed.

[Brief explanation of the drawing]

FIG. 1 is a layout diagram showing an example of a processing line according to the present invention.

FIG. 2 is a block configuration diagram showing an example of a control system of a processing line.

FIG. 3 is an external view showing a robot according to the present invention.

FIG. 4 is a flowchart showing an example of the operation of a mark detection step.

FIG. 5 is a flowchart showing an example of the operation of a polishing process.

FIG. 6 is an explanatory diagram showing an example of setting detection points.

FIG. 7 is an explanatory diagram showing an example of transfer data.

FIG. 8 is an explanatory diagram showing an example of coordinates of marks within a detection point.

FIG. 9 is an explanatory diagram showing an example of setting detection points on an automobile body.

FIG. 10 is an explanatory diagram showing an example of setting detection points on an automobile body.

FIG. 11 is a flowchart showing an example of the operation of the polishing robot.

FIG. 12 is a flowchart showing an example of the operation of the polishing robot.

FIG. 13 is an explanatory diagram showing the relationship between detection points and detection windows.

FIG. 14 is a flowchart showing an example of mark detection operation.

FIG. 15 is an explanatory diagram showing an example of the relationship between a detection point and an adjacent point.

FIG. 16 is a flowchart showing a first example of an operation for polishing a plurality of marks at once.

FIG. 17 is a flowchart showing a first example of an operation for polishing a plurality of marks at once.

FIG. 18 is a flowchart showing a second embodiment of the operation for polishing a plurality of marks at once.

[Explanation of symbols]

1 Automobile body 2 Transport line 5 Detection control panel 6 Polishing control panel 11 Bonnet surface 12 Roof surface 13 Trunk surface 14, 15 Interference avoidance points 51-53 Mark detection section 54 Vehicle type detection section 55-57 Detection robot control section 61-63 Polishing Robot controllers 64-66 Polishing tool controllers 1a, 1b,..., 1c,... Detection points P1-P3
Vehicle type detection sensor R1-R3 Mark detection robot R4-R6 Polishing robot S1-S3 Mark detection camera T1-T3 Polishing tool

Claims (3)

[Claims] 1. A control method for a robot that processes a processing part of a workpiece transported from upstream by a processing means, wherein each processing surface is divided into a plurality of detection points, and these detection points are connected to a moving means of the processing means. A non-contact detection means corresponding to the processing means is provided, and the detection means detects the processing part at each of the detection points, and the position of the detection point where the processing part is detected. 1. A method for controlling a robot, comprising: determining the detection point, and then controlling the moving means to move the processing means to the detection point. 2. The method of controlling a robot according to claim 1, wherein the range of the detection point is made to coincide with the detection range of the detection means. 3. The method of controlling a robot according to claim 1, wherein the control unit of the moving means stores the processing posture together with the detection point.